CLIENT - Pakistan Poverty Alleviation Fund · Client Attrition in Microfinance: Experience and...

CLIENTATTRITION IN MICROFINANCE:Experience and Practice

Pakistan Poverty Alleviation Fund

RESEARCH & DISCUSSION SERIES

Foreword ..................................................................................................1

Chapter 1: Defining Client Exit ..............................................................2a) Definitional Issues.................................................................................4b) Measuring Client Exit ............................................................................7c) Acceptable Attrition Rates ...................................................................17

Chapter II: Reasons for Client Desertion ............................................18a) Causes of Client Exit...........................................................................20b) Classification/Categorization of Exit Clients:.......................................26c) Socioeconomic Profile of Exit Clients:.................................................26d) Timing of Client Exit: ...........................................................................28e) Limitations of Literature:......................................................................28f) Conclusions:.........................................................................................29

Chapter III: Monitoring Client Exit .......................................................32a) Building Effective Monitoring Systems................................................34b) Institutionalizing the Process ..............................................................45

References .............................................................................................49Annexures ..............................................................................................52Annex 1: Exit Client Case Studies (Pakistan)Annex 2: Practices of PPAF POs (Pakistan)Annex 3: Exit Monitoring Form of Partner (Bosnia-Herzegovina)Annex 4: Exit Monitoring Form of Prizma (Bosnia-Herzegovina)Annex 5: Exit Monitoring Form of AIMS Project (USA)Annex 6: Client Exit Monitoring at SEF (South Africa)

Contents

Pakistan Poverty Alleviation Fund

PPAF has launched a Research and Discussion Series to commemoratecompletion of ten years of its operations. The R&D series aims to foster debateand discussion on poverty and its reduction with special reference to communitydriven grass root development. This publication, the third in the series, focuseson an important dimension of microfinance.

The perceived abundance of demand for micro-loans initially detracted attentionfrom efforts aimed at measuring client attrition. With microfinance marketsachieving a measure of growth and consolidation, the phenomenon of clientexit (and conversely client retention) has received greater attention. Subsequently,literature examining effects of dropout on a range of associated issues suchas financial sustainability, portfolio growth and quality, institutional viability, staffproductivity, risk assessment, as well as socio-economic impact has sinceproliferated. More recently, the microfinance industry has achieved substantialprogress in deciphering factors contributing to client desertion in a variety ofcultural, economic and organizational contexts. However, these advances havebeen researched and adopted with varying degrees of success in differentmarkets at the regional, national and sub-national levels. Pakistan for instanceshows tremendous growth in the sector that is yet to be matched by acorresponding level of research into factors underlying client desertion.

This study seeks to develop an understanding of borrower attrition with particularfocus on the following fundamental questions: What constitutes client exit?Who is an exit client? How can exit clients be monitored? What are the maincauses behind client exit? It attempts to answer these questions in the backdropof institutional best practices, theoretical literature and empirical evidence.

Focused on sixteen partner organizations and eight districts of Punjab andSindh, the study captures diversity of delivery (specialized/multisectoral),geographical concentration (urban/rural) and client focus (male/female). It wascarried out under supervision of Ahmad Jamal (Chief Strategy Officer). Theresearch team was led by Najeebullah Khan (Evaluation, Research &Development unit) who was assisted by Madiha Mumtaz (CSO office) andSameen Shahid (ERD unit). The facilitation and support extended by partnerorganizations is gratefully acknowledged.

Kamal HyatChief Executive/Managing Director

Foreword

1Pakistan Poverty Alleviation Fund

Chapter 1:

Defining Client Exit

recruit new clients all the time, theirefforts can be more meaningful if itcontributes to greater outreachinstead of merely keeping up existinglevels due to high client exits.

As organizations would in most casesprogressively increase loan size withthe number of loan cycles, existingclients who might have stayed withthe programme for any length of time– with their creditworthinessestablished and by contributing moreto organizational revenue throughbigger loans – are potentially lessrisky and more profitable.

Additionally, while client exit cannotbe always perceived as perfectlycorrelated with client dissatisfaction,a high rate of desertion does not bodewell for the organization’s public imageas well as for the morale of field staff.For most of the above reasons, it isnot only crucial to measure client exit,but also to do it in a way so as not tooverestimate (or underestimate) thephenomenon.

Client Attrition in Microfinance: Experience and Practice


To a large extent, the growth andsustainability of microfinanceinstitutions can be viewed as ameasure of their ability to forge long-term relationship with clients. Highdrop-out rates, by making the usuallyexpensive venture of deliveringfinancial services to low-incomehouseholds even more expensive,can be particularly disconcerting asmarkets mature and become morecompetitive. Additionally, thephenomenon can impede anorganization’s social objectives, whichat most times are directly proportionalto the length of time a clientparticipates in microfinance services.

An organization accrues severalimportant benefits from keeping oldclients for the longest possible lengthof time: a larger number of new clientsmean additional costs in terms ofmarketing and more staff time inorientation sessions and clientscreening. A higher level of clientretention further bolsters staffproductivity as older clients in mostcases would have developed strongerties with organizational staff. Althoughorganizational staff is expected to

the mere delivery of microcredit:clients might continue to take activepart in group meetings, contribute togroup savings and be ‘active’ in thatsense without contracting a follow-up loan. One group of expert opinionhas accordingly defined drop-out asa ‘client who has had no transactionwith the MFI for the last six months,’where the term transaction is usedto cover all client engagements withthe service provider includingcontributing savings, attendingmeetings, as well as contracting andrepaying loans (M-CRIL 2005).1

However, by virtue of the very factthat microcredit forms the core activityand the very raison d être for mostMFOs, such broad based attemptsat defining client status incorporatingthe full range of microfinance servicesand institutional ties is less favoredby most organizations. Nonetheless,it is important to note that availablemeasures of client exit can beaccommodated to include the abovebroad-based definition of whatconstitutes an active client.

a) Definitional Issues

Broadly speaking, three major issuespervade available literature on howbest to define client exit: Who is anactive client? When does he/shecease to be active? And whatstandard time-frame can best capturean adequate, policy relevant andprecise incidence of client exit?Irrespective of the differences thatexist on each of the above, addressingthese fundamental questions is thecrucial first step in the developmentof a reasonably accurate measure ofclient exit/retention. Accordingly, wewill take up these questions in thissection as a lead up to discussing thepros and cons of the various existingmeasures for measuring thephenomenon.

i. Microcredit vs. microfinanceIn general terms, client exit in micro-finance organizations (MFOs) can besimply defined as the failure tocontract an additional/follow-up loan.While this might furnish a reasonablebasic premise, the proposition hasbeen contested on the grounds thatit restricts microfinance services to

4 Pakistan Poverty Alleviation Fund


1 The definition comes from Micro-Credit Ratings International Limited (M-CRIL), an international rural financingconsultancy providing specialized services to the microfinance industry including credit rating of MFIs, monitoring,evaluation and sectoral research services.

(Imp-Act 2004).3 However, stretchingthe resting period that far will delayeffective policy response in dealingwith client desertion as decisionmakers would be waiting for a verylong period to get empirical evidenceshowing the full extent of desertion.Waterfield (2006) points to a usefulapproach to get around this problem:

“An alternative approach is to develop“aging categories” to measure the riskof losing clients, just as MFIs do forthe risk of loan defaults. By developingdifferent client aging categories, wecan generate different retention ratesthat can tell us things about clientbehavior in a more responsive fashion.We may therefore choose to definetwo or more retention ratios, each witha different cutoff date for resting. Suchratios could, for example, track:Regular clients (borrow again within2 months); Resting clients (2–6months); Likely deserters (7–12months); Confirmed deserters (morethan 12 months without a loan).”

The MFI cited above had in factadopted such an approach (Pawlakand Matul 2004). After allowing for a



ii. Time lapse between successiveloansBarring forced exit imposed by themicrofinance organization, twopossible scenarios can emergefollowing the successful completionof a loan cycle: The client mightimmediately contract a follow-up loanor he/she might wait for some timebefore doing so.2 While defining clientexit, how far should this ‘resting’ periodbetween successive loan cycles bestretched before categorizing amember as an ex-member?

The literature does not provide anyone uniform answer as organizationshave followed different formulationseither with or without preliminaryresearch on the subject. For instance,one MFI, after reviewing its data onclient exit and return patterns, foundout that 75% of all exiting clients inthe past, if ever they returned, didso within no longer than 250 dayswhile 50% came back after no longerthan 150 days. Following internaldiscussions with staff, a drop-outclient was defined as ‘a client thatdidn’t come back to the programmeat all or came back after one year’

2 This might be caused by several factors including the seasonality effect of many businesses (including thosedependent upon a particular agricultural crop) experienced by clients. Such clients might either need credit at aparticular time of the year or feel that they can repay loans only under certain conditions specific to a particular season.3 Improving the Impact of Microfinance on Poverty: Action Research Programme (Imp-Act) is a global actionresearch programme designed to improve the quality of microfinance services and their impact on poverty. It is acollaboration between MFIs, academics, NGOs and its secretariat is located in the Institute of Development Studies,University of Sussex. The MFI mentioned above is Partner, which operates in Bosnia and Herzegovina.

resting period, which can be a crucialingredient in measuring retention/desertion rate more effectively. Incertain cases, defining resting periodcan be tied to an organization’s lendingmethodology. For instance, in asituation where a group/village lendingmethodology specifically ties loanrenewal to a group’s loan cycle, theresting period can be simply set tomatch the loan term. In case the clientdoes not renew his/her loan at the endof the group loan cycle in which he/shedid not participate, there might besufficient grounds to consider him/heras a drop-out after that.

The above discussion reflects thatwhereas it is important to account fora specific time-frame after which a‘resting’ client can be conceived asan exit client, interpretations on theactual length of this period differconsiderably, depending on what aparticular organization considers asreasonable keeping in view itsexperience and history.

iii. Reference period for calculatingthe exitThe preceding discussion has

one year resting period, resters wereclassified on the basis of their returnbehavior as, a) hard users – thosewho returned within 60 days (andpresumably did not experienceseasonality in their businesses) and,b) resters – those who returned after60 days (and had a greater likelihoodof experiencing seasonality).

Concluding that the likelihood of hardusers to return was much higher, sucha categorization allowed MFI to focusmore intensively on the second category.

Prizma, another MFI located in Bosnia,adopted an identical approach.Following a similar survey of historicalclient exit data, a drop-out was defined‘as a person who has repaid any typeof loan but has not taken any newloan during the next 90 days.’ Thisconclusion was merited by data, whichshowed that the drop-out rate wasonly slightly less at over 500 days thanwhat it was at 90 days: most of thosewho wanted to return did so within 90days (Matul and Vejzovic 2004).4

These efforts gave implementingorganizations a time-frame to define



4 Taking the view that every drop-out can be retained, Prizma denoted drop-outs as ‘sleepers’ who were profiledinto three categories: ‘Voluntary – satisfied’ (leaving for reasons external to Prizma), ‘Voluntary – dissatisfied’(leaving for reasons internal to Prizma), and ‘Forced Out’ (for bad character or bad services)

monthly comparisons. One way ofdoing this is to compare dropout/retention rate in a particular month tothe same month in the previous year.Further, better trends can be identifiedthrough calculating annual rates forsubsequent months or other periodsof time, or using rolling averages.5

b) Measuring Client Exit

The Microfinance industry is yet toreach consensus on a standardformula for measuring client desertion/retention. To a large extent, the delayhas been induced by the perceivedfutility of measuring client exit asmicrofinance organizations rarelyexperienced demand constraints inthe initial decades. Even afterparticular MFOs took up the issuemore seriously following increasedlevels of saturation and competitionin specific markets, differences overa standardized approach to client exithave continued. Additionally, somenoteworthy contributions to theliterature have admittedly taken aless than ideal approach to measuringclient exit in order to compensate for



focused on defining exit status as thefailure to contract a follow-up/repeatloan within ‘x’ number of days (withthe ‘x’ denoting the time allowed forresting period). Equally important isthe criteria for arriving at a standardreference period over which clientdrop-out rate is measured.

Again, no particular standard isfollowed with microfinanceorganizations adopting monthly,quarterly, six monthly and annualreference periods for measuring clientexits. Although an annual rate willusually reflect better trends, it isimportant to note that shorter referenceperiods might be reasonably usefulfor the quick identification anddiagnosis of problems in the shortterm, particularly with reference toaddressing specific issues in somebranches or operational clusters.

Given seasonal variations, monthlyexit rates might vary sharply incomparison to rates calculated overa longer reference period. For thisreason, it has been at timesconsidered more fruitful to incorporatea measure of trend analysis to simple

5 For a useful discussion on incorporating trend analysis in client desertion/retention rates, see Pawlak andJahic (2004).

gaping holes in the data availablewith most MFOs.

In this section, we set out to delineatethe various approaches in measuringclient desertion/retention, whilefocusing on their respective limitationsand advantages. As the practicalapplication of any such approach isfundamentally dependent on theavailability of necessary data, it isimportant to cite the specific datarequirements for each approach.

i. The ACCION FormulaACCION6 conducted the first widelyrecognized effort to measuredesertion rate in response to the highincidence of client exit experiencedby one of its affiliates in the mid-1990s.The following formula was used tomeasure Desertion Rate (DR):7



Where,

x1= Active Clients at the beginning of the period;

x2= Active Clients at the end of the period;

NC= New Clients joining during the period.

DR= Desertion Rate

x1 + NC – x2

x1

DR –

6 ACCION is a leading microfinance organization with a network of lending partners in Latin America, Africa,Asia and the United States

7 The ACCION formula can be reformulated in terms of a Retention Rate (RR):

x1 ( x1 + NC – x2

)x1

RR = 1 – DR = =x2

– NCx1

The ACCION formula is a simple

measure that calculates DR as thenumber of clients dropping out in theperiod, expressed as a ratio of thetotal number of active clients at thebeginning of the period. Immediately,several shortcomings can bedeciphered:

i) the formula is inadequate to

measure DR for neworganizations in their first

year/period with x1 = 0 ;ii) even for relatively older

organizations, as we wouldshortly work out, the formulabecomes illogical when a

large number of NC enter anddesert during the samemeasurement period. As theyare added to the numeratorbut not to the denominator,the formula would show an

inaccurately large DR.8

iii) the formula does not accountfor resters, although thefollowing adjustment can bemade to account for them(Pawlak and Matul 2004):

since RR tries to determine thenumber of clients an MFO retains ofthose who are clients during the year,it makes more sense to put NC in thedenominator (Waterfield 2006).

Correspondingly, as DR calculatesthe number of clients that havedropped out of all those who coulddropout during the period, it is equallyimportant to account for NC in thedenominator. Schreiner (1997)accordingly made the necessaryadjustment (See Fig-A):

The formula calculates DR as thenumber of clients dropped out duringthe period as a ratio of all those whocould drop out (or, who were activeduring the same period). DR calculatedas thus is more accurate as itincorporates, unlike the ACCIONformula, NC in both the numeratorand the denominator. This reformulationof the Schreiner formula has in fact



Where,

R= Clients who contracted a follow-up loanduring the reference period after resting for

x number of days

x1 + NC + R – x2

x1

DR =

iv) Lastly, the formula also doesnot account for dropouts thatcould occur amongst NewClients

While many organizations continue

to use the above formula – either inits original form or with one of theseveral adaptations mentioned in thisnote – ACCION itself gave up theformula for the one proposed by MarkSchreiner as part of his Ph.Ddissertation (Schreiner 1997).

ii. The Schreiner FormulaIt has been accurately noted that

8 To account for this deficiency, an adaptation has been proposed (M-Cril 2005) to the original ACCIONformula by including a better, much larger, denominator:

, where the denominator represents anaverage of the number of clients at each interval during the reference period. This resolves the problem of alarge number of deserters, including from NC, through a more representative denominator. Yet, the averagedoes not cover all clients during the period. In any case, a better denominator is available in the SchreinerFormula, which works approximately similar data requirements.

x1 + NC – x2

Avg ( x1 ... ... x2 )DR =

Fig-A

Number of Dropouts

Number who could DropoutDR =

NC – (x1 – x2 )

x1 | NC–

x1 + NC – x2

x1 | NC=



been adopted by M-CRIL, noting that‘calculating drop-outs as a proportionof all the clients that MFI has comeacross during the period,’ accountsfor drop-outs that can occur from NCas well (M-Cril 2005).

The same formula can be reformu-lated for calculating RR (See Fig-B):

The above formula calculates thenumber of clients retained (numerator)out of the total number of clients thatcould have been retained(denominator). As such, whenadjusting the formula for resters, thenumerator must include only thoseresting clients who are resting at theend of the period.

Conversely, the denominator includesonly those resting clients who wereresting at the beginning of the periodand may have changed their statusduring the period.

Fig-B

RR =x1 + NC – [{( x }]1 + NC – x2 )

x1 + NC=

x2

x1 + NC=

x1 + NC – x2

x1 + NC1 –

The formula can be adjusted forresting clients (Waterfield 2006):Waterfield (2006) makes aninteresting comparison between theACCION and Schreiner formulas,which brings out the latter’scomparative accuracy more clearly,and is worth reproducing here. Thereformulated ACCION formula for RR(see discussion above) is denoted asthe “Old” formula by Waterfield, whileSchreiner’s is referred to as the “new”formula:

Where,

x1= Active Clients (Loans) at thebeginning of the period

x2= Active Clients (Loans) at the end ofthe period

R1= Resting Clients at the beginning ofthe period

R2= Resting Clients at the end of theperiod

x days = Resting period

RR =x2 + R2

(x days)

x1 + R1 + NC(x days)



Retention Rate(RR)

“Old” FormulaEnd Clients – New Clients

Beginning Clients

“New” FormulaEnd Clients

Beginning Clients + New Clients

Case 1Begin = 4New = 2End = 4

44 + 2

46

= = 66%4 – 2

424

= 50%


20001000 + 2000

20003000

= = 66%2000 – 2000

10000

1000= 0%


1900 – 2000100

–100100

= –100%1900

100 + 200019002100

= = 90%

=

=

=

In Case 1, where 2 clients drop outfrom the 6 that were active during thereference period, the Schreiner

formula returns a more accurate RR(66%). While working withcomparatively more extreme data inCase 2, where 1000 clients out of the3000 that were active during thereference period have apparentlydropped out, the Schreiner formularemains consistent by returning the

expected 66% RR. Interestingly, the

ACCION formula fails by returning

0% RR. This, however, is not all. InCase 3, which can be indicative of anew and growing organization with alarge number of new clients duringthe reference period, the ACCION

formula returns a negative 100% RR,indicating a complete collapse. TheSchreiner formula, on the other hand,returns the more plausible 90% rateof retention. The reason for ACCION’sfailing is simple: a large number of

measure of client retention thatfocused on decision points:9

new clients, if subtracted from thenumerator and not covered by thedenominator, will seriously distort theratio. The Schreiner formula doesnothave this weakness as new clientsare added both to the numerator andthe denominator.

In light of the above discussion, theSchreiner formula could be easilypassed off as a better bet. However,both share a ‘logical inconsistency’:the RR, in both formulas, is overstatedby including new clients that have notyet had the chance to either desert orbe retained (Waterfield 2006). AlthoughSchreiner was the first to concede theargument, the formula was justifiedon the grounds that it makes less datademands on MFOs and works withinformation that is easily available(Rosenberg 2001). The CGAP-Waterfieldformula was designed to specificallyaddress the above stated drawback.

iii. The CGAP/Waterfield Formula

In 1997, while developing the“Microfin” business planning tool forCGAP, Waterfield formulated a



9 This is also sometimes referred to as the Microfin formula10 If the MIS lacks provision for directly stating LP and FL, then the formula can be restated as:

,where L = Number of Loans disbursed during the period(Rosenberg 2001). The Delinquency Management Framework of MFOs account for delinquents through measuressuch as Amount overdue and Portfolio at Risk (PAR). Different MFOs might apply different time-periods for whichsuch clients are shown as active, before their loans are written off and they can be conceived as dropouts.

L – NC

x1 + L + x2RR =

Where 10,

FL = number of Follow-up Loans during the period

LP = number of Loans paid off during the period

RR =FL

LP

DR = 1FL

LP

to calculate the drop-out rate.

Put simply, if 80 clients receive newloans out of a 100 clients who havesuccessfully paid off their loans duringthe reference period, the retentionrate would be 80 percent. Theformula, however, has some seriousshortcomings:

i) A specific focus on retentionunderstandably leads tocomplications that can be betterperceived in terms of desertion.For instance, while trying to divest

This could be reformulated as



his formula from the problem ofdecision points vis-à-vis newclients, Waterfield took new clientsout of the equation altogether.This logically led to the associatedproblem of accounting for drop-outs amongst new clients. Withthe Waterfield/CGAP formula, afailure to do so results inoverstating RR.

ii) Similarly, the denominator LP didnot account for delinquents whodid not pay their last loan withpotential for further inflating RR.11

Recognizing the weakness,Rosenberg (2001) proposed thefollowing adjustment:

RR =FL

LP + WO

have received a follow-up loanduring the period (hence reflectedin the numerator). Conversely,RR is deflated if clients repay aloan during the reference period(hence reflected in thedenominator) and decide to restfor the remainder of the currentreference period (hence notreflected in the numerator eventhough he/she is consideredretained). The effect of thisphenomenon is more pronouncedwhen the reference period is smallrelative to the resting periodallowed by particularorganizations, at times yieldingover 100% RR or DR dependingon weather resting is occurring atthe beginning or the end of theperiod. In such instances, a simplecorrection can be made:

11 The Delinquency Management Framework of MFOs account for delinquents through measures such asAmount overdue and Portfolio at Risk (PAR). Different MFOs might apply different time-periods for which suchclients are shown as active, before their loans are written off and they can be conceived as dropouts.

R =LP + R1

FL + R2where WO accounts for loansWritten Off during the period.

iii) Furthermore, RR (FL/LP) isexaggerated if the denominatordoes not cover for resting clientswho had paid their last loan in aprevious reference period and

, where R2 and R1 denote restingclients at the end and beginningof the reference periodrespectively. In cases where a

remain after each cycle whichmeans an annualized RR of 41% (.8*.8*.8*.8 = .41).12

Since loan terms vary most acrossdifferent loan products, one wayto decrease the above complexityis to calculate the ratio for differentproducts along with the overallRR. Viewed together, such ascheme might be a better andmore meaningful rendition of theincidence of retention within anorganization.

iv. The New Waterfield Formula

The Shreiner formula includesnew clients who have yet to reacha decision point. TheWaterfield/CGAP formula soughtto address the weakness byadopting a new approach basedon (See Fig-C), rather than onthe number of clients (See Fig-D).In doing so, it became open tocomplications resulting fromvarying loan terms.

The new Waterfield formula went back



12 If 100 clients receive the initial loan with a 12 months loan term, then an 80% RR will mean that 80 clientswere retained at the end of the year (assuming that there were no defaulters and resters, or that the formulahas been adjusted to account for the same). With a three month loan term (4 loan cycles per year), the 80%(un-annualized) RR will mean that only 41 of the original 100 were retained at the end of the year. Similarly, an8 months loan term will require an adjustment by a factor of 12/8, which would mean an average loan cycle of1.5 per year.

sufficiently large reference periodis taken relative to the restingperiod, the unadjusted formulawill yield reasonably accurateratios, provided the otherdeficiencies have beenaddressed.

iv) The formula works well when theloan term offered by anorganization is equal to thereference period and consistentacross various loan products.This is because the formula

calculates RR per loan cycle. Anyincrease in loan cycles relativeto the reference period cancomplicate the formula. Forinstance, for an annual referenceperiod where the MFO has onoffer either 2, 3 or 4 loan cycles

per year, the calculated RR (say80% or .8) has to be accordinglyadjusted by the factor of 2, 3 or4 respectively [.82, .83, .84] toget the correct annualized rate.Mathematically, with 4 loan cycles

per year, the calculated RR of80% through this formula wouldmean that 80 % clients would

period as well as clients repaying thesame loan during the whole referenceperiod (possible only if the loan termis longer than the reference period).Active Clients that were repaying thesame loan but have reached at leastone decision point during the period(active clients with a second, third,fourth, etc. ongoing loan in the year)were deemed to have reached therequired decision point for the periodand have decided to remain in theprogram.

To adjust the formula for resting,Waterfield suggested the followingcorrection (See Fig-F):



to the original appraoch applied inSchreiner by simply adjusting thelatter for those who had yet to reacha decision point. Essentially the newformula calculated retention as thenumber of clients who reached adecision point during the period andremained with the program, as aproportion of all those who had adecision point during the period.

The Schreiner formula was thusadjusted (Waterfield 2006) as(See Fig-E).

It is important to note that clients whohave not yet reached a decision pointinclude only those with an ongoing“new” (first time) loan during the

Fig-C

Number of Follow on Loans during the period

Number of Loans Paid off during the period( )

Fig-D

Number of Clients Retained at the end of the period

Number of Clients Active during the period( )



Fig-E

End Clients – Clients with the Same Loan

Begin Clients + New Clients – Clients with the Same LoanRR =

x2 – Clients with the Same Loan

x2 + NC – Clients with the Same Loan=

Fig-F

– Clients with the Same Loan

– Clients with the Same LoanRR =

Where,

x1= Active Clients (Loans) at the beginning of the period

x1= Active Clients (Loans) at the end of the period

R1= Resting Clients at the beginning of the period

R2= Resting Clients at the end of the period

x days = Resting period

Same Loan = a) NC with an ongoing loan at the end of the periodb) Cleints repaying the same loan for the whole of the period

x2 + R(2)(x days)

x1 + R(1) + NC(x days)

The new Waterfield formula istheoretically more accurate than otherformulas cited above. However, asWaterfield (2006) cautions, the degreeof error will increase if:

the loan term is long relative tothe measuring period (i.e., moreclients are included who have not

yet reached a decision point)the institution is growingsignificantly and there are a largenumber of first-time clients whohave not yet reached a decisionpoint

In the final analysis, the most accurateformula any organization can adopt



at a particular point in time isdependent on the information it canreadily generate from its MIS.

c) Acceptable AttritionRates

Intuitively, the lesser the incidence ofclient exits the better. However,particular microfinance organizationsare well placed to make thatjudgment, depending upon thespecific spatial and temporal contextin which they operate. For instance,an organization might be content witha comparatively less than ideal drop-out rate if the same could berationalized in terms of forces that itcannot control. These might include,a) a highly stressful environ-mentcharacterized by systemic shocks intimes of economic recession, civilunrest, and severe weatherconditions; and/or b) an unusuallyhigh incidence of idiosyncratic shocksexperienced by clients owing to acombination of factors, e.g., illness,death, business failure, economicdistress of family members,ceremonies, etc.13

Similarly, microfinance organizationsthat are at a comparatively lessmature stage in their developmentmight justify varying degrees of clientloss to stronger competition in theshape of more sustainableorganizations that are in a betterposition to attract more qualified staffand have greater funds, experience,and product diversity/ flexibility.

Additionally, acceptance of a certainlevel of client desertion can be tiedto organizational objectives: a higherdrop-out rate, for instance, can bepalatable to an organization if it helpssustain zero portfolio at risk (Imp-Act2004).

13 For a summary of the push and pull factors cited in the literature on microfinance client exit,see Pagura (2003)

Chapter 2:

Reasons for Client Desertion

the sustainability and profitability ofmicrofinance organizations.Information on the causes of clientexit can further assist MFIs inunderstanding client preferences andmarket trends, leading towards thedevelopment of more appropriate,user-friendly and feasible MFIproducts.

Significant theoretical and empiricalresearch has been conducted on thecauses of client exit from themicrofinance sector. The principalobjective of this chapter is to reviewwork that has already been conductedon microfinance attrition globally andin Pakistan particularly. In view ofthe sector’s rapid growth in Pakistan,the identification and assessment ofreasons for client exit has becomecritical to ensure that greatercompetition does not constrain orartificially limit the natural progressionof the sector.

Attrition need not only be viewed asa negative. The sector has in factfacilitated a large number of clientsto graduate beyond microfinance intothe formal sector. Such client exit can



A full understanding of factors leadingto client dissatisfaction is a vitalmilestone on the road to achievingfinancial sustainability for all MFIs.Why an MFI’s target clientele, or apart thereof, chooses to stay away isan important – yet a logistically and/orfinancially problematic – researchquestion to answer. It is thus morefeasible, and arguably more effective,to alternatively study clients whochoose to exit microfinance servicesas a way to developing a betterunderstanding of the sector’slimitations.

It has been argued that repeatborrowing is critical to the long-termviability of MFIs as it lowers risks,reduces administrative costs andenhances organizational productivity.Contrarily, a high dropout rate mightforce an MFI to continually expandits lending base in order to maintaina viable portfolio size. In mostinstances, the recruiting and trainingcosts incurred before making the first(small) loan to new clients preventsprofitable lending before the third orfourth loan cycle (CHIP, 2005). Forthis reason, client retention is vital to



important force behind dropoutsin the sector. Churchill (2000)identified competition in the LatinAmerican setting as a significantexplanation of client exit fromparticular MFIs. In a similar vein,Shreiner (2001) found thatcompetition from Chileanconsumer finance companies leddropouts to double for a largeBolivian MFO offering individualloans in Bolivia. Similarly,Maximambali (1999) found thatin areas with significant MFIconcentration, clients tended toleave one MFI for another.Montgomery et al (1996) alsoidentified client perceptions ofother MFIs offering betterservices as an important reasonfor drop out. Multiple studiesconducted in Bangladesh(Hassan and Shahid, 1995; Khanand Chowdury, 1995; Hulme andMosley, 1997; Evans et al., 1999),have found that a significantnumber of former clients typicallyshift to another MFO in the areabecause of better products andservices. The CHIP study (2005)in Pakistan similarly found that

be classified more appropriately asa successful outcome than a failure.

a) Causes of Client Exit

i. Market-Based Factors:Market-driven attrition resultsfrom either industry competitionor client maturity and are oftencited as a sign of healthycompetition or successful servicedelivery. Such factors are usuallyconsidered a positive forceinsofar as they reflect the client’sability to choose between differentalternatives, i.e., by eitherallowing a movement towardsMFIs with more competitive termsand better services or facilitatingclients to graduate from themicrofinance sector altogether toa higher quantum of financeoffered by the formal financialsector (Pagura, 2003). This kindof exit from the microfinancesector is regarded as a positiveforce.

Sufficient research has beenconducted globally identifyingmarket-driven factors as an

One of the most frequentlyidentified reasons for voluntaryclient dropout has been rigidityof the MFI’s products and policies(Painter and MkNelly, 1999;Wright, 2000; Murray, 200114).Strict rules concerning savingsrequirements and restrictions onwithdrawing savings even intimes of need have also beenidentified by former clients (Kuwikand Mashaba, 2000; Churchilland Halpern, 2001; Stark andNyiumburinga, 200215). Lowcredit ceilings and rules againsttaking multiple loans have alsobeen identified as important factorsleading to clients dropping out(Wright et al., 1999; Mosedale,2001; Garuba, 2004 16).

Another factor cited was thatintervals between installmentswere too short (Stark andNyiumburinga, 2002; Garuba,2004).

Methods of Recruitment andLoan Disbursement: Stark andNyiumburinga (2002) found thatrecruitment methods are a major



15 % of dropouts were causedby market-driven factorsincluding a lack of demand forservices of the MFO, borrowersfinding other providers with betterterms, and client needsgraduating beyond what theNGO can provide.

ii. MFI-Related FactorsForced Expulsions: In manyinstances, the MFI decides toexpel clients if the latter defaulton their repayments and/or if thecredit officer feels that a particularclient might not be able to repayin the future.

A study of Kashf clients inPakistan found that about 60%of the attrition was due to clientexpulsion for poor performance(Mosedale, 2001).

MFI Products/Policies: Themajority of dropout cases (61%)interviewed for the CHIP studyexited because of reasonsrelated to the lendingorganization’s product design ororganizational policies.

14 This Also see ASA (1996); Churchill and Halpern (2001); Stark and Nyiumburinga (2002)15 Also see Montgomery et al (1996); Painter and MkNelly (1999); Kashangaki et al. (1999); Maximambali(1999); Wright et al., (1999); Wright (2000)16 Also see ASA (1996); Murray (2001); Kuwik and Mashaba (2000); Churchill and Halpern (2001)



Some MFIs require all membersof a group to continuously takeloans of increasing size (Matin,1998), and often a client can onlytake a loan on the prescribed dayand time of disbursement (CHIP,2005). This assumes constantlyhomogeneous needs of allmembers and is likely to forcerelatively poorer members – whocannot afford the installments –out of the group. MFIs appearunwilling to recognize, especiallyin rural areas, that there areseasons when loans are notrequired. The clients are thenleft with no option but to borrowand try against the odds toservice the loan, or to leave theMFI altogether (Wilson, 2001).Zeller et al (2000) in fact, foundthat the average size of borrowinggroups fell over time becausedropouts exceeded newmembers. Specifically the studyshowed the average size of ASAgroups falling from 25 to 18individuals over a five-year period,and that of BRAC groups fallingfrom 56 to 37.

contributor to client exit at PSAand SEF. PSA recruits clients byretaining agents who areresponsible for finding clients intheir immediate neighborhoods.On the other hand, SEF requiresthat members of poor householdsinterested in taking part in theprogram form groups and createa full center. Clients from bothMFIs identified tactics used by theagents as problematic, includinghumiliating clients by visiting theirhomes at odd hours. Agents werealso found to encourage, andsometimes force, clients to takelarger loans than they needed ortheir ability to repay. Clientsinterviewed as part of the CHIP(2005) study in Pakistan alsoreported MFI staff either forcingor sweet talking people into takingloans. In Bangladesh as well, akey factor for client failure to repaya loan (which would automaticallyexclude them from being able totake additional loans) wasidentified as the insistence of fieldstaff that clients take loans (Wright,2001; and PromPT, 1996).

interviewed for the Stark andNyiumburinga (2002) studyreported a large number ofrequirements for groupmembership: group memberswere required to be within 15years of age of each other, notrelated to each other, be friends,live in close proximity, and fallwithin the MFI’s ‘poverty criteria’.Such condionalities significantlylimit not only the number ofindividuals eligible to becomegroup members, but are morelikely to result in group membershaving to guarantee a much lessdesirable set of co-members.

Stark and Nyiumburinga (2002)found cases where defaulterswould taunt their group membersthat they cannot go to the policeor local council and presscharges against them. Also, theunwillingness to pay for defaultinggroup members (Hassan andShahid, 1995; and Mustafa et al.,1996; Stark and Nyiumburinga,2002) has also been at timesidentified as a reason for clientexit.



Transaction Costs: Meetingsthat were too frequent and/or toolong were also identified as amajor reason for dropouts (Murray,2001; Khan and Chowdury, 1995;Kashangaki et al., 1999;Maximambali, 1999; Wright et al.,1999; Kuwik and Mashaba, 2000).At times, the distance betweenclient residence and thedesignated place for groupmeetings have also causedapprehensions, especially in thecase of women from conservativerural backgrounds. Cost ofborrowing being too high (Murray,2001; Khan and Chowdury, 1995)has been cited as a cause ofdropping out in many countries,especially when there are cheaperalternatives.

Problems with GroupFormation: In some instances,MFI policies on groupcomposition have been found tonegatively impact the quality ofborrowers. For example, clientsfrom two MFIs - Provident SouthAfrica (PSA) and SmallEnterprise Foundation (SEF) -



A CGAP study in 2000 also foundthat dropouts increase when thereis a downturn in the nationaleconomy. The vulnerability ofMFI clients to external shocksunderscores the fact that MFIsare serving a precarious market. Such vulnerability, which canadversely affect both retentionand repayment of clients,highlights how important it is forMFIs to address other needs oftheir clients besides credit. MFIscan either develop in-housecapacity or otherwise formstrategic alliances to providevoluntary savings, micro-insurance and businessdevelopment services. Offeringaccess to these valuablewraparound products could bean important dimension of anyclient retention strategy (Murray,2001).

iv. Life Cycle EventsPersonal and family illness anddeath also lead to client exit(Maximambali, 1999; Simanowitz,1999; Kuwik and Mashaba,2000)17

Staff Behavior: In the LAPOstudy (Garuba, 2004), staffattitude was found to be animportant factor in the problemof client exits. In particular, mis-targeting of clients, inadequatepre-loan training for borrowersand disrespect for customersranked high among the issuesthat former clients reported. Thefailure of some staff to strictlyfollow procedure for selection ledto the inclusion of clients, whoreally do not value the amount ofloan offered by LAPO because itis inadequate for their businesses,and who are therefore not thetarget clients of LAPO.

iii. Systemic IssuesResearchers have also found thatclient exit is provoked by systemicshocks. Some of the studiesdocument natural disasters, e.g.,drought or excessive rains, theclosing of key industries, andgeneral macroeconomicdownturns as factors that provokeclient exit (Kashangaki et al.,1999; Maximambali, 1999; Wrightet al., 1999; Simanowitz, 1999).

17 Also see Hassan and Shahid, 1995; Khan and Chowdury, 1995; Painter and MkNelly, 1999; Churchilland Halpern, 2001; Wright et al., 1999; ; Stark and Nyiumburinga, 2002.

factor for dropout and the inabilityto repay. At Provident, clientsadmit that since they all do thesame types of business, theycompete against each other andno one makes a significant profit(Stark and Nyiumburinga, 2002).

vi. Peer Based ExpulsionGroup based expulsions aremostly common under a jointliability system where groupmembers are made to pay theinstallment in case any memberfails to do so. In someorganizations, loan officers don’taccept repayment installments ifany one member is defaulting,putting the whole group injeopardy. In many cases, suchmembers are usually expelledby the group itself.

In the LAPO research study, aquarter of the ex-clients claimedthat their MFOs and groups hadexpelled them. In Kibaha,Tanzania, it was similarly foundthat expulsion of clients from agroup was often initiated bycenter members. In fact, the



v. Enterprise ViabilityAlthough microfinance loans aregenerally extended to clients forthe purpose of investing inproductive activities, theinvestments are not alwaysprofitable. Business problemsincluding cash flow issues,seasonality factors, and lack ofbusiness skills can therefore alsocause clients to exit the sector(Maximambali, 1999; Wright etal., 1999; Kuwik and Mashaba,2000; Simanowitz, 1999;Churchill and Halpern, 2001).The lack of business skills hasbeen particularly highlighted asa reason for failure of businessesby credit officers and borrowersalike. Many take loans andexpand their business withoutadequate consideration of thecapacity of their businesses togenerate the cash to repay theloan. The lack of, or poor, recordkeeping is a major weakness inmost informal sector businesses(Stark and Nyiumburinga, 2002).

Business competition has alsobeen highlighted as a major



at the time (Churchill, 2000).Many researchers have identified,however, that often clients thatare resting immediately followinga loan repayment do intend tosecure a loan in the future, butthat monitoring systems at mostMFIs are not able to distinguishsuch clients from non-voluntarydropouts.

More importantly, there is oftenno single clear reason of clientexit. For instance, it is possiblethat a client drops out becausethe loan product is not feasible,in addition to a lean patch inbusiness or illness in the familymaking repayments difficult.

c) SocioeconomicProfile of Exit ClientsGaruba, 2004 found thatindividuals who fall in the categoryof Least Poor, Less Poor andAverage Poor groups constitutednearly 82% of MFI clients, whilethe Poor and Very Poor onlymade up 18% of clients. A CGAPstudy similarly found that themajority of clients were relatively

research team’s impression wasthat group members were quiteruthless when it came to theexpulsion of defaulters and otherdifficult members (Matin, 1998).

b) Classification of ExitClientsConceptually it is possible todistinguish between voluntary,non - voluntary and forced outdrop outs. In practice, however,it is often difficult to categorizedropouts into neat categories.For instance, while it is easier toseparate forced out client exits(those expelled by the MFI) fromthe other two categories citedabove, a clear distinction betweenvoluntary (those who don’t needa loan at the time) and nonvoluntary (those who need theloan but do not contract itbecause they cannot repay dueto some reason) drop outs mightnot always be possible. Manyclients, for example, choose torest from the loan repayment aftera cycle is complete due toseasonality in businesses and/oras a result of no expressed need

downturns faced by the poor, ledto dropout.

On the other hand, the relativelybetter off clients of MFIs show atendency to drop out, based moreon factors such as the desire forlarger loans as even themaximum loans offered by MFIsare often too small for theirgrowing businesses.

Borrower dissatisfaction due todelays in entering the next loancycle due to other groupmembers’ reluctance or being inarrears, and frustration with thelength of time spent in groupmeetings and in trying to recruitnew members to replacedropouts were also importantcauses of dropout.

It has been argued that thepoorest cannot join most MFIsto begin with. This is based onseveral factors including the factthat the poorest do not havebusinesses – a pre-requisite formost MFIs, instead they oftenengage in casual labor or work



less poor (though not necessarilynon-poor) – those that haverelatively stable income, sufficientlivelihood diversification thatallows them to make regularinstallment payments, and whoare not as vulnerable to smallexternal shocks.

Although clients that drop outhave been found to belong to awide range of differentsocioeconomic backgrounds, thereasons why clients decide todrop out vary greatly betweenthese classes. CGAP, forinstance, found that poorer clientstend to drop out when theaverage size of loans within thejoint liability group rises to highlevels and the members have totake the risk of guaranteeingmuch larger loans than theythemselves can take. In addition,it was found that poorer clientsare particularly vulnerable to theincreasing size of weeklyrepayment installments. Such“program-design-induced” risk,when coupled with the generalvulnerability to economic



Such a significant number ofdropouts during the initial periodindicate that a large proportionof clients were ‘testing’ the MFIin the case of voluntary dropouts,and those being weeded out bythe MFI in the case of forced outdropouts (Garuba, 2004).

CGAP study results indicated thatdropout rates tended to be higheraround religious festivities,periods of harvest and the timefor payment of school fees –when clients would likely needaccess to savings or would nothave the capacity to make regularloan repayments. The period rightafter a major change in the policiesof the MFI was additionally citedby staff as resulting in a largenumber of client exits from theirrespective programs.

e) Limitations ofLiteratureBecause no uniform study designhas been used to examine clientexit, it is difficult to makemeaningful comparisons acrossMFIs and it has been found that

at minimum wages. Many amongthe poorest that do ownbusinesses make very low return,and therefore cannot meet theonerous conditions such assavings requirement, group-guarantee and other transactioncosts associated with accessinga loan.

Evans et al. (1999) show thatformer clients were less educatedand had smaller households thancurrent clients, demonstrating thatoverall household vulnerabilitymay be playing a role. Thereasons from dropping out did notsignificantly vary based on genderof the client (Mutesasira et al,1998; Maximambali, 1999).

d) Timing of Client ExitA study on Nigeria (LAPO) foundthat a significant number of theex-clients exited after their first(47%) and second loan cycles(31.8%).

Cumulatively, these two groupsof clients constitute as high as78.8% of all the exited clients.

that offer loans to individuals.Although it is logical that this bethe case since it is the largestinstitutions receiving the mostresearch attention andBangladesh and Bolivia (bothpredominantly following thegroup lending model) dominatetheir peer groups in terms ofscale, it is important not toassume that all findings fromsuch research are relevant andapplicable to the microfinancesector at large.

f) ConclusionsWhen an MFI client drops out,the organization is negativelyimpacted in many ways. Thereare tangible financial losses fromclient attrition – the resourcesthat had been invested inrecruiting and training thatmember and also the loss offuture cash flows that would havecome from the increasing loansizes. The larger the potentialfuture cash flow from a givenclient, the larger the loss fromthe exit. Exit, especially in largenumbers, damages the growth



regional differences do exist(Pagura, 2003). Clients in Africaappear to be more susceptibleto exit due to idiosyncratic andsystemic shocks, indicating ahigh level of vulnerability at theindividual household level.Conversely, market-driven factorsplay a comparatively larger rolein explaining client attrition in LatinAmerica and Asia.

Pagura (2003) believes that theliterature on microfinance clientdropout rates is biased bygeographic location and in termsof the particular type of loanproduct being offered. Client exitwithin each region varies greatlyand is often concentrated in oneor two countries. She points outthat most of the exit studies havebeen carried out in Bangladeshand Bolivia given their relativelylong history of microfinanceactivity. In addition, she pointsout that almost all of the researchwork that has been conductedhas focused on group lendingand the village banking model –with very little focus on institutions



MFIs are found to not bemaintaining any formal record ofwhy borrowers are exiting theirprograms (CHIP, 2005).

Much of the microfinance sectoris currently being funded throughexternal concessionary finance,donor funding, and organizationsmore often than not havedemonstrated a tendency not tofollow self-identified objectivesbut those identified externally.CHIP (2005) points out that sincecurrent donor policies do notreflect a priority concern for clientretention. NGOs benefiting fromdonor funds are not required toreport their respective clientretention rates for theorganization as a whole or fordifferent loan cycles, as part ofthe required set of performanceindicators to benefit from futuredonor funding.

potential and trajectories of theMFI. They negatively impactmotivation in the centers wheredropouts occur and may also leadto a sense of failure among MFIstaff and a reduction in staffmorale could negatively impactproductivity and further slowgrowth.

Murray (2001) found that issuescausing dissatisfaction amongstcurrent customers were the sameas those that had causeddissatisfaction amongst formerclients. If this is the case, then itis even more important that MFIslearn from experience of otherorganizations. The literaturereviewed reflects a need formicrofinance institutions tosystematically collect informationon client exit with the intention ofunderstanding the limitations ofexisting products. Some studiesindicate that middle managementat many organizations are still ofthe opinion that dropouts are nota serious enough issue to requirespecial attention, reflectedparticularly in the fact that many

Microcredit Financing: The Generic Cycle

Ex-anteAssessment

Supervision &Monitoring

BorrowerIdentification

Ex-postEvaluation

Appraisal /Bankability

Financing /Disbursement

Processing &Documentation

Chapter 3:

Monitoring Client Exit

systems serve the decision makingprocess, it aids field workers inimproving their performance and drawmaximum benefits from incentivestructures, wherever the latter aredesigned to reward client satisfactionand retention.

Information from exit clients furtherassists in gauging client response toa variety of procedural, administrative,and policy changes within theorganization. Consequently,organizations with sustainedinvestments in exit monitoringsystems - and a willingness to actupon the information – have invariablyachieved declining rates of dropoutand delinquency with positive impacton portfolio growth and institutionalsustainability. While such investmentshave resulted in more client orientedpolicies, they have also played acritical role in revitalizing falteringdecision making structures throughthe infusion of a learning culturecritical to continued success incompetitive markets.

Essentially, the EMS design revolvesaround a single key concept –



Exit Monitoring Systems are vitallearning tools furnishing a continuousfeedback loop for decision makersat the helm of a microfinanceinstitution. A well designed ExitMonitoring System (EMS) helps inthe contextualization of successfulmicrofinance products, models andmethodologies imported from avariety of cultural, social andeconomic settings. Without a soundknowledge of local needs andpreferences that an EMS is designedto provide, an MFI is alwayssusceptible to an almost bureaucraticinsistence on straitjacket solutionsthat seldom work in the long run.

While facilitating institutions in thedevelopment of better business plansand targeting methodologies, regularinterviews with exit clients can furtherhelp in conceptualizing better loyaltybuilding and promotional strategies.Simultaneously, ongoing access toclient exit data provides managementteams with requisite knowledgecritical in adjusting product designsto market demand. Not only does thedevelopment of exit monitoring



a) Building EffectiveMonitoring Systems

Every EMS that tracks reasons of exitshould be carefully designed, testedand implemented. Any organizationinvesting in an EMS should carefullyconsider the following three questions:What information do you need fromdropouts? How that information willbe collected? How often do you needto collect such information?

The first question is perhaps the mostimportant and will usually take up themost time. The team responsible fordeveloping the EMS would need toconduct several preliminary sessionsat different tiers of management andwould need to talk extensively to fieldworkers and clients (both active andformer). Once a decision is made onwhat information will best facilitatebetter decision making within theorganization, further deliberations needto be carried out on ways in which thatinformation will be collected. Dependingon its preferences and needs, anorganization will either choosebetween a quantitative and qualitative

learning from clients who for somereason stop using the services of anMFI. It seeks to unearth the causesof disengagement with a keen eyeon rectifying what has gone wrongand deliberating upon what can bemade better. There is no one way ofcollecting information from exitclients: organizations have usedmethods ranging from samplesurveys to one off in-depth casestudies and focus groups. An EMS,while using some of the sametechniques that are now commonknowledge, follows a set processwith a clearly defined set of datacollection methods that are routinelyimplemented. Further, unambiguousprocedures are laid down to integratethe EMS in the decision makingstructure through precise channelsof communication for follow up actionin a host of interconnected spheresincluding product design, stafflearning, business planning,employee incentive structures,promotional and client satisfactionstrategies, customer services,targeting methodologies, etc.

carried out. It is equally critical tocarefully plan and not rush theimplementation of an EMS. As withany information system, latermodifications not conceptualized atthe system design phase are costlyand difficult to implement. The systemshould thus be tested and necessaryrevisions made before roll out.

Any system developed to monitor exitclients on a routine basis is usuallyconstructed out of a few importantbuilding blocks: a) a strong MIS, b)a well designed exit monitoring form,c) a mix of data collecting methods,and d) an institutionalized structureof data flow, follow up and effectiveaction through clearly defined comm-unication channels. Figure 1 showshow one MFI has brought the abovepieces together in its EMS design18.

For effective use and implementationof EMS, MFI first devised a definitionof what constitutes an exit client. TheEMS itself can be conceptualized asdivided into three components:

a) A routine component consistingof: i) A loan tracking system with



approach or will prefer a combinationof both.

Taking the latter option is also costeffective. For instance, a survey toolbased on a structured or a semistructured interview and administeredto a representative sample of exitclients on a periodic basis can becomplemented by Focus GroupDiscussions and detailed casestudies for a more in-depth diagnosis.It is also important to carefully planthe skill sets and human resourcerequired for collecting, inputting, andanalyzing data. Lastly, a decision onhow to collect data will guide thefrequency with which data iscollected. An organization will makea decision keeping in view the costsassociated with different datacollection methodologies and its owncapacity in terms of financial andhuman resources.

It is important for MFIs to embed exitclient management in their strategicvision and organizational objectivesto induce full commitment from staffand shake off any confusion withregard to why the exercise is being

18 Adopted from Pawlak and Jahic (2004). The organization alluded to is Partner in Bosnia-Herzegovina.



c) An action component that ensurestimely reaction and response toaddressing the problem and whichincludes: i) a research teamheaded by a marketing managerto develop proposals and takeinitiatives in response to theinformation received from thesystem; ii) meetings involving topmanagement to take timelydecisions and initiate proactiveaction, prompt follow-up andexpedite research; iii) regionalmeetings to meet specific regionalissues identified by the system.

ongoing information on clientstatus (new, active, exit, retained);ii) an Exit Monitoring Formfurnishing a better understandingof exit reasons,19 and; iii)anecdotal information aboutclients, competition and the localenvironment complementing theother two sources. While the firstpart of the routine componentgathers information from anexisting MIS, the second partforms the centerpiece of the EMSand answers questions relatedto the reasons of dropping out.Together, these help in buildingdetailed profiles of exit clients.The third part of the routinecomponent complements theinformation generated throughthe MIS and exit forms byfocusing on relevant informationcollected from the field

b) An ad hoc component that seeksto collect follow-up in-depthinformation on particular areas ofconcern identified by the routinecomponent through qualitativetechniques like focus groups,detailed case studies, etc.

19 In this example, the concerned organization (Partner) uses a self completion form to be filled in bythe clients themselves. Other organizations have used loan officers and/or other field and administrativestaff for the purpose.

Communication,Delegation,Implementation

Decision making

Consolidation,Analysis,

Follo

w u

p in

– d

epth

qua

litat

ive

rese

arch

with

pro

files

of i

nter

est

Tracking – routine datacollection

MIS – exit rates withbreakdown

Exit Forms – identifyingdesertion

Identifying dropout profiles External reasons – related “Resters” Internal reasons – related Left fo competition Expelled

Exit Definition

Table 1 reflects the above design from an operational perspective delineatingin detail the scope of activities together with timelines, delegation of responsibilities,data collection methods, etc.20



20 Adopted from Pawlak and Jahic (2004).

What?

AnecdotalInformation onexit

Reasons fordropout –information onstructure of exits

In-depthunderstanding ofreasons fordropout/problems experienceby dropoutgroups ofinterest

Information onexit magnitudeand its dynamics(exit andretention rates)

WhoCollects?

Loanofficers

Loansofficers

ResearchTeam

MIS officer/LTS Officer

How?

Informal talkswith clientsthat justbecameresters (after2 monthswithout aloan)

Self-completionforms withresters (after2 monthswithout aloan)

QualitativeResearchwith identifiedsegments ofdropouts(FGDs,PRAs, III)

Analysis ofMIS

WhoCompilesthe data?

Regionalmanagers

ResearchTeam

ResearchTeam

ResearchTeam

MIS Officer/LTS Officer

How?

At monthlyregionalmeetings in formof astandardizedreport (lastworking day ofthe month)

At organizationallevel

Basic statistics inExcel fororganizationallevel

Narrativefindings in a formof presentation

Reports toResearch TeamandManagement

Reporting

At regional level to otherregional staff by regionalmanagers

At an organizational levelto Management

Propositions presented tomanagement atorganizational level(Research Team –Headquarters members)

Results presented atregional level anddiscussed at monthlymeeting with regional staff(Research Team-regionalrepresentatives)

Propositions presented tomanagement atorganizational level(Research Team –Headquarters members)Results presented atregional level anddiscussed at monthlymeeting with regional staff(Research Team – regionalrepresentatives)Send to Research Teamand management

Detailed Analysispresented to themanagement

Howoften?

Monthly

Monthly

Semi-annually

Semi-annually

Monthlyquarterly

When?

Every monthexcept AprilandSeptember

Every Monthexcept AprilandSeptember

May,September

June,November

January, April, July, October

Table 1:

Table 2: Possible Criteria for Exit Client Profiles (MIS) 21

Loyalty Active as a client for how long?Age cohort Joined in which year of operation? How soon after opening a

branch in that locality? What sort of group do they belong to?Revealed indicators of How much have they saved? What is their revealed debt potentialcapacity?Individual client characteristics Records of arrears? Credit rating?

Age, gender, marital status, earner/dependency ratio, Education,training and business experience, Place of residence Poverty

Characteristics of clients main status (requires some form of routine means testing)business Sector or sub-sector, Age of business, Location of business



Knowing what kind of clients areleaving – as opposed to just howmany are leaving and after howmany loan cycles – can yieldimportant additional inputs thatcan be utilized in modeling betterproduct designs and targetingstrategies. Table 2 showspossible criteria for developingclient profiles based on differentsets of indicators that a good MISshould be able to record. Whenthe MIS show clients to haveexited, useful analysis could bedone on the basis of thesecategories to evolve profilesshowing what percentage of exitclients belongs in which category.

i. The Management InformationSystemA strong Management InformationSystem (MIS) that incorporateskey socio-demographic data inaddition to a client’s loan historymight be extremely useful as aquick resource for adding analyticalvalue to drop-out rates. Such datacan be especially helpful inconjunction with informationgenerated from exit monitoringforms that track reasons fordropping out. An MIS with thecapability of integrating informationfrom such forms is particularlyuseful. However, such integrationis desirable, not necessary.

21 See Copestake (2001).

comparative likelihoods of clientswith particular socio-demographiccharacteristics to leave or stay withinthe program, while helping the MFIto design better products for clientsin a specific regional, economicand demographic context.

However, an MIS that regularlytracks loan and client portfoliosof an MFI does not usuallyanswer why a particular client isleaving. For that purpose, oneneeds to have a separate toolwhich can preferably beintegrated with the MIS.



Table 3 below shows theexit/retention data of one MFIthat has been disaggregated bya number of variables includingsex, region (rural/urban), typeand status of business,household size, number of loancycles and marital status. In thisinstance, data shows that a largenumber of drop-outs are leavingthe program after the first loancycle and are predominantlymarried with a household size of3-4 members. These and othersuch pieces of information canbe valuable in ascertaining

Table 3: Exit and Retention Disaggregated by MIS Variables 22

Retained Drop-out Active Retained Drop-out ActiveStatus of Business Marital StatusNew 14.6 29.2 25.4 Married 83.9 80.1 79.5Existing 85.4 70.8 74.6 Non Married 16.1 19.9 20.5# Previous Loans Household Size0 40.1 62.7 50.8 1-2 18.6 17.9 22.41-2 54.3 32.3 35.3 3-4 64.1 68.7 63.2>2 5.6 5.0 14.0 >5 17.3 13.4 14.4Business Type Target GroupTrade 37.6 21.1 27.3 Women/rural 31.0 33.4 34.8Service 29.1 30.6 28.4 Women/urban 24.9 19.4 20.0Manufacturing 7.9 4.5 4.7 Men/rural 24.9 31.5 28.4Agriculture 2.1 5.9 4.8 Men/urban 19.3 15.7 16.8Animal Husbandry 23.3 37.9 34.8Total 734 378 8281 Total 734 378 8281

22 The table has been adopted from Matul and Pawlak (2004).



field tested in one or severalbranches before implementation.

For organizations that alreadyhave a strong MIS capable ofproducing lengthy client portfoliosand the capability to integrate itwith information from exit forms,the latter need not be veryexhaustive and repeat informationthat can be retrieved from theMIS. However, profiles based onexit forms are different from MISprofiles in that they are developedon categories based on reasonsof desertion. For example,Prizma, a MFI based in Bosnia,

ii. Exit Monitoring FormsThe causes of client desertionare many and complex. It isimportant for a microfinanceprovider to exhaust all informationresources – existing literature ondropout causes, focus groupdiscussions and preliminaryinterviews with field and officestaff, as well as with active anddropout clients, etc. – to developa sense of all possible reasonsthat could lead to client desertion.The information collected canthen be used as a basis for thedevelopment of an Exit MonitoringForm, which should ideally be

Table 4: Exit Client Profiles of Prizma 24

Exit Client Category Profile(Voluntary - Satisfied) Bosnian, single women, service providers, those who have takenLost due to external reasons more than 3 loans in Prizma;Sleepers25 Croats, married women, traders, those with new businesses,

exiting after repayment of the 1st loan, small last loan size, thosewho never had any repayment problems

(Voluntary – Dissatisfied) Serbs, younger women, big number of dependents, traders, thoseEager to return if services having “old” businesses, big last loan sizeimprove(Voluntary – Dissatisfied) Serbs, older women, big number of dependents, service providers,Lost to competition those who have taken more than 3 loans in PrizmaForced out Sample too small to run any analysis on forced out

24 Adopted from Matul and Vejzovic (2004)25 The term “sleepers” is used as a designated category for those exit clients who were not using any otherfinancial services at the time of the interview and planned to resume their relationship in the future. A greatmajority of such ex clients is usually composed of those who experience seasonality in their businesses.

profiles that have a higherlikelihood of dropping out andthose that could be retained.

iii. Data Collection TechniquesThe type and mix of datacollection techniques employedwill vary with organizationalpreferences that are usuallyguided by organizationalobjectives and benefit-costanalysis. There are severaltechniques available that can bereadily used by organizationsinterested in developing aneffective EMS: surveys based onstructured/semi-structuredinterviews or self completionforms are examples ofquantitative tools while focusgroup discussions and individualinterviews (case studies) arequalitative tools that are usefulfor an in depth analysis.

However, each tool has somelimitations that need to be kept inmind before implementation(Table 4)26.



has developed five main dropoutprofiles on the basis of how itdefines dropout clients and theirreasons for leaving theorganization (Table 4). Both thecategories (formed on the basisof reasons for exit cited) andprofiles in the table come frominformation collected through exitforms in a sample survey of exitclients (See Annex 1).

Notably, the profiles could haveeasily come from a strong MISif the MFI had the capability tointegrate the two sources ofinformation in a meaningful way.Even in the absence of such acapability, the MIS alone canperform a useful functionwherever an EMS exists. Havingdefined categories of exit clientson the basis of reasons for exitand having simultaneouslydeveloped exit client profiles onthe basis of information from exitforms, an organization can thenhunt for existing clients withsimilar profiles in the MIS.Subsequently, designated staffcan focus more on clients with

26 Pawlak and Matul (2004).



Technique/Method

Focus group discussiondriven by discussionguide or by PRA tools(reasons to drop-outranking, productattribute ranking,financial sector trendanalysis, seasonalityanalysis, etc).

Individual in-depthinterviews

Semi-structuredinterview

Structured interview

Self completion survey.

Applications

To collect in-depth information on specificissuesLearning toolIdentifying wider range of reasons for exitBetter defining exit for measurementUnderpinning this process with a tool thatprovides good segmentation allows thistechnique to go into depth and reduce thenumber of focus groups to be organized, thusreducing the cost.

To collect very in-depth information on specificissues including more sensitive onesLearning toolUnderpinning this process with a tool thatprovides good segmentation allows you tolimit the number of interviews to identifiedmain categories of exited clients, thusreducing the cost of administration.

To collect relatively in-depth, representativeinformation. It combines advantages of bothqualitative and quantitative methods.If combined with MIS data the interview canbe very short as there is no need to ask forall credit history and socio-demographicinformation.Good to precede with qualitative research toidentify possible reasons and adjust working.

To collect general representative informationLow skills needed to collect the dataIf combined with MIS data the interview canbe very short as there is no need to ask forall credit history and socio-demographicinformationGood to precede with qualitative research toidentify possible reasons and adjust wording

To collect general representative informationLow cost of data collection as no need forinterviewersIf combined with MIS data the form can bevery short as there is no need to ask for allcredit history and socio-demographicinformationGood to precede with qualitative research toidentify possible reasons and adjust wordingHelps to avoid staff bias in an internally carriedout investigation.

Limitations

Not representativeRequires homogenous groupsDifficult administration and data processingthat requires appropriate training, skills andpractice in qualitative researchDifficult to capture sensitive issues and talkat the community level about exit reasons ofparticular individuals.Difficult to mobilize groups with ex-clients.

Not representativeCostlyLimited possibility to identify wide range ofreasons (unless high number of interviewsconducted)Difficult administration and data processingRequires appropriate training, skills andpractice

Needs relatively large number of interviewerswith qualitative interviewing skillsNeeds a focused standardization effort andclose supervision during the data collection.Needs good analytical skills (both qualitativeand quantitative)Needs a strong ”learning culture” of theorganization to mitigate potential biases.

Needs quantitative analytical skills

Needs quantitative analytical skillsLack of control over a client when filling inthe form.Needs pilot testing and find tuning instructionsfor a client

Qua

litat

ive

Qua

ntita

tive

Table 5: Data Collection Methods

surveys that are self-administered by clients, there isthe extra advantage of lowercosts making it more attractiveto organizations with scarceresources.

Qualitative tools can either takethe form of focused discussionson specific issues with a groupof exit clients or in-depthunstructured interviews withdropouts. These have theadvantage of being moreparticipatory with tremendousscope for providing staff withunique learning opportunities.However, such methods are notrepresentative and requirespecial skills to extract optimalvalue from the informationgathered. It is often also notadvisable to touch upon sensitivesubjects during focus groupdiscussions. Similarly, conductingin-depth sessions with individualex-clients can be costly if carriedout in large numbers. Conversely,conducting such interviews withonly a few exit clients will notbring out the whole range of



All data collection techniques canbe broadly divided into twocategories, i.e. quantitative andqualitative. Quantitative datacollection techniques are mainlysurvey based and collectinformation through a structuredor a semi-structuredquestionnaire, which could eitherbe self-administered orenumerated through designatedindividual(s). These survey basedtools have the advantage ofbeing representative of thepopulation or a subset of thepopulation if proper sampling isdone in advance. Essentially,semi-structured interviews givethe interviewee more freedom toanswer whereas a structuredinterview restricts him/her toeither giving short answers orchoose from a given number ofoptions. Whereas the formermethod has more scope forgathering qualitative data thatcan at times produce usefulunexpected answers, bothmethods have the advantage ofmaking statistical analysispossible. With self completion



an overarching scheme of routineexit monitoring forming what canbe passed off as an EMS.

Although the choice of datacollection techniques wouldusually guide the samplingmethodology, it is important tocollect data from groups that arecomparable in terms of servicesused and/or the period ofdropping out. Just as clients usingdifferent services are more likelyto have different patterns ofdropout causes, those exiting indifferent periods will reflect trendsthat are equally dissimilar ifimportant changes have beenintroduced in the meantime. If thepopulation is constituted ofdiverse groups such as thosecited above, it is critical to capturethat diversity through the use ofproper stratification methods.Further, it is equally critical to userandom sampling whenquantitative tools are being usedto extrapolate findings to thewhole population. Even wherenon-representative qualitativetechniques like focus groups are

causes responsible for highdesertion rates.

For most of the reasons citedabove, an effective EMS makesuse of a mix of tools to betterunderstand the problem in waysthat would also make decisionmaking easy and moreproductive. However, a feworganizations have successfullyused qualitative methods aloneas a way of routinely investigatingthe causes of client desertion.For instance, the Small EnterpriseFoundation (SEF) in South Africahas adopted a two stagequalitative approach based ongroup discussions and individualfollow up interviews to routinelystudy and monitor dropouts(Annex 6)27. Yet, a large numberof organizations have usedquantitative techniques with mostpreferring a sample surveyapproach: there are severalinstances where organizationshave conducted a one-off clientexit survey28. Very feworganizations, however, havecombined both approaches within

27 Simonowitz (2000)28 Many organizations have used the client exit survey tool developed by the AIMS AIMS project (Annex ____)



used, care must be taken toensure that the latter arehomogenous.

Moreover, it is important tocarefully weigh the question ofwho will interview exit clients. Anex-client will perhaps not be ableto open up if his/her previousloan officer is the interviewer.Similarly, the latter might betempted to distort intervieweeresponses if the same reflectsbadly on him. A few organizationshave tried to overcome suchbiases by making loan officersinterview clients that are notknown to them. However, thisimmediately throws up thedisadvantage of the interviewer’sinability to fully detect falseanswers. Further, getting loanofficers to interview ex-clients ofother loan officers might notcompletely eradicate bias. Theclose camaraderie that loanofficers enjoy, particularly in smallorganizations, might prompt themto protect each other in situationswhere their interests, as a cadre,are protected. This might make

the whole exercise biased andfutile.

A few organizations have tried todeal with such biases byminimizing the role of loan officers(and other office staff for thatmatter) in the actual conduct ofthe interview through introducingself-completion survey methodswhere ex-clients are themselvesmade to fill exit forms. Anotherway to minimize staff bias andmaintain the sanctity of findingsis through outsourcing the wholeexercise to a credible third party.This, however, is always anexpensive option and one that isseldom taken by organizationsplanning to monitor exit clientson a routine basis.

b) Institutionalizing theProcess

For any EMS to bear fruit, it has tobe strategically integrated within thefunctions, operations and culture ofthe organization. Not only will anorganization need to functionally



entrench exit client monitoring withinits strategic objectives, it will furtherneed to procedurally integrate theprocess within its operations.Additionally, true institutionalizationwill not be possible unless bothmanagement and staff understandhow important client retention is tothe organization’s social and financialobjectives.

Operationally, it is critical to developthe necessary work plans andtimelines; to assign and delegateresponsibilities; to clearly defineinternal communication channels anddelineate precise paths for verticaland horizontal data flow; to addressthe important question of whatmeetings will be held, at whatintervals, and at which levels fordiscussing issues identified by theinformation collected; and to finallydelegate responsibilities for takingdecisions and overseeingimplementation (Box 1).



Box 1: Institutionalizing the Process: An Example

Different Levels and Functions InvolvementTo make sure that the information is used effectively there is a need for involving different organizational levels,functions, positions in different phases of the information flow through assigning different roles and responsibilities.…all the organizational levels and functions have their roles to play in system maintenance, and in particular:

Central level:Management

Set strategies based on the organizational goals, opportunities and challenges. These strategies are the basis forthe system goals in terms of information collection priorities.Use the information from the system to revise the effectiveness of their strategies.Use the information from the system to revise strategies and make decisions about the initiatives that [the organization]needs to undertake.Make decisions on allocation of resources to allow the system to function.Make sure that other processes and systems are compatible with the information system (support activitiesundertaken under the system).Ensure that results are used and initiatives undertaken based on them implemented in a timely manner by responsiblemanagers.Make sure that all inefficiencies in the system’s functioning are eliminated in a timely manner.Make sure that the system is reviewed on an ongoing basis and provides reliable and high-quality information.Supervise exit-reducing and relationship-building strategies at an organizational and regional level

Regional level:Middle Management

Regional Managers supervise informal chats with exit clients and distribution and collection of forms twice a year.Regional Managers hold monthly meetings to discuss anecdotal information from the field.Regional Managers develop regional level strategies to fight exit as well as help frontline staff to build their individualrelationships with clients.Regional Managers aggregate information on exit from informal talks of frontline staff and other feedback.Regional Managers pass the regional and frontline feedback to the central level.Supervise implementation of exit reducing and relationship building strategies at a regional and frontline level

Frontline-level: Staff

Loan Officers develop individual strategies building relationships with clients from the moment of seeking informationuntil a client becomes a dropout (after 12 months from last repayment).Loan Officers hold informal talks with exit clients turning to potential resters every month and twice a year distributeand collect self-completion forms.Loan Officers feed information back from the field and their ideas for new initiatives to Regional Managers.Feed the information back to clients about decisions made in the program.Implement individual, regional and organizational-level strategies to reduce dropouts and build long-term relationships.

Cross-level and cross-functional:Research Team

Administer the system – makes sure that all elements are working well, and problems are spotted and solved timely.Collect or supervise collection of data, input the data, analyze, come up with propositions to management anddisseminate the results at different organizational levels.Marketing Manager manages the Research Team, drafts operational plan, prepares and update job descriptions(together with HR manager), recruits new team members (together with HR manager), evaluates the team work,identifies capacity building needs and makes sure that they are being met in timely manner.Regional Representatives of the Research Team (one representative for one branch) coordinate system implementationat the regional level, spot any inefficiencies, problems and inform headquarters about them; gather feedback anddisseminate results.

Source: Pawlak and Jahic (2004)



interests. Nor will any system of exitclient monitoring generate enthusiasmwhen incentive structures do notreward retention in the same way asoutreach. In order to institutionalizea system of exit client monitoring, itis important not only to reform staffincentive structure by giving moreweight to retention, but also torationalize growth targets.

Insofar as exit client monitoring – asa learning and decision making tool– requires an accommodation of clientpreferences into organizationalpolicies, products and procedures, itdiscourages staff uptake andenthusiasm in a supply drivenorganization that does not preferflexibility in the first place.

An institutionalized EMS functioningas a decision making tool guides anorganization’s strategic vision towarda more client-based and demanddriven orientation accepting asubstantial degree of flexibility inproducts and procedures, in additionto focusing on building long termrelations with clients. A supply drivenorientation focusing on rapidexpansion, centralization, and thestandardization of products andprocedures usually produce a culturethat does not gel well with exit clientmonitoring as a learning and decisionmaking tool.

For instance, a strategic emphasison rapid expansion generally leadsto high operational targets and a staffincentive structure that rewardsoutreach growth and punishes highdelinquency, portfolio at risk andamount overdue in an employee’sportfolio. Such incentive structuresare likely to produce more dropoutsas loan officers (as well as groupmembers under a joint liability system)will be less interested in retainingweak and/or comparatively riskyclients detrimental to their own

References

ASA. (1996).“Dropouts in Microcredit Operations”. Dhaka: ASA

Burki, H. B. and Shah, M. (2007). “The Dynamics of Microfinance Expansion in Lahore”. PakistanMicrofinance Network. Islamabad: PMN

CGAP (2000). “Those Who Leave and Those Who Don’t Join: Insights from East African MicrofinanceInstitutions”. Focus, No. 16 (May). Washington D.C.: CGAP

CHIP (2005). “The Causes of the High Rate of Dropout Amongst Micro-finance Borrowers”. - CivilSociety Human and Institutional Development Program.

Churchill, C. (2000). "Banking on Customer Loyalty". Journal of Microfinance. Volume 2: pp. 1-21

Churchill, C.F. and Halpern, S. (2001). “Building Customer Loyalty: Measuring and MaximizingCustomer Satisfaction”. Micro Finance Network. Technical Guide No. 2.

Copestake, James. (2001). “Towards a General Framework for Client Monitoring Within MicrofinanceOrganizations”. Imp-Act Thematic Report No. 3. Brighton: Institute of Development Studies,University of Sussex.

Evans, T., Adams, A.M, Mohammed, R. and Norris, A.H. (1999). "Demystifying Nonparticipationin Microcredit: A Population-Based Analysis". World Development, Vol. 27, Issue 2: pp.419-430

Garuba, S. (2004). "Report of the Findings from the LAPO Client Exit Study Nigeria 2003". LAPO,Benin.

Hassan, G.M. and Shahid, N. (1995). “A Note on Reasons of Dropout from Matlab VillageOrganizations”. BRAC-ICCDR,B Joint Research Project. Dhaka: BRAC

Hulme, D. and Mosley, P. (1997). “Finance for the Poor or the Poorest ? Financial Innovation,Poverty and Vulnerability” in “Who Needs Credit ? Poverty and Finance in Bangladesh”.(eds.) G.D Wood and I. Sharif. Dhaka: University Press Limited

Imp-Act. (2004). “Learning from Client Exit”. Imp-Act Practice Notes no.3. Brighton: Institute ofDevelopment Studies, University of Sussex.

Kashangaki, J., Hulme, D., and Mugwanga, H. (1999).“Dropouts Amongst Kenyan MicrofinanceInstitutions”. Kampala: Microsave

Kashf. (2008). "Kashf Exit Client Study". Aftab Associates (Pvt) Ltd. Unpublished report

Khan, M.K. and Chowdury, A. (1995). "Why do VO Members Drop Out". Dhaka: BRAC



Kuwik, B., and Mashaba, K. (2000). “Dropout Study: Who is Dropping and Why?". Small EnterpriseFoundation.

Matin, I. (1998). “Rapid Credit Deepening And A Few Concerns: A Study of Grameen Bank inMadhupur”, mimeo

Matul, M. and Vejzovic, S. (2004). “Beyond Numbers: Prizma’s Exit Monitoring System” MicrofinanceCentre Spotlight Note # 10. .

Maximambali, F. (1999). "Client Exits (Drop-Outs) Amongst Tanzanian Microfinance Institutions".Kampala: MicroSave

Meyer, R. (2002). "The Demand for Flexible Microfinance Products: Lessons from Bangladesh".Journal of International Development, no. 3 (April). pp 351-368

Micro-Credit Ratings International Limited. (2005). “Client Drop-Out Rate”. Technical Note 1,Gurgaon, India: M-CRIL

Micro-Credit Ratings International Limited. (2007). “Estimating Client Exit Rate”. Technical Note1 revised. .

Montgomery, R.D.B., and Hulme, D. (1996). “Credit for the Poor in Bangladesh”. Finance AgainstPoverty: Volume II. Routledge: London.

Mosedale, S. (2001). "Case Study: Kashf Foundation-Pakistan". Enterprise Development ImpactAssessment Information Service.

Murdoch, J. and Haley, B. (2002). "Analysis of the Effects of Microfinance on Poverty Reduction".NYU Wagner Working Paper No. 1014. New York: New York University.

Murray, Inez (2001). “Cultivating Client Loyalty: Exit Interviews form Africa and Asia”. MicroBankingBulletin: Focus on Productivity, no. 6 (April).

Mustafa, S.I.A., Banu, D., Hossain, A., Kabir, A., Moshin, M., Yusuf, A. and Jahan, S. (1996).“Beacon of Hope: An Impact Assessment Study of BRAC’s Rural Development Programme”.Dhaka: BRAC

Mutesasria, L., Wright, G., Sempangi, H., Hulme, D. and Rutherford, S. (1998). "Drop-outs AmongstUgandan Microfinance Institutions". Kampala: MicroSave.

Nagarjan, Geetha. (2001). "Productivity Drivers and Trends”. MicroBanking Bulletin: Focus onProductivity, no. 6 (April): pp. 35-39.

Nelson, C. et al. (n.d). “Learning from Clients: Assessment Tools for Microfinance Practitioners”The Small Enterprise Education and Promotion Network, Washington D.C.: The SEEPNetwork

Omar, A. (2008). "Multan Exit Client Study: Survey Findings and Analysis". Unpublished researchstudy.



Pagura, Maria, E. (2003). “Examining Client Exit in Microfinance: Theoretical and EmpiricalPerspectives”. Unpublished Ph.D dissertation. The Ohio State University

Painter, J. and MkNelly, B. (1999). “Village Banking Dynamics Study: Evidence from SevenPrograms”. Journal of Microfinance, 1(Fall): pp. 91-116

Pawlak, k. and Jahic, S. (2004). “Promoting Client-focussed Organizations – Partner’s Exit MonitoringSystem”. Microfinance Centre Spotlight Note # 9. .

Pawlak, K. and Matul, M. (2004). “Client Desertion in Microfinance: How to Diagnose it Successfully?”Microfinance Centre Spotlight Note # 11. .

PromPT (1996). "Financial Services for the Rural Poor - Users' Perspectives". Dhaka: PromPt

Rosenberg, Richard. (2001). “Measuring Client Retention”. MicroBanking Bulletin: Focus onProductivity, no.6 (April), pp. 25-26

Schreiner, M. (2001). "Scoring Drop-Out at a Microlender in Bolivia". Center for Social Development.St. Louis: Washington University

Schreiner, Mark. (1997). “A Framework for the Analysis of the Performance and Sustainability ofSubsidized Microfinance Organizations with Application to Banco Sol of Bolivia and to theGrameen Bank of Bangladesh”. Unpublished Ph.D. dissertation, The Ohio State University.

Simanowitz, A. (1999). “Client Exit Surveys: A Tool for Understanding Client Drop-Out”. Journalof Microfinance: Volume 2, Number 1: pp. 112-136.

Stark, E. and Nyirumuringa, P. (2002). "Dropouts in Northern Province, South Africa". Kampala:MicroSave.

Waterfield, Chuck. (2006). “The Challenges of Measuring Client Retention”. Putting Client Assessmentto Work Technical Note # 2. Washington: Practitioner Learning Programme, The SmallEnterprise Education and Promotion Network.

Wilson, K. (2001) “Exodus: Why Customers Leave”. MicroBanking Bulletin: Focus on Productivity,no. 6 (April): pp 17-19

Wright, G. (1997). “Drop-outs and Graduates – Lessons from Bangladesh”. Nairobi: Microsave

Wright, G. (2000). “Microfinance Systems: Designing Quality Financial Services for the Poor”. NewYork: The University Press Limited

Wright, G. (2001). “Market Research and Client-Responsive Product Development”. Kampala:MicroSave.

Wright, G., Mutesasira, L., Sempangi, H., Hulme, D., and Rutherford, S. (1999) “Drop-outs AmongstUgandan Microfinance Institutions”. Kampala: MicroSave.



Annexures

Annex 1: Exit Client Case Studies (Pakistan)

Case Study 1:In January 2008, Mohammad Sajid Khanrequested for a loan of Rs. 18,000 fromone of PPAF Partner Organizations activein his area. At the time of the request,Sajid shared a rented shop with his brotherin one of the busiest market places ofShahdadpur and needed the money forhis mobile business. Unfortunately, theorganization wouldn’t lend more than Rs. 8,000 to first time clients. Sajid had little choicebut to accept the amount ofRs. 8,000 for which the concerned loan officer approved him.

At the time of the loan contract, Sajid and his brother were the only earning hands in a familyof ten. Sajid’s parents were old and unable to work, while his four sisters were of marriageableage. His elder brother, the other earning hand in the family, was already married and alsohad to cater for his wife and one child. Given all this, the Rs. 5,000 a month that Sajid usedto earn from his existing business was not enough to sustain the family.

In the loan officer’s assessment, the approved loan amount was expected to add anincrement of Rs. 2,000 to Sajid’s monthly income. Given the nature of the business, thiscould only have been realized on a sustainable basis through consistent reinvestmentsand/or access to a steady credit line. As household expenditures were continuouslymore than the pooled income of Sajid and his brother, the former was never a seriousalternative. The family’s reliance on the provision of a steady credit line was thus crucialto their long term prosperity.

Previous Loans NoneLoan Size Rs. 8,000Interest Rate 20 % FlatLoan Duration 12 MonthsNo. of Installments 12Loan Purpose RetailingLocation Shahdadpur, Sanghar




Sajid’s small business yielded enough income for him to pay back the Rs. 800 monthlyinstallment on time. This was despite the fact that business was generally consideredto be good only in winters after the harvesting of cotton when consumption levels in thearea increased. Sajid’s problem, however, was of a different nature. For a youngenterprising individual with little capital assets, the fear of taking the additional risk ofjoint liability in case of loan default by a group member was a constant worry.

Organizational policy required that Sajid be part of a group of 3-5 individuals, all ofwhom would jointly be responsible for making sure that everyone paid his installmenton time. Additionally, all group members would contribute in paying the installment ofa fellow member in case the latter fails to submit his installment. The above policyfollows the traditional model of group liability which leverages social collateral to reducethe risk of default. As such, it is absolutely crucial to the long term sustainability andhealth of the organization.

However, in order to make the model attractive to individual entrepreneurs averse totaking on additional risks, it is also crucial to adopt specific policies geared towardsincreasing group cohesion and trust. The traditional model tries to achieve that goalthrough selecting individuals in groups who are already known to and comfortable witheach other. These preexisting bonds are then further cemented through regular meetings.In case of Sajid, the above safety valves were missing.

The other four group members were not known to Sajid before the formation of the group:he in fact had problems in remembering the name of one group member despite the factthat they were part of the same group for one year. The only group meeting that was everheld was at the time of the sanctioning of loan at the organization’s office. Additionally,the usual meetings that are generally held for the purpose of collecting monthly installmentswere also not held in case of Sajid’s group: the loan officer would collect monthly installmentsthrough individual visits to borrowers. Hence, the sort of group cohesion that the modelheavily relies on could never mature at least in the case of Sajid’s group.

A lack of trust in group members had other implications: In case other group membersare holding larger loans than Sajid, the latter would have a disproportionate liability.Sajid was particularly concerned about such an eventuality. The above fears led himto look for other alternatives. He found one in a bank with a nearby branch that he couldeasily access. Despite the fact that the paperwork and other procedures of the bankwere more cumbersome, the latter was willing to lend a loan of Rs. 30,000 to Sajid.More importantly, he did not have to worry about group liability.


ANNEXURES

Case Study 2:Janat Bibi, age 33, wife of Ali Hassan isa lady health worker who left her schoolingin 8th grade. Her husband owns someland and makes around Rs. 20,000 –25000 every six months. They have threesons and two daughters and they all goto school. Janat Bibi and her family arewell off and are living very comfortably.She requested for a livestock loan of Rs. 12000. According to her she applied for a loanbecause she thought it was a good way of keeping herself busy and earning money atthe same time. She had to return the loan as a lump sum with 20% interest. She boughtthree lambs, six months before Eid-ul-Azha and sold them for kurbani with a profit ofRs. 2500 on each lamb. She had no problems repaying her loan and had no issues withinterest rate. According to Jannat Bibi repaying her loan as a lump sum amount waseasy as her livestock was ready to be sold by that time and she also made some profitafter repaying her loan.

This was a very profitable experience for Jannat Bibi and she wanted to continue takingloans. But due to a sudden change in PO policy she had to change her plans. Accordingto the new policy if Jannat Bibi applied for another loan she would have to repay it onquarterly basis (every 3 months). Jannat Bibi is not happy with the new policy; she feelsher lending organization has taken a very unintelligent step. How is it possible for herto save money for an installment if she is still spending money on her livestock. Thelivestock she buys is not ready to be sold until after six months and people only buy herlivestock for Eid ul Azha so this quarterly installment process does not make any sense.

This change in policy has been causing a lot of problems. Jannat bibi stopped takingmore loans as she could not afford to repay her loan on quarterly basis. Other peoplein Jannat Bibi’s village were facing similar problem. According to them, if they had aregular cash flow they would be able to repay their loans on quarterly basis but sadlythat was not the case. Jannat Bibi and many others have dropped out of lending programdue to change in policy of their lending organization.

Previous Loans NoneLoan Size Rs. 12,000Interest Rate 20 % FlatLoan Duration 12 MonthsNo. of Installments 1 / lump sumLoan Purpose LivestockLocation Pano Akil, Sukkur



Case Study 3:Lahri is a 48 year old woman and has alarge family of five daughters and twosons. Her husband Dhangi died 10 yearsago. Lahri makes and sells pappar fromher home. Both her sons are tailors andearn around Rs. 7000 per month. Lahri,although illiterate, still makes around Rs.4000 per month.

Lahri has completed 2 cycles and did not face any problems in returning her loans. Lastloan she applied for was of Rs. 50,000 which was rejected. She wanted to buy a tailorshop for her sons. Her loan was rejected because her lending organization consideredthe loan limit requested too high.

Perturbed at the development Lahri has stopped taking loans. According to her shealways payed her installments on time and never caused her lending organization anyproblems. She feels that her Lending Officer does not trust her with their money. Shefeels insulted and has decided never to take another loan.

Lahri’s lending organization has its own policy but could have discussed their policiesand constraints more proactively enabling her to understand the underlying reasons forrejection of her loan request. In the process they lost a performing microfinance client.

Case Study 4:Perveen Bibi, 34, is a house wife and hasa 12 year old son. Her husband, AllahDittah is a laborer and earns Rs. 200 aday. Their son, Salamat, has a heartcondition and all that Allah Ditta earnsis used up in doctor’s fee and hospitalbills.

Perveen Bibi took a livestock loan ofRs. 10,000 with 18% service charge. She bought three goats (kids) thinking that shewill make some profit in goat raising and will be contributing in household expenditures.But her plans did not work out, two of her goats died. Perveen spoke to her social

Previous Loans 2Loan Size Rs. 30,000Interest Rate 15 % FlatLoan Duration 12 MonthsNo. of Installments 12Loan Purpose LivestockLocation Mithi, Tharparkar

Previous Loans NoneLoan Size Rs. 10,000Interest Rate 18 %Loan Duration 12 MonthsNo. of Installments 2Loan Purpose LivestockLocation Pano Akil, Sukkur


ANNEXURES

mobilize about this but she was very strict and made it clear to Perveen that she hadto return the loan within the given time frame. There was no way out for her. She soldher husband’s mobile to repay the loan.

Perveen said that she had applied for the loan because she had heard that it was agood way of earning money. But for her it was not profitable in fact she had to give herown money to pay off the loan. And representatives of her lending organization werenot at all helpful. Perveen has decided not to take another loan as she cannot go throughthe tension and frustration of repaying the loan all over again.

Case Study 5:Although illiterate, Ramo Harji, 48, is anenterprising woman with three daughtersand four sons. She makes and sells rilis(handmade traditional bedcovers) athome. Ramo has completed three cycleswith her lending organization. Andaccording to her all the loans she tookwere very profitable. The last loan (three years back) she took was of Rs. 20,000. Shereturned her loan in 12 installments and her social mobilizer confirms that her repaymentrecord was good.

Ramo is member of a CO which has 39 members with more than Rs. 300,000 of groupsavings. Ramo and her CO, keeping in mind the large amount of savings, have startedinternal lending. They call their organization Darzi Paru and work along the same linesas their previous lending organization. Darzi Paru has regular borrowers even thoughit charges higher interest. These borrowers are willing to pay higher service charge asthe profit Drazi Paru makes will eventually be distributed among the members and theirmoney will remain within their group.

Darzi Paru feels that internal lending was a very good idea. According to Ramo theyhad learned what they needed to learn from their previous lending organization and itis always better to take a loan from people who are familiar and can be trusted. Ramoand her CO members feel quite comfortable applying for a loan because they know thatDarzi Paru will help them out if they faced any problems returning the loan.

Previous Loans 3Loan Size Rs. 20,000Interest Rate 15 % FlatLoan Duration 12 MonthsNo. of Installments 12Loan Purpose LivestockLocation Mithi, Tharparkar



Case Study 6:Momal Bai, 37, a matriculate used to bea teacher. Jodho Mal, her husband, didnot allow her to continue teaching aftermarriage. They have four sons; three arecurrently enrolled in a nearby school andthe youngest, 3, will start attending schoolwhen he is old enough.

Not so long ago Momal Bai and her family were facing hard times. Her husband usedto make around Rs. 5,000 per month, which was not enough to cover their expenses.It became obvious to Momal Bai that she had to, somehow contribute and help herfamily.

In January 2008, Momal Bai requested one of PPAF POs for a loan of Rs. 10,000. Shewanted to start her own business of selling bangles to women in her village. Afterreceiving the loan she started a small bangle shop in her house.

Momal Bai started earning through her business but was not left with any profit afterpaying her installment each month. She realized that service charge on each installmentwas too high. She kept on paying her installments and after one year when her loanwas repaid she decided that there was no point taking another loan as there was noprofit in doing so. Even if she did make any money it was used up in paying servicecharges.

But Momal Bai’s family is better off now. Her husband now works two jobs, (combinedearnings: Rs 22,000 per month) and earns enough to support the family. So now,according to Momal Bai there is no need to take any more loan as her family does notneed it. And even if she did need to take another loan, she would take from a relative.Her reason is that taking another loan from a PPAF PO would add to her problems andshe could not afford to do that.

Previous Loans NoneLoan Size Rs. 10,000Interest Rate 15 % FlatLoan Duration 12 MonthsNo. of Installments 12Loan Purpose Selling BanglesLocation Mithi, Tharparkar


ANNEXURES

Annex 2: Practices of POs (Pakistan)

MIS* Is DR Reporting Formula Reference Resting Can MISReported Document (DR) Period Period Compute(Y/N) (for DR) (for DR) (for DR)** DR

Formulas?SindhTRDP Operational N - - - - YSRSO Operational N - - - - YMRDO Not N - - - - -

OperationalIRC Not N - - - - -

OperationalSAFWCO Operational N - - - - Y AMRDWO Not N - - - - -

OperationalOCT Operational N - - - - Y

PunjabKashf Operational Y QPR Exit / Month, 10 days Y

Active QuarterAssassah Operational Y QPR Exit/Total Cumulative 6 months Y

10 daysDaman Operational N - - - - YPRSP Operational N - - - - YCWCD Operational N - - - - YCSC Operational N - - - - YBunyad Operational N - - - - NRCDS Operational N - - - - YJWS Not N - - - - N

Operational

DR: Dropout Rate



Annex 3: Exit Monitoring Form of Partner(Bosnia-Herzegovina)

Source:Pawlak, K. and Jahic, S. (2004). “Promoting Client-focussed Organizations – Partner’s Exit MonitoringSystem.” Microfinance Centre Spotlight Note # 9. www.mfc.org.pl

I. INTRODUCTIONII. INSTRUCTIONS FOR RESPONDENTSIII. CORE QUESTIONS

1. How have your business activities been affected by the loan last taken from Partner?

PLEASE MARK THECIRCLE WITH MOSTRELEVANT ANSWER

1. Helped to start a business2. Helped to expand my / the business3. Helped to keep my / the business going4. Decreased the scope of my / the business activity5. Influenced on my / the business closure6. Other – please specify what:


ANNEXURES

2. Now please think about products and services offered by Partner. What is your opinion on eachcharacteristic of the loan(s) that you got from Partner presented in this table.IN EACH VERSE MARK WITH THE CIRCLE THE MOST RELEVANT ANSWER.

WHEN ANSWERING PLEASE USE THAT SCALE: 1-definitely 2-rather 3-neither 4-rather 5-definitelyBad Bad good nor good good

badA. Office location (accessibility of the office location) 1 2 3 4 5B. Working hours (convenience of working hours) 1 2 3 4 5C. Loan officer’s behavior (his/her politeness, being nice) 1 2 3 4 5D. Reactions to your suggestions and complaints 1 2 3 4 5E. Simplicity of procedures (how easy it is to comply with 1 2 3 4 5

all the requirement to get a loan)F. Loan size 1 2 3 4 5G. Loan term (the time within which one can repay the loan) 1 2 3 4 5H. Method of loan repayment (repayment through a bank) 1 2 3 4 5I. Installment size (how well the amount of monthly 1 2 3 4 5

repayments are adjusted to your needs and businesscapacity).

J. Repayment in even installments 1 2 3 4 5K. Collateral, guarantee requirements 1 2 3 4 5L. Processing time (how quickly one can get a loan, 1 2 3 4 5

when one submits all the required documents)M. Interest rate 1 2 3 4 5N. Provision fee 1 2 3 4 5O. Grace period (holidays in payment at the beginning 1 2 3 4 5

of the loan)

3. What is your overall opinion on your co-operation with Partner?

PLEASE MARK WITH THECIRCLE THE MOSTRELEVANT ANSWER

Definitely bad 1Rather bad 2Neither good nor bad 3Rather good 4Definitely good 5



4. Please complete the sentence choosing ONE statement that describes youbest: Using Partner loan...

PLEASE MARK WITHTHE CIRCLE THEMOST RELEVANTANSWER

A. Experienced business problems

B. Experienced personal problems

c. I have found a better source of credit

D. I am not satisfied with your staff

E. I am not satisfied with your loan product

F. I am not satisfied with your procedures

G. My business doesn’t need any

additional

H. I have no business any more

I. Other reasons – please specify what?

Here are given possible factors that mighthave influenced your decision of not takingnext loan from Partner after repaying theprevious one:

6. In this column pleasemark the main cause thatinfluence your decisionnot to take next loan?

PLEASE MARK WITHTHE CIRCLE ONLY ONE– THE MOST RELEVANTREASON.

7. in this column pleasemark any other, lessimportant cause(s) thatinfluenced your decisionnot to take next loan?

PLEASE MARK WITHTHE CIRCLE ALL OTHERRELEVANT REASONS.

1

23

4

5

6

7

8

1

23

4

5

6

7

8

I was not able to satisfy any of my business needs forborrowingI was able to satisfy hardly any of my business needs forborrowingI was able to satisfy only few of my business needs forborrowingI was able to satisfy most of my business needs for borrowing

I was able to satisfy all of my business needs for borrowing

1

2

3

4

5


ANNEXURES

8. Do you plan to take another loan from Partner within the next 10 months?Please indicated how likely it is that you will borrow from Partner again?


Definitely no 1Rather no 2Rather yes 3Definitely yes 4

10. Would you recommend in the future taking a loan from Partner to your colleagues, friends orany other people?


Definitely no 1Rather no 2Rather yes 3Definitely yes 4

9. Choose the sentence, which describes you best:

Partner is the only available source of credit for my businessChanging the source of credit (Partner) would demand toomuch effort from meI got used to partner and I like this institution. I see noneed to look for another onePartner is the best source of credit I can imagineAs a good client I am offered special terms by PartnerNone of the above

1

2

3

4

5

6

PLEASE MARK WITHTHE CIRCLE THE MOSTRELEVANT ANSWER



11. If you could improve something about Partner, what would would it be?

PLEASE MAKE SURE YOU ANSWERED ALL THE QUESTIONS!!!THANK YOU VERY MUCH FOR YOUR HELP.

iv. INSTRUCTIONS AND PART TO BE FILLED OUT BY A LOAN OFFICER (separate piece ofpaper)


ANNEXURES

Annex 4: Exit Monitoring Form of Prizma(Bosnia-Herzegovina)Source:Matul, M. and Vejzovic, S. (2004). “Beyond Numbers: Prizma’s Exit Monitoring System.” MicrofinanceCentre Spotlight Note # 10. www.mfc.org.pl

X1. Client ID X2. Client Name X3. Telephone No. X4. Interviewer Name

(This form is supposed to be administrated by loan officers guided by core questions. The form supportsan informal telephone interview with a use of the probing techniques. Nor the questions or pre-codedanswers should be used to prompt answers from the respondents.)INTRODUCTION:Hello! I am calling from Prizma – microcredit organization. In Prizma, we always want to improve outservices to adjust them to our clients’ needs. We are now in the process of calling active and former Prizmaclients to learn about your opinion on Prizma and its services in order to better tailor our offer to your needsand preferences. I would appreciate if you can devote 10 minutes of your precious time. I would like toemphasize that all replies will be treated in strictest confidence. The answers provided by you will bereported in general statistical tables. They will never influence any of your applications to Prizma. Do youagree?(encourage, but if not willing ask when you can call again)

WARM UP QUESTIONS:Do you remember Prizma? With who have you worked in Prizma? How was it? How did Prizma credit helpyou? How is your business? Have you constructed the house? Etc.

A-1. What are 2 things that you did not like the mostabout Prizma and its products and services? (if youcould improve Prizma services what two things willbe priority?)

Do not prompt! If difficulty in responding. You canmention broad categories: promotion, place, people,product, price, process….

A-2. Which of the external reasons (not linked toPrizma services) discouraged you to take anotherloan?(multiple responses possible)

Ask the question, if no response list possible reasons.Even if the respondent mentions spontaneously thereasons probe delicately on other listed reasons.

1. [_] loan size2. [_] repayment period3. [_] repayment frequency4. [_] installment size5. [_] collateral6. [_] guarantors7. [_] eligibility requirements (access)8. [_] group methodology policies9. [_] interest rate level10. [_] application fees11. [_] penalty system

40. [_] seasonality41. [_] business/household has no need for further Financing42. [_] other financing from informal sources43. [_] approval of credit in formal institution44. [_] closing down the business45. [_] lack of market demand46. [_] other business problems47. [_] family health problems48. [_] travel/migration49. [_] other personal / family reasons



OTHER:

A-3. Do you feel that problems in repayment (yoursor other group member) influenced the fact that youhave not taken the next loan in Prizma?

It is not a question if the respondent had delays butrather if the delays caused her desertion. Look in therepayment history and probe further if you feel therespondent is not honest.

60. [_] No (go to question X6 / A4)61. [_] Yes, because of my repayment problems62. [_] Yes, because of other group member

repayment problems

12. [_] staff professionalism13. [_] approach to client14. [_] staff flexibility15. [_] handling of non payment16. [_] application process (incl. simplicity of forms)17. [_] waiting time between application and

disbursement18. [_] way of loan repayment19. [_] office location20. [_] office opening hours & days21. [_] promotion and communication channels used22. [_] incentives for loyal clients23. [_] range of products24. [_] availability of other then credit products25. [_] difficulty in getting money back from other

group members

OTHER:

X5. CHECK if all the reasons mentioned in X5 are reflect in A1, A2, A3. If not get back to relevant questionand probe. If there is no space to put the respondent answer put it here:

100. [_] __________________________________________________

A-4. SUMMARY – decide together with respondent which are the 2 most important reasons (from A1, A2,A3, X6) why she has not taken the next loan:(put code and description; use the same codes as in questions A1, A2, A3, X6)

1. [ ]_________________________________ 2. [ ]_____________________________________

A-5. Which other financial services are you and your household members using now (are about touse)? (this question is totally independent of Prizma)

99. [_] I do not use any financial services

1. [_] Enterprise credit (incl. supplier credit) 2. [_] Leasing 3. [_] Housing credit4. [_] Consumer credit (incl. hire purchase agencies / department store credit)

A-6. which institutions are sources of the above (A-5) mentioned loans?(after spontaneous answer list all sources and probe carefully if the respondent or any of her householdmembers is using 9is about to use) any of the credit services)

1. [_] 2. [_] 3. [_] 4. [_] 5. [_] 6. [_]Family/Friend Suppliers Moneylenders Hirepurchase Other MCOs Banks

(loan sharks,Private persons)

A7. Do you plan to take another loan in Prizma? 1. [_] Definitely No (go to question A8) 2. [_] Probably No (go to question A8) 3. [_] Probably Yes 4. [_] Definitely Yes


ANNEXURES

A-8. SUGGESTIONS: What will encourage you to take another loan from Prizma? Other suggestions?

A-9. for interviewer only: please classify a respondent as a drop-out who: (select one statementin each column A-9a and A-9b)

A-9aWas voluntary and satisfied 1. [_](reasons not related to Prizma)

Was voluntary and dissatisfied(Prizma related reasons) 2. [_]

Was forced out by Prizma 3. [_]

A-9bDoes not have any need for credit 1. [_]services and will not come back toPrizma in a near future

Is not using now any other credit 2. [_]services and plans to come back toPrizma in near future (“sleeper”)

Will come back, to Prizma only if 3. [_]services will be improved/expanded(can use or not other services)

Is using now credit services of 4. [_]competition and do not plan to comeback to Prizma



Annex 5: Exit Monitoring Form of AIMSProject (USA)Source:Nelson, C. et al. (n.d). Learning from Clients: Assessment Tools for Microfinance Practitioners. TheSmall Enterprise Education and Promotion Network, Washington D.C.: The SEEP Network

Data quality revision:

Form reviewed by field team leader (date and initials)_________________________

Data entered on computer by ____________________(name) on ___________(date)

Form reviewed by data cleaning team (date and initials)________________________

Fill in before meeting with ex-client:

21. Client identification number________________ 22. Survey identification

number_____________

23. Interviewer number______ 24. Date of interview_______________________

25. Name of Client __________________________________________________

26. Address________________________________________________________

27. Type of borrower: Individual loan_______ Group loan________ Other______

28. Name of group (if any)_____________________________________________

29. Sex (circle): M or F

30. Entry date: <___/___/___> 31. Exit date: <___/___/___>

32. Number of program loans taken _______ 33. Size of last loan __________

34. Was final loan repaid by borrower? (circle): Y or N

35. If NO, amount in arrears or default?__________________________________

36. Amount of savings withdrawn_______________________________________

37. Amount of withdrawn savings used to pay off the last loan? ______________

38. Loan officer (who last worked with client): ____________________________

39. Circumstances of departure according to program MIS (Mark only one answer):

[__] 1. Client voluntarily left group/program

[__] 2. Loan group failed so client left

[__] 3. Group/program expelled the client (because of inability to pay, loan default)

[__] 4. Other


ANNEXURES

40. Type of business financed by last loan (Mark only one answer):[__] 1. Retail [__] 5. Animal raising[__] 2. Service [__] 6. Fishing[__] 3. Production/industry [__] 7. Other (specify):________

(Read to Ex-Clients ):“We would like to find out a little about why you are leaving our loan program so we canconsider future improvements. Please think of all the main reasons you decided to leavethe program. We will combine your answers with those of others to understand why clientsleave our program. Your answers will not be shared with anyone else. This will take onlya few minutes. Thank you for helping us.”

1. Who primarily made the decision that you will no longer be participating in the program(or continuing as a member of this group)? (Do not read answers. Mark only one answer)

[__] 1. I made the decision. (go to question # 3)[__] 2. Someone else in my family decided. Specify who__________________

Why?____________________________________ (go to question # 3)

[__] 3. The group made the decision. (go to question # 2)[__] 4. The program made the decision. (go to question # 3)

2. (If marked answer 3 in previous question) In your opinion, what factors led the group todecide to exclude your continued participation? (Do not read answers. Multiple responsespossible.)

[__] 1. Repayment problems[__] 2. Attendance problems[__] 3. Difficulties with other members of the group[__] 4. Other reason (specify): ______________________________________

3. What are the main reasons that you are leaving or left the program? (See the followinglist of possible answers. Do not read answers. Multiple responses are possible)



A. Problems with program policies orprocedures:

[__] 1. The loan amount is too small.[__] 2. The loan length is too short.[__] 3. I do not like the repayment schedule.[__] 4. The loan became too expensive (such

as interest, fees).[__] 5. The disbursement of the loans is not

efficient.[__] 6. I was unwilling to borrow because of

other conditions (such as obligatorysavings, obligatory training).

[__] 7. I did not like the treatment by the staffor had personal conflicts with staff.Who?______________________

[__] 8. I found a program with better terms.

Which one?_________________________Why is it better?_________________________________________________________

B. Problems with group lending:

[__] 9. The group told me to leave.[__] 10. The group disbanded.[__] 11. I had personal conflicts with other

members of the group.Explain______________________________________________________________

[__] 12. I was unhappy about group leadership.[__] 13. I was unable or unwilling to attend all

the group meetings (such as take toomuch time; have schedule conflicts)

[__] 14. I did not like the rules and/or the pressureestablished by group.

C. Client’s business reasons:

[__] 15. I have enough working capital now formy business.

[__] 16. My business is seasonal; I will borrow again when I need it.

[__] 17. I am graduating to a loan program thatmakes larger loans.

Which one?_____________________________

[__] 18. I am unable to repay the loans becauseof the weak condition of my business(for example, poor profits, low sales).

[__] 19. I decided to close the business and dosomething else (for example, get ajob, start a new business).Why? _______________________

[__] 20. I sold the business.

D. Personal reasons:

[__] 21. I cannot continue because I spent themoney on a crisis (such as illness,death) or a celebration (such asmarriage) in my family.

[__] 22. My spouse (or other adult incomeearner) left me so I do not have theability to continue the business.

[__] 23. I am pregnant or now have anotherperson to care for (lack of time or abilityto continue the business at the samelevel).

[__] 24. I am moving out of the area.[__] 25. A family member told me to stop

borrowing from the program.

E. Community and economic reasons:

[__] 26. My business was ruined by a disaster (such as robbery; fire; flood; hurricane).[__] 27. A major new competitor moved into the

area and many of my customers nowbuy from the competition.

[__] 28. Poor economic conditions have left mycustomers with less money with whichto buy my goods or services


ANNEXURES

4. In thinking about all the reasons why you have said you left the program, which categorybest describes your most important reason? (Read the answers. Mark only one answer.)

[__] 1. Program reasons—Client has problems related to the program requirements orpolicies (does not want to borrow again under present program terms; does not liketreatment by program staff; needs loan but decided to borrow from competitive source ofcapital).

[__] 2. Problems related to borrowing in a group (internal conflicts; does not like group pressure,frequency of meetings, or group leaders; was expelled; group failed).

[__] 3. Does not need capital now (has enough capital now; seasonal business is not active now;has graduated to larger loans from another source).

[__] 4. Business reasons—Related to economic activity for which client borrowed (was not profitableenough to continue borrowing; decided to sell or close business).

[__] 5. External reasons—Problems beyond client’s control that are not related to either the loanprogram or business (for example, personal reasons such as illness or death in family,leaving area, pregnancy, lack of time, departure of spouse; or economic reasons such asdestruction of business, new competitor, poor economic conditions affecting purchasingpower of customers).

The following questions are about your use of the loan:

5a. How did you spend your last loan? (Multiple responses possible. Mark the 3 largest categoriesof expenditure. Do not read answers.)

[__] 1. Start a new business [__] 6. Improve/expand business site[__] 2. Change type of business [__] 7. School fees[__] 3. Buy more inputs/stock [__] 8. Medical/funeral expenses[__] 4. Buy equipment/tools, and the like [__] 9. Savings[__] 5. Hire more workers [__] 10. Other (specify)____________[__] 99. Don’t know, or unwilling to answer

5b. Did the loans help your family? If yes, how? (Do not read. Multiple responses possible)[__] 1. More and better food [__] 6. Furniture, utensils, goods for your house[__] 2. Educate children/self [__] 7. Recreation; leisure activities[__] 3. Improve your housing [__] 8. Other (specify)_________________[__] 4. Medical costs/improved health [__] 98. Loans did not help family[__] 5. Clothing [__] 99. Don’t know



6. Which of the following best describes your experience in paying your last loan...?(Read the answers. Mark only one answer.)

[__] 1. Difficult to pay [__] 2. Within my capacity to pay[__] 3. Easy to pay [__] 99. Don’t know

7. During the last 12 months, did your income in the business…?(Read answers. Mark only one answer.)

[__] 1. Increase greatly [__] 4. Decrease some[__] 2. Increase some [__] 5. Decrease greatly[__] 3. Stay the same [__] 99. Don’t know

8. Which answer best describes the impact for you of these program loans...?(Read answers. Mark only one answer.)

[__] 1. Helped me quite a lot [__] 4. Loan was a burden[__] 2. Helped me a little [__] 99. Don’t know[__] 3. Didn’t help me at all

9a. (For group members only) Do you think you benefited from being a member of the group?

[__] Yes (go to #9b) [__] No (go to #10)

9b. (For group members only) Please tell me the specific ways in which being in a group helpedyou. (Do not read answers. Multiple responses possible.)

[__] 1. Helped me to make my repayments [__] 5. Allowed me to develop myleadership skills

[__] 2. Provided advice and support when I [__] 6. Gave me training and newneeded help personally information

[__] 3. Gave me business ideas and contacts [__] 7. Other[__] 4. Offered me new friendships (specify):_________________________________

(Read to Ex-Clients):“We are coming to the end of the survey. The next set of questions is about your opinionof the overall program.”

10. Which best describes your experience of participating in the program?(Read answers. Mark only one answer.)

[__] 1. Very good [__] 4. Bad[__] 2. Good [__] 5. Very Bad[__] 3. No Effect


ANNEXURES

11. Please name the two things you liked best about the program.

1. ____________________________________________________________________________2. ____________________________________________________________________________

12. Please name the two things you liked least about the program.

1. _____________________________________

_______________________________________2. ____________________________________________________________________________

13. What do you think should be done to improve the program for clients?_______________________________________________________________________________________________________________________________________________________________________________________________________________

14a. Do you think that you would rejoin the program in the future? (Read answers.)

[__] 1. Yes [__] 3. No (go to #15) [__] 99. Don’t know

[__] 2. Probably [__] 4. Only if specific changes are made (go to #14)

14b. Note the specific changes in the program that the ex-client desires before returning to theprogram.__________________________________________________________________________________________________________________________________________

15. Would you encourage a relative or friend to join this program the way it is now?

[__] 1. Yes [__] 2. No [__] 99. Don’t know

16. Any other comments?__________________________________________________________________________________________________________________________________________

(Read to Ex-Clients ):“Thank you so much for your time. We will use your answers to help us improve our programfor other borrowers. Good luck.” END OF INTERVIEW WITH EX-CLIENT.Observations by the loan officer about the ex-client and reasons for leaving:



17a. Does the information given above match with your understanding of the situation and theprogram records?

[__] 1. Yes (go to #17c. [__] 2. No (go to #17b.)

17b. (If no) Why not?_______________________________________________________________________________________________________________________________________________________________________________________________________________

17c. Was there any noticeable difference between this ex-client and other clients in his or hergroup?

[__] 1. Yes (go to #17d.) [__] 2. No (go to #17e.)

17d. (If yes) What was the difference? (poorer, richer, more outgoing, more shy, and so on)_______________________________________________________________________________________________________________________________________________________________________________________________________________

17e. Other comments:_____________________________________________________________________

__________________________________________________________________________________________________________________________________________

Interviewer: If you need to shorten the survey, consider eliminating some of the following, whichmay be of lesser importance: questions # 4, 6, 8, 10, 11, 12, 16, and possibly the last few questionsdesigned for the loan office


ANNEXURES

Annex 6: Client Exit Monitoring at SEF(South Africa)Source: Simonowitz, Anton. (2000). “Client Exit Surveys: A Tool for Understanding Client Drop-Out.” Journalof Microfinance, no. 6 (spring): pp. 112-137.

1) Group Meeting: A meeting is set up with as many of the drop-outs as possible. In this meetinga general discussion is held. At this stage it is important not to ask the reasons for drop-out.Clients who have left the program feel bad or feel pressured to rejoin. By explaining the povertymission of SEF and the fact that SEF is worried when people leave, the staff should makesure that clients feel relaxed and free to talk about their experiences in the project. The staffthen explain that they want to learn what the former clients thought was good and bad aboutthe project—former clients are the best people to learn from because they have nothing tolose if they tell the truth. Existing clients, on the other hand, may feel they will jeopardize theirposition if they say what they think.

A. The group meeting starts with looking at the participants’ experiences at SEF. What did they like about SEF (what was good)? What did they not like (what was bad)?

Participatory methods, such as voting forms, are used to ask specific questions as they arise.For example, how was the loan term; how was the loan amount; how was the support fromthe staff; how was the support from the group? Voting forms give a quick view of the rangeof opinions, which can then be used to facilitate a discussion about why people voted the waythey have and why there are differences of opinions.

B.The second stage looks at participants’ business experiences. Matrices or voting forms canbe used to look at participants’ business strength before the loan, at present, and at a numberof points during the loan. If the business status has changed, the staff ask why. They try tounderstand why it improved or if there were problems. This general discussion generates agood understanding of the clients’ experience in the program, and it is likely that the staff willhave a good idea of the reasons for drop-out without actually having had to ask.

The group meeting should not be too long (about 1 hour). The aim is to finish when peopleare still active, not when they are getting tired. In this way they will be happy to come backfor a followup meeting. At the end of the group meeting the staff facilitator explains that the



meeting yielded a lot of information which is very helpful to SEF. Participants are then askedif they would be available for individual discussions at a later date.

2) Information From Files: Following the group meeting, the files for drop-outs are gathered.Information can be triangulated with the group meetings, for example on loan sizes andbusiness types. Credit discipline, performance, impact monitoring, and comments on thedebtors card can also help triangulate information and add to the understanding.

3) Individual Meetings: Individual meetings allow for a more in-depth understanding of anindividual’s experience in the program and reasons for drop-out. Prior to this meeting, thefacilitator looks at the information gained in the group meeting and relates this informationto the four areas of potential problems: personal reasons, problems with the business, problemsin group/center, and problems with SEF procedures.

In the individual interview, the staff member probes the issues raised in the group meeting,trying to get a good understanding of the member’s experience and opinions. Finally, as thelast question, the drop-out is asked why she or he left. At this point, there will be a goodunderstanding of the experience, but not necessarily how these related together and whatwas the final motivation for her or his leaving the program.

4) Interview with the Field Worker: Finally, the staff member talks with the field worker (FW)and discusses the dropped members. Again, this helps to triangulate previous information,as well as improve the overall understanding.

5) Writing of the Report: Reports are written using these headings in the following format:A. Introduction and description of the process for the monitoring – how did you do it, whatproblems did you have, how was the group meeting (were people open and free?), etc.B. Description of the members who dropped – names, center, group, business, loans received,etc.C. Group discussion (according to the four headings).D. Individual members – results from discussion and information from files (according to thefour headings).E. Information from field worker.F. Analysis and conclusions – from the meetings, interviews, files, and FW, what can youconclude are the main reasons for drop-outs (according to the four headings)?G. Recommendations for the members who dropped (is any more follow- up necessary?), forthe center where the member dropped, and for SEF.


ANNEXURES

Recommendations from SEF’s Dropout Study

I. Personal ReasonsThere are many personal reasons given for drop-out, which include the following: death inthe family; personal or family illness; conflict in the family (e.g., husband stops memberparticipating); moving away (temporarily or long term). Personal reasons are important tonote for two reasons:

A. Often they are not the real reason but are given as an excuse, either because the memberis ashamed of failure or the member has not complied with TCP rules and is afraid to admitit (for example, poor loan utilization).B. Where the reason is because of something temporary, the member may want to return inthe future, or may return with encouragement.

Action to Reduce Drop-Outs

1. Understanding reasons for drop out: Where “personal reasons” are given it is importantto allow the member time to talk freely about her or his reasons for dropping and to talk aboutthe success or problems in the business. This may give FWS a chance to discover otherreasons that they may be able to help solve.2. The “personal touch”: Members should feel that TCP staff care about them as people notjust as loans. For example, if a member is ill or has a death in the family, the FW should visitthe member and perhaps a fund should be set up to make a small contribution toward funeralcosts. This contact will also help in encouraging the member to return once the mourningperiod is over, or once they have recovered from illness.3. Allowing the member to return: If the member gives reasons that indicate a temporaryproblem, the FW should encourage the member to continue to attend center meetings andmaybe to save. The FW should make an effort to maintain contact and give the member anopportunity to return in the future.Important: The FW should never try to force or convince a member to remain in the programor to return.4. Group formation: During group formation, it is important to discuss issues of potentialconflict at home created by the member starting or expanding her business.

II. Business FailureFailure of a business may be reported directly or can be seen from other information given:A.Business does not grow or goes down—business value does not increase, or it decreases.B.Member leaves business to take up employment—this employment is mostly not well-paid(such as a farm or domestic laborer), so it shows that the business was not succeeding inproviding a living income.



C.A member often has to be patched by the group or the center.

Common Reasons for Business Failure:

1. Too much sales on credit.2. Inappropriate loan size—too big for manageable repayments, or too small to do plannedbusiness.3. Too much competition.4. Lack of business skills—support not given by group/center/field worker to develop skills.5. Poor loan utilization.6. Profits not re-invested in business—due to high demands from the family, or poor businessmanagement.7. Unforeseen disaster—such as robbery, rain damage, or family crisis that takes money frombusiness.8. Part-time business—member works as well and is not serious about the business.


Through close monitoring and support of businesses, joined with good problem-solving, wecan help reduce the chances of business failure; we can deal with problems early so as tosolve any problems before the business fails and the member drops.

1. Understanding of the business: spending time before the first loan discussing the businesswith the member and group (looking at the market for the business, how it should be run, andhow it could grow) helps the member develop skills for running their business and gives theFW and group a focus for the type of support they should be giving.Included in this discussion should be competition, selling on credit, the need to re-invest inthe business, and often the need to diversify the business in order to grow. From this themember will develop a business plan, which will not just be how much and what she will buy,but how and where she will sell, and how she will grow her business (this need not be writtendown, but should be discussed).2. Appropriate loan sizes: Using the business value and impact monitoring information, theFW can assess the strength of the member’s business and her progress. Based on thisassessment, an appropriate installment plan for the business should be set. During thebusiness plan discussions, the FW, member, and group should discuss what loan size wouldbe appropriate for the business type, and the member’s planned activities.3. Support to business development and monitoring: The group, center, and FW should supportmembers in following through with their business plans. This support may sometimes includebusiness skills training, but most skills will be developed “on the job” through discussion of


ANNEXURES

problems and sharing of experience within the groups and centers and by the FWs.Regular checks need to be made by members/groups/ centers on the performance ofbusinesses so as to deal with problems immediately as they arise.4. In-depth discussions at center meetings: The financial and reporting side of center meetingsshould be kept short, and at least half an hour should be allowed for detailed discussions andoccasional workshops. Reports from group chairs should show problem areas. These reportscan be used to encourage discussion about the issues raised—for example, selling on credit,diversification, good business practice tips, etc. The FW can facilitate this. Sharing experienceof problems and solutions is the best form of business skills training and can help reducedrop-outs due to business failure.5. Good monitoring of loan utilization: Loan utilization checks and loan supervision visits mustbe taken seriously. It is important that the group chairs take responsibility for this job and thatthey do it well. FWs must ensure that this happens. Monitoring of loan utilization helpsmembers to take their businesses seriously and to avoid destroying the business by takingmoney from the business for their families. The monitoring also gives an early warning ofproblems, which can then be dealt with.6. Dealing with disasters: Members must be helped to deal with disasters and not to feel thatthey must leave the program because they are struggling to repay the loan and their businessesare failing (see personal problems).

III. Problems in Group/Center

A.Conflicts within groups or centers: Conflicts often arise from members not making theirrepayments. This results in other members having to spend time trying to find the memberto make them pay or having to make payments on their behalf.B.Patching for other members: the feeling of “working for others” is a major reason for drop-out in centers with patching problems. The costs of members making additional paymentsfor others, on top of their other costs, may be enough to cause business failure and drop-out, or it means that they are unable to make savings.

Patching results from deliberate nonpayment or problems of some members whose businessesfail. A major reason for this is poor group formation, where the members do not know andtrust each other well. This may result from the following:1. Rapid growth of a center: The centers grows too fast for the FW to ensure that the groupsare well formed.2. Inexperienced FW: Many drop-outs are from groups that were formed by trainees or newlyqualified FWs. Again, this is due to failure of the FW to recognize poorly formed groups.3. Pressure of targets on FW and branch: The need to reach targets can result in a FW orbranch manager (BM) pushing through groups that are poorly formed.



4. Deliberate “cutting of corners”: FWs may form groups that they know are not well formedand then train the group to answer questions from the branch or zonal manager in a way thatthe poor group formation is disguised.


Good group formation is the key to reduce conflict within groups and centers, and is one ofthe keys to reducing drop-out.

1. Deal with repayment problems immediately: Field workers must find out in each meetingwho is being patched and work to assist the member to pay, or settle the reason they are notpaying. When patching occurs, this must be dealt with immediately. Patching one week is aproblem; repeated patching every meeting causes a lot of discontent and leads to dropouts.2. Pressure from targets: Targets must not be set so high that they create a pressure on theFW or CO to pass badly formed groups. Other targets – such as dropout rate – should be setand the link between group formation and success in these other targets should be madeclear.3. Fast center growth: Fast growth is not good in TCP. This fact should be stressed and fastgrowth should be checked by BMs and the zonal manager. Again, targets should be developedwhich are more holistic, reflecting impact and keeping of members, not just numbers.4. Deliberate “cutting corners”: Checking of group formation is very important, however, thecurrent procedures should be reviewed to see if there are better systems to detect groupswhich have been trained to pass the group recognition test, despite being poorly formed.

IV. Problems with SEF Procedures

Many former TCP members complained about various aspects of the program. For manyissues, strong opinions were expressed, but there were no issues for which there was 100%agreement – even where most people were strongly against something, there was someonestrongly in favor of it. The following are recommendations based on majority and strongestopinions – any changes implemented should be piloted prior to being adopted.A.Repayment terms: Most drop-outs agreed that repayments should be monthly rather thanfortnightly. However, a few members do prefer fortnightly repayments. Some businesses,where income is spread throughout the month, seem more suited to fortnightly payments, butthe women running these businesses still express strong desire for monthly payments.B.Transport costs: Many members are having to pay high transport costs, which in somecases amount to far more than the interest payments on the loan. For people with small loansand new businesses, this may place great burden on their ability to succeed.C.Loan periods: Long repayment periods for small loans result in difficulties in maintaining


ANNEXURES

the business. Particularly at the start of the business, it is easier for a member to managerepayments over a shorter time period. As loan sizes increase, so should the repaymentperiod. However, some members are also concerned that 10 fortnight loans are too short.D.Staff support: Support and regular contact with FWs is valued by members and is importantto provide moral support as well as advice and skills.E. Loan size: A loan which is too large for a business may create problems and lead tobusiness failure, however, no cases of this have been reported from the people interviewedso far. Similarly a loan which is too small, for example, to buy enough stock to be viable, mayresult in the member having to spend household money which puts a strain on the householdand results in money being taken from the business.The fact that under the Visual Indicator of Poverty Test many members came into TCP whoare richer than the cut-off line under Participatory Wealth Ranking means that these peoplemay put pressure on TCP to give larger loans. Loan size did not come across as a very strongissue.F. Failure to re-form groups: When a member leaves a group they must be replaced. Thisbecomes very difficult if three or four members leave. In many cases the remaining membersare forced to drop because of their inability to re-form the group.


1. Monthly repayments: It is important for centers to meet fortnightly so as to establish regularcontact between members and with the FW, so that problems can be discussed and businessessupported. However, monthly payments should be piloted as either an option or as standard,either from the first or second loan.2. Transport costs: Alternative forms of disbursement should be implemented to reduce thecosts of members collecting their disbursements.3. Loan periods: There is agreement that 20 fortnights is too long for a first loan. Shorterperiods should be reviewed based on the pilots currently being done, and also in relation tothe issue of monthly payments. More time should be spent reviewing whether 10 fortnightsis too short.4. Staff: Staff commitment to the success of their members is important (see “personal touch”above). In addition, it is important for staff to be strict in following TCP procedures.5. Loan sizes: Maximum first loan sizes should be reviewed in the context of the businessprofiles being developed to ensure that they enable members to start a viable business.6. Re-forming groups: We should look at how to make it easier for a member left on her ownto continue with TCP. The possibility of allowing centers to re-group themselves once theyhave been members for some time should also be looked at.

Date post:	28-Jul-2020
Category:	Documents
Upload:	others
View:	5 times
Download:	0 times

CLIENT - Pakistan Poverty Alleviation Fund · Client Attrition in Microfinance: Experience and...

Documents