+ All Categories
Home > Documents > The BI Survey 8 participant summary - CXP

The BI Survey 8 participant summary - CXP

Date post: 11-Feb-2022
Category:
Upload: others
View: 4 times
Download: 0 times
Share this document with a friend
40
Complimentary participant summary This document is summarized from the full 489 page Survey document The world’s largest independent BI survey by Nigel Pendse Survey The Not for sale, not to be quoted
Transcript
Page 1: The BI Survey 8 participant summary - CXP

Complimentary participant summaryThis document is summarized from the full 489 page Survey document

The world’s largest independent BI survey

by Nigel Pendse

SurveyThe

Not for sale, not to be quoted

Page 2: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center ii

The BI Survey 8by Nigel Pendse

March 2009BARC Ltd

27 FalaiseWest QuayNewhaven

BN9 9GGUK

+44 (0) 20 8879 [email protected]

BI-Survey.com

The information contained in this report has been gathered through primary interviews withknowledgeable participants in the computer industry, through secondary research ofgenerally available documents, reports, and other published media, and from earlier studiesconducted by BARC’s BI research program. BARC has relied upon the validity and accuracy ofall information so obtained. BARC assumes no liability for inaccurate or omitted information.

This report is the exclusive and confidential property of BARC and is subject to limiteddistribution and restricted disclosure only, as specified in the license. Persons andorganizations reviewing copies of this document must regard this information as confidentialand proprietary. Any unauthorized use, reproduction, or transfer of this material is strictlyprohibited. Depending on the license level, excerpts from this report may be released to otheremployees and non-employees of the client’s organization or quoted in materials issued bythe client by obtaining permission to do so from BARC and the author.

BARC and its affiliates, partners, and members of their families may engage in performingservices for, and/or engage in business with, and/or hold equity positions in the companiesmentioned in this document.

BARC is not liable for errors contained in this document, its data and analysis, and/or forincidental or consequential damages in connection with the use of this document and itsinformation.

Copyright ©2009 Business Application Research CenterReproduction and public quotations forbidden unless authorized

Page 3: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 1

1 IntroductionThank you for participating in The BI Survey 8 last summer. The analysis ofthe data has now been completed, and the full Survey has just been published.As promised, we are now providing you with a free summary of the keyfindings.

In total, 3269 of you took part in the Survey, broken down as follows:

Less RemainingTotal responses 3269

Filtered during data cleansing 84 2.6%

Remaining after data cleansing 3185

Not yet considered buying a BI product 176 5.4%

Considered, but not yet bought a BI product 152 4.6%

Don't know 176 5.4%

Bought, but unable to answer detailedquestions on any BI product

59 1.8%

Total unable to answer questions 563

Total answering detailed questions 2622

Of which:

Users answering questions about their use of a BI product 1894 57.9%

Consultants answering questions about a customer's useof a BI product

256 7.8%

Total answering detailed questions about customer BIusage

2150

Vendors answering separate set of questions 472 14.4%

Figure 1 – Sample make-up

The BI Survey 8 follows on from seven successful editions of The BI and OLAPSurveys. From the previous edition, we widened the range of productsincluded from just OLAP to any BI tools, and changed the name appropriately.The seventh edition was therefore able to cover more products than everbefore, and this eighth edition has increased the product range still further,including not only products from the well-known BI giants, but also specialistproducts from much smaller vendors. But this change of scope does not meanthat we have reduced the range of OLAP products covered, nor the depth ofanalyses.

The BI Survey 8 benefits both from the experience of the seven previouseditions and the ability to analyze trends based on up to eight years of data.This in-depth analysis allows many myths about the BI industry to bedebunked.

The increased sample and wider range of products covered has allowed newanalyses to be added, so this year’s Survey document is not only the largestyet, but the growth alone this year is more than the size of the whole of thefirst edition.

Page 4: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 2

Edition: 1 2 3 4 5 6 7 8 Growth

Date published:Jul

2001Oct

2002Nov2003

Nov2004

Nov2005

Nov2006

Feb2008

Mar2009 %

Users and consultantsanswering detailed questions

589 669 1047 971 2100 1679 1901 2150 249 13.1%

Vendors answering a differentquestionnaire

365 472 107 29.3%

Total pages 143 225 282 325 365 380 420 553 133 31.7%

Editorial pages 109 192 248 288 323 341 368 489 121 32.9%

Total words 32,125 52,257 66,374 74,139 81,333 83,950 123,149 183,725 60,576 49.2%

Tables and charts 76 135 175 200 222 235 250 351 101 40.4%

Figure 2 – Survey growth since 2001

1.1 BI buyers compared with non-buyersFigure 3 shows the proportion of respondents who said their organizationshad purchased, considered but not yet bought, or not yet considered buying aBI product.

Clearly, larger organizations were more likely to have purchased BI tools thansmaller ones. But the purchase rate stabilizes at about 90 percent fororganizations with revenues from $100m and above, with a peak of 94.1percent in the $1bn to $15bn range. The purchase rate may seem high, but thisis a self-selected sample: people who work for smaller organizations with noBI usage are much less likely to be interested in this sort of Survey than thosewho have experience of BI.

Asian organizations were the least likely to have purchased BI products. Moresurprisingly, the purchase rates were slightly lower in Europe and NorthAmerica than in the rest of the world.

Page 5: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 3

BoughtNone yet

purchasedNot considered

BI purchase

All (2537) 87.1% 6.0% 6.9%

Organization revenue

<$500k (115) 73.0% 10.4% 16.5%

$500k to $1m (58) 75.9% 6.9% 17.2%

$1m to $20m (280) 74.3% 11.4% 14.3%

$20m to $100m (282) 81.9% 6.7% 11.3%

$100m to $500m (353) 90.1% 4.5% 5.4%

$500m to $1bn (180) 90.6% 7.2% 2.2%

$1bn to $15bn (494) 94.1% 4.0% 1.8%

$15bn to $25bn (76) 89.5% 7.9% 2.6%

>$25bn (180) 88.9% 4.4% 6.7%

Organization headcount

<500 staff (791) 79.3% 8.1% 12.6%

500 to 5,000 staff (774) 90.2% 5.3% 4.5%

>5,000 staff (817) 90.9% 4.7% 4.4%

Region

Asia (67) 74.6% 11.9% 13.4%

Europe (1313) 87.6% 6.7% 5.7%

Rest of the world (70) 91.4% 4.3% 4.3%

North America (766) 86.8% 5.0% 8.2%

Figure 3 – Purchase rates

1.2 Respondents’ perspectivesYou were asked whether you played a technical or business role when selectingproducts. Perhaps not surprisingly, purely business respondents were in theminority — questionnaires of this sort are more likely to be completed bypeople with a technical interest. Figure 4 shows the breakdown for severalgroups.

The first row covers respondents who answered the numerous detailedquestions in the Survey, so it includes both users and consultants, but notvendors or non-users. The second row shows all employees of actual orpotential user organizations, regardless of whether they have yet purchased.The third row analyzes consultants who answered questions on behalf of theircustomers, while the fourth row covers vendors who participated in theSurvey, but who were not asked to answer product usage questions. Finally,the last row analyzes the data from people who work for organizations thathave not yet considered using a BI product.

As might be expected, consultants generally had more technical roles, withmore than half having both business and technical roles. Conversely, non-users tended to have business rather than technical roles, although the purelybusiness users were still in a minority.

Page 6: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 4

Business TechnicalBoth business

& technicalOther

All (3185) 27.4% 29.4% 41.9% 1.3%

Not considered using BI yet (176) 42.6% 28.4% 26.1% 2.8%

BI Users (2212) 27.7% 31.1% 40.0% 1.2%

Vendors (472) 26.9% 25.2% 46.4% 1.5%

Consultants (325) 18.2% 24.3% 56.9% 0.6%

Figure 4 – Respondents’ roles overall

By geography, German respondents were relatively more likely to come from abusiness background than those from other countries, with US and UKrespondents much more likely to be technically oriented. This meant thatrespondents to the German language version of the questionnaire were morelikely to be business-oriented, with respondents to the English languagequestionnaire being more technical.

Business TechnicalBoth business

& technicalOther

Region

Europe (1387) 35.1% 20.7% 43.3% 0.9%

South America (60) 23.3% 33.3% 40.0% 3.3%

North America (842) 18.1% 43.5% 37.2% 1.3%

Asia (72) 16.7% 45.8% 33.3% 4.2%

Rest of the world (73) 16.4% 42.5% 41.1% 0.0%

Questionnaire language

German (797) 39.4% 14.9% 44.8% 0.9%

Spanish (123) 26.8% 31.7% 39.8% 1.6%

English (1084) 18.5% 42.1% 38.3% 1.2%

Other (430) 30.2% 28.6% 39.8% 1.4%

Major countries

Germany (617) 41.8% 14.1% 43.4% 0.6%

Italy (71) 33.8% 38.0% 26.8% 1.4%

Austria (87) 32.2% 14.9% 51.7% 1.1%

Spain (74) 27.0% 33.8% 39.2% 0.0%

UK (128) 22.7% 31.3% 46.1% 0.0%

USA (789) 17.9% 43.6% 37.4% 1.1%

Figure 5 – Respondents’ roles by geography

Page 7: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 5

2 DemographicsThis section analyzes the demographics of your fellow participants in theSurvey.

2.1 Geographic distributionParticipants in The BI Survey 8 worked for organizations located in about 60countries. As in previous years, the on-line questionnaire was available inEnglish and German, but this year a Spanish version was added for the firsttime. There were a significant number of responses from German-speakingcountries, though the largest number were, as usual, in the US.

Countries constituting less than 0.2 percent of the total sample are onlyshown in aggregate.

Page 8: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 6

Location Share of sample Region totals

Germany 26.4%

Europe total58.1%

UK 5.7%

Switzerland 3.9%

Austria 3.6%

Spain 3.1%

Italy 3%

France 2.3%

Sweden 2.1%

Denmark 1.8%

Netherlands 1.6%

Israel 0.7%

Belgium 0.6%

Portugal 0.4%

Norway 0.4%

Hungary 0.3%

Czech Republic 0.3%

Finland 0.3%

Ireland 0.3%

Luxembourg 0.3%

Greece 0.2%

Russia 0.2%

Slovenia 0.2%

Other European countries 0.8%

USA 31.6% North Americatotal

34.2%Canada 2%

Mexico 0.6%

Australia 1.9%

Rest of world total7.7%

Japan 0.8%

South Africa 0.7%

India 0.6%

Brazil 0.5%

New Zealand 0.4%

Chile 0.3%

Argentina 0.2%

Pakistan 0.2%

Singapore 0.2%

Colombia 0.2%

Peru 0.2%

Other countries 1.7%

Figure 6 – Location of respondents’ parent organizations(based on 1977 responses)

Page 9: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 7

Apart from being widely dispersed, your employers are also very likely to beinternational, with almost 60 percent being described as either internationalor global in scope, and only just over five percent described as local in scope.

2.2 Organization sizes by revenueGiven the multinational nature of the sample, it is no surprise that many ofyou work for very large organizations. The median revenue was about $425m,and the biggest (modal) group was the one to 15 billion dollar revenue bracket.

Figure 7 – Geographic scope of respondents’ organizations(Based on 2081 responses)

Local5.3%

Regional11.7% National

24.8%

International26.9%

Global/worldwide

31.3%

Page 10: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 8

Total organization revenue Percentage

Less than $500,000 4.9%

$500,000 to less than $1m 2.6%

$1m to less than $20m 12%

$20m to less than $100m 13.3%

$100m to less than $500m 18.5%

$500m to less than $1bn 9.1%

$1bn to less than $15bn 26.7%

$15bn to less than $25bn 4%

$25bn or more 8.8%

Figure 8 – Organization total revenues(Includes data from 1695 respondents answering product and organization revenue

questions)

2.3 Organization sizes by headcountJust as you typically work for big organizations with large revenues, they alsohave large numbers of employees. It is hard to say whether these largecompanies are representative of the market as a whole, but we did find somedifferences in the experiences and preferences of the larger organizationscompared with the smaller ones. In most analyses, we group the organizationsinto three roughly equal sized employee bands (in terms of numbers oforganizations), shown in Figure 9. The median number of employees is justover 1900, which is also in the most populated band of 1000 to 2499employees.

Page 11: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 9

Organization headcount Percent Size band

1 to 24 4.8%

Small30.8%

25 to 99 7.7%

100 to 249 9.2%

250 to 499 9.1%

500 to 999 11.5%Medium33.8%1000 to 2499 12.7%

2500 to 4999 9.7%

5000 to 9999 8.9%

Large35.4%

10,000 to 24,999 9.2%

25,000 to 49,999 6%

50,000 to 99,999 5.6%

100,000 or more 5.7%

Figure 9 – Organization size by employees(Includes data from 2014 respondents answering product and organization headcount

questions)

2.4 Vertical marketsWe asked users and consultants, “Which of the following best describes your

organization’s industry sector?” Figure 10 on the next page shows the results,based only on data from respondents who answered product questions in theSurvey.

Clearly, the sample comes from a wide range of vertical markets. No singleindustry dominates, but some are obviously more heavily represented thanothers. Banking, retail and insurance are the top three industries, as they werelast year (though with retail in third place last year). These industries havelong understood the value of BI, although one wonders if banking will remaintop after the credit crunch banking disasters, which largely occurred after thedata capture for this Survey was complete. However, collectively,manufacturing customers were the largest vertical segment. Healthcare rosefrom eighth to fourth this year, while automotive dropped from sixth to ninth.

Page 12: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 10

Industry Percentage of sample

Banking 8.2%

Retail 6.6%

Insurance and financial services 6.1%

Healthcare 5.6%

IT - services 5.2%

Government 5%

Manufacturing - industrial equipment 4.9%

Manufacturing - consumer packaged goods 4.3%

Automotive 3.9%

Business consulting and professional services 3.9%

Pharmaceuticals 3.4%

Food/beverages 3.2%

Wholesale 3.1%

IT - software 2.9%

Telecommunications 2.9%

Transportation 2.6%

Education 2.6%

Manufacturing - process 2.5%

Utilities and energy 2.5%

Oil, gas and chemicals 2.4%

Media and publishing 2.1%

Construction and architecture 1.5%

Marketing, advertising and PR 0.9%

Non-profit, trade association and charity 0.9%

Entertainment 0.7%

IT - hardware 0.7%

Agriculture 0.7%

Aerospace 0.6%

Hospitality and travel 0.5%

Mining 0.4%

Airline 0.3%

Military 0.2%

Figure 10 – Industry sector analysis(Includes data from 2663 product buyers providing vertical market information)

Page 13: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 11

3 Key findingsThis section is based on the Key findings section from the full Survey. Itsummarizes the main findings of the research, but can only skim the surface ofall the material in the 489 editorial pages of the full document.

3.1 The market Only 8.2 percent of employees in the typical BI-using organization regularly

use the BI products being reported on. The rate is slightly lower in largerorganizations.

Vendors and consultants think that BI deployments are significantly widerthan users actually report. For example, users report a typical (median)deployment rate of eight percent, but consultants report a rate of 9.5 percentand vendors, 14 percent; 36.4 percent of users report deployments to lessthan five percent of their organization’s employees, while only 17.2 percent ofvendors think that their products are used by less than five percent ofemployees. It seems that vendors believe too much of their own propagandaabout mass deployment of their products.

Page 14: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 12

Some products have a distinct vertical bias. For example, MicroStrategy isstronger than average in retail and telco; Information Builders in insuranceand healthcare; Oracle BIEE in IT services; and QlikView in manufacturing.

The biggest single group (42 percent) of respondents had a mixed businessand technical role, while 27.4 percent had a purely business role, and 29.4percent had a purely technical role. Respondents using MIK, Inform PM,Excel, Cubeware and Bissantz were the most likely to have a businessbackground, while respondents using Actuate, BusinessObjects,MicroStrategy, Microsoft AS and Cognos Reporting were most likely to have atechnical background. Cognos Analysis and Essbase users were the mostlikely to have a mixed business/technical background.

Vendor respondents overwhelmingly reported that BusinessObjects (withinSAP) and Cognos (within IBM) were their most significant competitors, with

Figure 11 – Vendor vs user perceptions of breadth of BI deployment

2.7

%

33

.7%

22

.6%

11

.7%

6.3

%

8.2

%

4.0

%

3.2

% 4.7

%

2.9

%

0.2

%

16

.9%

20

.1%

15

.7%

11

.4%

8.7

%

7.0

%

5.1

%

10

.4%

4.4

%

None 1 to 5% 5 to 10% 10 to15%

15 to20%

20 to30%

30 to40%

40 to50%

50 to75%

>75%

Users (1738)

Vendor (413)

Page 15: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 13

Microsoft some way behind. The fast growing QlikTech was more often citedas a significant competitor than Oracle (excluding Hyperion), SAP (excludingBusiness Objects), SAS, Information Builders, SPSS, Actuate or open sourceBI.

3.2 The selection process The most popular reasons for product short listing are Internet research,

previous experience, consultant advice and industry analyst research.

Conventional marketing activities like mail shots, webinars andadvertisements were much less influential. Advertisements were cited by only12 percent of respondents in 2008, compared to 20 percent in 2002.

Overall, respondents cited an average of 3.42 influences each. Sites spendinglarge amounts on license fees were more influenced than average by industryanalyst research, existing deployments and strategic vendor relationships, butless influenced than average by Internet research, product bundling, andadvertisements. Consultant advice was particularly popular in Europe,whereas industry analyst research was most popular in the US.

Page 16: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 14

To

tal

infl

ue

nce

s

Inte

rnet

rese

arch

Pre

vio

us

exp

erie

nce

Co

nsu

ltan

tad

vice

Ind

ust

ryan

alys

tre

sear

ch

Pro

du

ctal

read

yu

sed

else

wh

ere

inth

eo

rgan

izat

ion

Tra

de

sho

ws,

conf

eren

ces

and

exh

ibit

ion

sA

tten

din

gfa

ce-t

o-fa

cese

min

ars

Pre

ssar

ticl

esan

dca

sest

ud

ies

Str

ateg

icve

ndo

rre

lati

on

ship

Wo

rdof

mou

th

Web

inar

s

Bu

nd

led

with

anot

her

app

licat

ion

Ad

vert

isem

ents

Ven

do

rm

ails

hot

san

dm

arke

tin

gm

ater

ial

Oth

er

All (2330) 3.42 42% 41% 36% 36% 27% 26% 25% 22% 21% 19% 15% 13% 12% 5% 4%

Parent organization revenue

<$100m (624) 3.23 45% 37% 37% 29% 19% 24% 21% 23% 17% 20% 16% 11% 14% 6% 3%

$100m to $1bn (509) 3.51 42% 38% 39% 32% 20% 33% 28% 26% 19% 22% 16% 16% 10% 5% 4%

$1bn to $15bn (479) 3.74 43% 47% 34% 53% 36% 28% 25% 20% 25% 19% 13% 13% 12% 4% 3%

>$15bn (241) 3.71 35% 49% 34% 42% 44% 24% 27% 18% 34% 15% 16% 14% 10% 3% 6%

Parent organization headcount

<500 staff (685) 3.20 43% 39% 37% 27% 19% 25% 22% 23% 18% 20% 14% 11% 13% 6% 3%

500 to 5,000 staff (726) 3.40 43% 38% 37% 34% 21% 30% 28% 21% 17% 19% 15% 15% 12% 6% 3%

>5,000 staff (777) 3.72 41% 47% 36% 46% 39% 24% 25% 22% 28% 17% 15% 12% 10% 4% 5%

Parent company region

Asia (57) 3.11 39% 44% 26% 35% 32% 9% 26% 16% 19% 16% 19% 11% 12% 4% 4%

Europe (1220) 3.37 40% 39% 40% 32% 25% 32% 26% 22% 18% 19% 11% 13% 12% 4% 3%

North America (696) 3.59 44% 45% 30% 42% 32% 19% 23% 21% 26% 19% 23% 13% 11% 5% 6%

South America (53) 3.53 43% 38% 38% 34% 23% 13% 25% 25% 30% 25% 15% 13% 15% 9% 8%

Rest of the world (65) 3.28 48% 43% 42% 45% 32% 5% 22% 23% 15% 12% 12% 12% 9% 5% 3%

Parent company country

Austria (78) 3.46 35% 35% 46% 37% 19% 46% 27% 21% 14% 26% 8% 17% 13% 1% 3%

Germany (555) 3.46 44% 36% 36% 37% 22% 44% 24% 24% 16% 17% 12% 14% 14% 4% 3%

Italy (60) 2.52 18% 43% 52% 15% 18% 15% 17% 20% 18% 8% 3% 5% 10% 3% 5%

Spain (66) 2.92 35% 29% 50% 18% 21% 15% 26% 23% 24% 18% 5% 11% 8% 9% 2%

UK (116) 3.38 45% 48% 39% 34% 35% 15% 24% 16% 25% 21% 12% 8% 9% 3% 4%

USA (652) 3.59 44% 45% 31% 43% 33% 19% 23% 21% 27% 19% 23% 13% 10% 5% 5%

Others (564) 3.40 41% 43% 39% 33% 28% 17% 29% 23% 21% 21% 13% 12% 12% 5% 3%

Questionnaire language

English (1451) 3.45 42% 45% 36% 38% 31% 17% 23% 22% 24% 19% 17% 11% 11% 5% 4%

German (780) 3.42 43% 35% 36% 33% 20% 44% 27% 22% 15% 17% 12% 16% 14% 3% 3%

Spanish (99) 2.84 30% 28% 41% 15% 21% 11% 32% 18% 29% 21% 4% 13% 7% 8% 3%

Figure 12 – Influences by organization demographics

Crystal Reports, Microsoft AS, BusinessObjects and Cognos BI were theproducts most likely to be evaluated. This varied by organization size: SAS,Hyperion and MicroStrategy more likely to be considered by the largestorganizations.

There is a clear tendency to favor local products. The small European vendorsare mainly only considered by local customers, while even long-establishedUS vendors like MicroStrategy, Information Builders and Actuate are muchmore likely to be evaluated by US than European organizations.

Page 17: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 15

Overall, 55 percent of BI buyers said they had conducted a multi-productcompetitive evaluation before buying, slightly lower than in 2007. Another 20percent said they had performed a formal single product evaluation, while theremaining 25 percent admitted to buying with no formal evaluation.

Products from Microsoft, Cognos and SAP were most likely to be boughtwithout a multi-vendor competitive evaluation. Panorama, Infor,MicroStrategy, arcplan and Actuate were most likely to have been chosen onlyafter a competitive evaluation.

As previous editions of the Survey have consistently found, organizations thathad performed a formal multi-product evaluation were more likely to achievetheir business goals, and much more likely to realize business benefits thanthose that did not. This may be partly because they were more likely to haveselected the most suitable product, and partly because their projects weremore thorough in other ways, too. This theory was supported by the findingthat those performing a formal single-product evaluation were significantlymore successful than those who performed no formal evaluation at all.

In competitive evaluations, Infor PM OLAP had the highest win rate, of 90percent. Targit and Panorama had competitive win rates of about 79 percent.

At the other extreme, Corporate Planner had a competitive win rate of just16.2 percent.

Win rates are higher in larger organizations, presumably because a higherproportion of evaluations lead to product purchases.

Overall, the primary reasons cited for selecting products were functionality(43 percent) and ease of use for end-users (38 percent). Fast queryperformance was only cited by 18 percent of buyers, low price by 16 percent,corporate standards by 11 percent and data scalability by 11 percent. The mostrarely cited reason was again the range of server platforms supported.

There were significant differences in the reasons given for selecting differentproducts. For example, functionality was frequently cited as a reason forbuying arcplan, Bissantz, Cubeware, Infor PM OLAP, MicroStrategy, MIK andWebFOCUS, with more than 50 percent of their buyers citing it as a principalreason for choosing those products. In contrast, only 12 percent of MicrosoftReporting Services and 17 percent of SAP BI/BW buyers gave functionality asa reason for choosing them. Ease of use for end-users was cited by more than50 percent of Cubeware, Excel, Panorama and QlikView users, but only sixpercent of SAP BI/BW and 13 percent of Microsoft AS users. Integration withexisting products was important to 54 percent of SAP BI/BW buyers, but lessthan ten percent of Bissantz, MicroStrategy and MIK customers. Fast queryperformance was a key factor for Essbase, TM1 and QlikView buyers, but foronly one percent of SAP BI/BW and two percent of Crystal sites. Userscalability was important for BusinessObjects, Essbase and Panorama buyers,but not to Bissantz, TM1 and QlikView customers. Low price, however, wasvery important for Microsoft customers, but did not concern buyers of CognosReporting, Essbase, MicroStrategy and SAP BI/BW.

SAP BI/BW buyers had largely corporate rather than product reasons forselecting it: integration with other SAP products (54 percent vs 19 percentoverall), corporate standards (54 percent vs 11 percent) and bundling (44percent vs 12 percent) were the main reasons cited for buying. Conversely,

Page 18: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 16

very few SAP sites chose it for product-related reasons such as functionality(17 percent vs 43 percent overall), ease of use (six percent vs 38 percentoverall) and performance (just one percent of BI/BW buyers vs 18 percentoverall). This was probably just as well, because SAP sites complained ofsignificantly more product problems than the norm.

3.3 Achievement of business goals Projects were slightly more successful in the 2008 sample than in 2007, but

about the same as in 2006: 29.5 percent reported that their projects had metor exceeded goals and 73.2 percent said they had at least largely met theirbusiness goals. At the other extreme, 7.3 percent had met no goals.

Business goals were most often met when projects were led by business end-users or specialist BI/OLAP consulting firms, and least likely to be met inprojects led by the software supplier’s consultants.

Implementation times made a big difference: the longer the time to go live,the worse the goal achievement rate. Projects that took over two years didextremely badly.

Among multi-product vendors, Business Objects and Cognos sites reportedthe lowest levels of goal achievement, while Microsoft and Hyperion sites hadthe highest levels.

Goal achievement increased as sites matured — the longer the period sincethe software was acquired, the greater the level of goal achievement. Thismeans that longer-established products tend to have higher goal achievementscores.

Exc

eed

edg

oal

s

Fu

llyac

hie

ved

go

als

Lar

gel

ym

etg

oal

s

Par

tial

lym

etg

oal

s

Har

dly

met

go

als

No

tm

etan

yg

oal

sye

t,b

ut

pro

ject

cont

inu

ing

No

tm

etg

oal

sat

all-

pro

ject

aban

do

ned

Met

or

exce

eded

go

als

No

tm

etan

yg

oal

s

All (2079) 9% 20% 44% 17% 2% 7% 1% 29.5% 7.3%

Time since purchase

5 years or more (454) 11% 25% 49% 12% 1% 2% 0% 35.9% 2.2%

3 to 5 years (391) 8% 22% 48% 17% 2% 2% 1% 29.9% 3.1%

2 to 3 years (298) 8% 18% 47% 17% 3% 6% 1% 26.8% 6.4%

1 to 2 years (358) 8% 17% 45% 21% 2% 5% 1% 25.1% 6.1%

6 to 9 months (120) 8% 23% 35% 23% 3% 9% 0% 30.0% 9.2%

9 to 12 months (162) 7% 22% 36% 22% 4% 10% 0% 28.4% 9.9%

3 to 6 months (121) 12% 18% 30% 17% 3% 18% 1% 30.6% 19.0%

<3 months (101) 10% 18% 30% 10% 1% 32% 0% 27.7% 31.7%

Figure 13 – Achievement of goals, analyzed by maturity

Page 19: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 17

Taking maturity into account, Panorama, Bissantz, Cubeware and TM1 siteshad particularly high levels of business goal achievement, while Crystal,Cognos Reporting and Targit sites’ levels were lower than expected.

Austrian and German sites were more likely to report higher levels of goalachievement and Spanish sites less likely.

There was very little correlation between license fee expenditure and theachievement of business goals.

3.4 Realizing business benefits ‘Soft’ benefits, like faster or more accurate reporting and better business

decisions are much more likely to be achieved than ‘hard’ bottom line benefitslike reduced costs and headcounts. The benefit least likely to be achieved is ISheadcount reduction. This has been the consistent trend each year.

A weighted Business Benefit Index (BBI) is calculated and used to compare asmany factors as possible throughout the Survey. This can be used by buyers tomaximize the business benefits from their own projects.

The overall BBI continued its gentle but steady decline in 2008. It peaked at4.314 in 2004, and was down to exactly 4.0 in 2008. This is its lowest valuesince the calculation was begun in 2003.

Panorama, Board, Bissantz, arcplan and QlikView sites were most likely toreport that they had realized benefits (ie, they had the highest BBI).

SAP BI/BW, Actuate and Cognos Reporting sites were the least likely toachieve business benefits.

SAP BW sites have consistently reported the lowest level of realized businessbenefits in each of the seven years that we have calculated the BBI.

No single product has topped the BBI league table in each year, and only oneproduct has beaten the overall average every year since 2003: Microsoft AS.

At the multi-product suite level, Hyperion sites were the least likely to achievebusiness benefits, and classic Oracle sites the most likely.

There is no clear link between BBI and license spend, but the BBI tends to fallas license fees rise.

Projects led by specialist BI consulting firms were the most likely to realizebusiness benefits, closely followed by projects led by business end-users,while those led by large, general-purpose consulting firms and independentcontractors were the least likely. This finding is consistent with previousyears.

Another consistent trend is that business benefits were most likely to beachieved if projects went live quickly. There was a steady fall-off as initialrollout times rose. The small number of projects that took more than twoyears to roll out achieved few business benefits

3.5 The power of the mega vendors As in previous years, SAP BI/BW sites had the lowest likelihood of realizing

business benefits. They also spent much more than average on both license

Page 20: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 18

and implementation fees, took by far the longest to go into production,reported worse than average levels of product support and encountered moretechnical problems than almost any other product. Its users complained thatthe product was slow, hard to use and was less reliable than average. All ofthese were also true in previous years. Despite these consistently poor resultsfor customers, BW projects again had rates of product abandonment barelyhalf the Survey average (though no longer the lowest).

The best-known BI vendors did not provide the best product support.Business Objects and Hyperion were far below average, and Cognos, Oracleand SAP were slightly below average. The best support was reported bycustomers of small regional vendors like Bissantz, with Panorama,Information Builders and MicroStrategy also doing well. TM1’s previousexcellent support rating has fallen after Applix was acquired by Cognos.Microsoft was the best of the mega BI vendors, with a score just aboveaverage.

Exc

ellen

t—

accu

rate

an

dti

mely

Gen

era

lly

accep

tab

le

Tech

nic

all

yg

oo

d,b

ut

slo

w

Tech

nic

all

yp

oo

r

Un

accep

tab

lyb

ad

Weighted supportscore

All (1900) 26.7% 52.5% 14.0% 4.5% 2.3% 6.44

Vendor size

Small vendors (607) 41.5% 45.0% 10.4% 1.8% 1.3% 7.49

Medium vendors (357) 33.6% 53.8% 8.7% 3.6% 0.3% 7.26

Large vendors (902) 13.4% 57.2% 19.0% 6.7% 3.8% 5.39

Figure 14 – Support quality ratings by vendor size

Since being acquired by SAP, BusinessObjects’ share of the SAP BI/BW clienttools market has shot up from 1.9 to 8.2 percent, and Crystal from 4.7 to 10.2percent.

Oracle’s BI Suite now connects to Essbase, but the current release is not yetoptimized for it (for example, it cannot yet handle Essbase’s sophisticateddimensional model). Nevertheless, 9.1 percent of Essbase users said they wereusing it as a client tool, far more than most third party tools, which workmuch better with Essbase.

Microsoft acquired Maximal Max in 2001, and soon re-launched it asMicrosoft Data Analyzer (MSDA). The product has not changed since 2001and it was quietly discontinued, with no formal announcement. Yet it shot upfrom being used by just three percent of Microsoft OLAP users in the firstOLAP Survey, to 28.3 percent in 2002, 36 percent in 2003, and down again to26.5 percent in 2004, still more than even the most popular of the third-partyclient tools. But in 2005, it had dropped to 12.3 percent, the first time it hadbeen below the most popular third-party client tools. In 2006, it droppedfurther to just 6.0 percent and rose slightly to 6.5 percent in 2007. Even in2008, the share was 5.8 percent, which remains higher than the shares ofmost of the actively maintained, much more functional, third-party client

Page 21: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 19

tools for Analysis Services.

Microsoft Reporting Services was first released in the second quarter of 2004,in a form that was far from optimized for Analysis Services. For example, ithad no ability to create MDX queries, and no interactive end-user reportdesign capabilities until the 2005 release. Despite this lack of suitability foruse with Analysis Services, and immaturity, more sites said they were usingReporting Services as an Analysis Services client than any other tool apartfrom PivotTables. Approximately six times as many Analysis Services siteswere using Reporting Services as a client tool than the most successful third-party client tool, Panorama NovaView. This remarkable success may havebeen helped by the fact that Reporting Services, like Analysis Services, isbundled with SQL Server.

3.6 Applications The five most common applications were ad hoc query and analysis; general

data warehouse reporting; sales and marketing analysis; budgeting andplanning; and dashboards and EIS, each being reported by more than 40percent of respondents.

Profitability/yield analysis; financial consolidation and statutory reporting;scorecards; and standard ERP reporting were also reported by more than 20percent of respondents.

Some much-hyped applications continue to be unpopular: for example, CRMhas declined to 12 percent (from 18 percent in 2002), and clickstreamanalysis is down to just three percent.

On average, respondents reported 4.26 deployed applications each.

Page 22: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 20

Analysis 1. Ad hoc query and analysis 63.4%

Reporting 2. General data warehouse reporting 50.0%

Analysis 3. Sales and marketing analysis 44.4%

CPM 4. Budgeting and planning 43.5%

CPM 5. Dashboards and EIS 40.6%

Analysis 6. Profitability/yield analysis 26.8%

CPM 7. Financial consolidation & statutory reporting 24.5%

CPM 8. Scorecards 23.4%

Reporting 9. Standard ERP reporting 20.8%

Analysis 10. Data mining 18.9%

CPM 11. Activity based management 17.5%

Analysis 12. Quality analysis 16.1%

CRM 13. CRM 11.9%

Analysis 14. Long-range forecasting 10.1%

Analysis 15. Scientific analysis 3.5%

CPM 16. Sarbanes-Oxley compliance 3.4%

Analysis 17. Clickstream analysis 2.8%

Other Other 4.0%

Average number of applications per site 4.26

Figure 15 – Applications categorized

With an average of 4.93 applications per site, Targit had the largest averagenumber of applications. Oracle BIEE (4.83), SAP BI/BW (4.79),MicroStrategy (4.73) and WebFOCUS (4.69) were just behind. At the otherextreme, Actuate (3.16), Excel PivotTables (3.59), arcplan and Infor PMOLAP (3.79) sites had the fewest applications.

Not surprisingly, the application profiles differed greatly between the 24products analyzed. This confirms that there is no single all-purpose BI tool,nor a single ideal BI vendor.

Organizations that conducted a multi-product competitive evaluation beforeproduct selection deployed the most applications.

The number of applications deployed rises with license fees, from 3.84 (under$5k) to 5.24 (>$1m).

Spanish organizations deployed the most applications (4.62) and Austrian thefewest (3.94).

3.7 Products In all, data was collected on 41 BI products. Many of these had only a handful

of respondents, but we generally do not separate out sub-samples of fewerthan 30 responses for analysis. For some more complex analyses, thethreshold for analysis is set higher, so the majority of the products are notanalyzed individually.

Page 23: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 21

Primary product Respondents % User organization/Consulting firm

Inclu

ded

inm

ostd

eta

iled,

pro

duct-

leve

lanaly

ses:

88.7

4%

MicroStrategy 256 11.91%

Infor PM OLAP 143 6.65%

QlikTech QlikView 143 6.65%

Board 141 6.56%

Microsoft Analysis Services 122 5.67%

Cognos Reporting 113 5.26%

BusinessObjects 103 4.79%

SAP BI/BW 102 4.74%

WebFOCUS 97 4.51%

Cubeware Cockpit 73 3.40%

Cognos TM1 Server 58 2.70%

arcplan 57 2.65%

Microsoft Reporting Services 57 2.65%

Panorama NovaView 54 2.51%

Bissantz 54 2.51%

Hyperion Essbase 48 2.23%

Crystal Reports 47 2.19%

Cognos Analysis 47 2.19%

Targit 45 2.09%

Oracle BIEE/BISEO 42 1.95%

Actuate Platform 39 1.81%

Microsoft Excel PivotTables 34 1.58%

MIK 33 1.53%

Genera

llyin

clu

ded

on

lyin

aggre

gate

dana

lyses:11.2

6%

LogiXML 28 1.30%

Oracle Discoverer 25 1.16%

SAS 22 1.02%

Palo 21 0.98%

Hyperion Interactive Reporting 17 0.79%

Microsoft PPS, ProClarity or BSM 14 0.65%

Other Business Objects product 11 0.51%

CyberQuery 10 0.47%

Other Cognos product 9 0.42%

Oracle OLAP Option 8 0.37%

Hyperion Planning 7 0.33%

SAP BPC (OutlookSoft) 7 0.33%

Hyperion Financial Management 6 0.28%

Others 57 2.65%

Total 2150 100.0%

Figure 16 – Products included in the Survey(The last column shows the customer/consultant mix of the respondents)

We asked respondents from multi-product sites to nominate the product theyknew best or which was most widely used for further detailed questions. We

Page 24: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 22

calculated a ‘prevalence rate’, based on the likelihood that products would bechosen in this way. WebFOCUS, MicroStrategy, Panorama and Board had thehighest prevalence rates, while Hyperion FM and Planning, Crystal Reports,Excel PivotTables and SAS had the lowest.

The size of the individual samples was affected by whether the vendor hadpublicized the Survey to its customers. In particular, the Business Objects,SAP, Oracle and SAS samples were smaller than we would have liked, whileMicroStrategy, Panorama and some of the smaller German products hadlarger than expected samples.

As in the last three years, the sample had a large German proportion, whichincreased the share of products like Bissantz, SAP BI/BW, Infor PM OLAPand Cubeware.

Several products from large vendors fell below the threshold and weretherefore excluded from individual analysis: Hyperion Interactive Reporting(former Brio), Cognos Planning (previously Adaytum), Hyperion FinancialManagement and SAS.

Crystal Reports, Microsoft Analysis Services and BusinessObjects were theproducts evaluated most often. Cognos BI was just behind.

Business Objects, Hyperion, MicroStrategy and SAP products were muchmore likely to be evaluated by larger organizations, whereas Microsoftproducts were more likely to be evaluated by smaller organizations.

Excel-based front-ends were the most popular for the OLAP servers thatoffered them. The percentage of sites using Excel for at least some users wasagain lowest for Microsoft AS at 80 percent, rising to 85.5 percent forEssbase, and 91 percent for SAP BI/BW and TM1.

3.8 Purchases Many respondents reported that their organizations had purchased

concurrent (33.1 percent) or unlimited server/enterprise (21.7 percent)licenses. Only 55 percent said their organizations had purchased named userlicenses. The totals do not add up to 100 percent, as some organizations had amixture of licenses.

Of those who had purchased named user licenses, the median seat count wasonly 60, and the mean 705, but as this excludes unlimited licenses, itprobably understates the typical seat count. Eleven percent of named usersites had over 1000 seats.

Median named user seat purchases varied greatly by product, ranging from825 in SAP and 250 in Business Objects sites to under 40 for QlikView,Targit, Board, Bissantz and Cubeware sites.

Page 25: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 23

Product Average seats Median seats% of sites with 1000 ormore licensed users

SAP BI/BW (32) 5173 825 46.2%

BusinessObjects (63) 1790 250 24.4%

Cognos Reporting (65) 2042 213 27.8%

Panorama (41) 824 150 30.3%

MicroStrategy (160) 571 150 15%

Cognos Analysis (31) 300 140 11.5%

Hyperion Essbase (29) 616 120 20%

Cognos TM1 Server (46) 531 98 10%

arcplan (39) 287 90 2.9%

All (1310) 705 60 11%

Infor PM OLAP (109) 129 50 3.8%

QlikView (119) 137 38 3.9%

Targit (35) 58 30

Board (112) 59 27

Bissantz (40) 43 27

Cubeware Cockpit (65) 47 19

Figure 17 – Seats sold, excluding unlimited licenses(Concurrent seats treated as three named users)

The typical license fees were almost 2.6 times higher on Unix than onWindows.

The average (mean) number of deployed (as opposed to purchased) seats wasonly 904, varying from 27 for Bissantz to 1553 for Oracle BIEE. The mediannumber of deployed seats was just 45, varying from 10 for Targit to 250 forSAP BI/BW.

We were worried by the high level of unused licenses, or shelfware, in earlyeditions of The OLAP Survey. Shelfware is inevitable in young sites, asorganizations will have purchased seats that they have not yet had time todeploy. Many of these may be deployed imminently. Very low shelfware ratesare not good, as it suggests that the vendor has made few recent sales.Equally, very high shelfware rates could be a sign that the product does notlive up to expectations.

We suggest that an overall shelfware rate of around 20 to 25 percent isprobably ideal, though the calculation is not possible in the many sites withunlimited user licenses. This may explain why our calculated shelfware rate,which necessarily excludes such licenses, has dropped to under 20 percent in2008. If correct, this is perfectly acceptable, but we suspect that it understatesthe true figure.

Of the few products with large enough named user samples to estimate theshelfware rate reliably, Panorama, Cognos Reporting and TM1 had thehighest rates, of 30 or more percent. Cubeware, Infor PM OLAP,MicroStrategy and BusinessObjects had rates below 20 percent.

Page 26: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 24

Two thirds of sites expect to deploy all purchased seats within a year (whichwill, of course, consume any shelfware they may have built up).

Almost 37 percent of sites expect to buy more licenses in the future and only21 percent expect not to. The ‘positive gap’ (the difference between those whodo and don’t expect to buy more seats) was 15.5 percent in 2008, the lowestfigure since the Survey started in 2001.

QlikView (50 percent) and MicroStrategy (46 percent) had the highest‘positive gap’ figures, and MIK much the worst with minus 30 percent.

The median time since the BI products were purchased was about 30 months,up from 27 months in 2007. It varied from nine months for Oracle BIEE tomore than 55 months for Essbase and MIK OLAP. A long average time sincepurchase suggests few recent sales and an aging customer base.

3.9 Cost of ownership The Survey attempts to provide some guidance on the costs of ownership per

deployed seat, but does not have access to data on internal costs that wouldallow a full cost of ownership to be calculated.

The median license fee per site was $65k, but fees ranged from over $349kfor MicroStrategy to $18.4k for Microsoft Analysis Services. Users of bundledproducts, like Analysis and Reporting Services, were asked to try to separateout the BI portion of their total license fees.

The median external consulting spend was $45.5, including only those usingexternal resources.

QlikView sites spent the least on consultancy ($8.5k) and SAP BI/BW sites byfar the most ($500k).

The high license and implementation fees in SAP BI/BW sites were balancedby the large median number of seats deployed, so its cost of ownership indexis not much higher than average.

The median (typical) number of full-time equivalent heads involved inrunning and administering BI projects is 1.0, with a mean of 3.2. SAP BI/BW,with much larger deployments than average, had a median of three heads anda mean of 6.6. The number of deployed users per admin head ranges from ahigh of 83 for SAP BI/BW down to eight for Bissantz. So, on a per capitabasis, the apparently high maintenance intensive SAP model works out asmore efficient than the low headcount required for products with few users.

3.10 Customer loyalty Customer loyalty is highly prized by vendors, because maintenance revenue

and sales of additional licenses to existing customers are much moreprofitable than winning business from new customers. Consequently, mostvendors work hard to achieve it and claim to have done so.

We measure customer loyalty indirectly by combining the quantified resultsfrom four other analyses: inclination to buy more seats, deterrents to widerdeployment, rate of product abandonment and standardization tendencies.

This year, of the products with enough data for all four component analyses,

Page 27: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 25

Cognos Reporting had the most to gain from potential standardizationinclinations; QlikView users had the highest inclination to buy more seats;Microsoft AS the fewest deterrents to wider deployment; and Bissantz thelowest discontinuance rate.

Overall, combining all four factors, QlikView customers were the most loyal,displacing MicroStrategy which had been the first for the last four years. Theywere closely followed by WebFOCUS and Microsoft AS. Of the products withenough data for analysis, Infor PM OLAP and arcplan customers were theleast loyal.

Some products that had shown low loyalty rates in the past no longer hadenough users remaining for us to analyze their loyalty rates, which suggeststhat this metric is a useful predictor of product failure. Products in thiscategory include the classic Oracle BI products (Discoverer, Express andOLAP Option).

3.11 Platforms Microsoft Windows server is by far the most commonly used platform for BI

servers; its share of 82.3 percent has been on a plateau since the 84.7 percent in2005. There was a continued migration from Windows 2000 (now 9.3 percent) toWindows 2003 (61 percent) and 2008 (2.4 percent), but even Windows NT’sremaining 3.5 percent share was comparable to the shares of individual Unixvariants.

Page 28: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 26

The Unix share (excluding Linux) rose to 11.9 percent from 9.9 percent in

Figure 18 – Server platforms

2.4%

61.0%

9.3%

3.5%

6.2%4.7%

2.3%

4.0%4.7%

0.8% 0.3% 0.4% 0.4%

Win

do

ws

Serv

er2

00

8

Win

do

ws

Serv

er2

00

3

Win

do

ws

20

00

Serv

er

Win

do

ws

NT

Win

do

ws

XP

or

Vis

ta

AIX

HP

-UX

Sola

ris

Lin

ux

Oth

erU

NIX

OS/

39

0(z

/OS)

OS/

40

0

Oth

er

Page 29: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 27

2007. Despite the rise in 2008, the Unix share has declined from 21.7 percentin 2003.

The Linux share of the BI server market grew from 3.2 to 4.7 percent, its bestyet, but still barely half the share of the elderly Windows 2000.

Oracle BIEE (22.5 percent) and WebFOCUS (13.3 percent) had the greatestproportion of sites running on Linux. Reflecting its former IBM relationship,Essbase had the highest proportion (22 percent) running on AIX, whileOracle BIEE had the highest on Solaris (20 percent) and SAP BI/BW on HP-

Figure 19 – Detailed platform trends, 2001-2008(For consistency with earlier years, HP-UX is included in ‘Other Unix’)

0%

10%

20%

30%

40%

50%

60%

70%

80%

Share

ofth

esam

ple

sin

the

eig

ht

Surv

eys

OS/400

OS/390 (z/OS)

Other Unix

Linux

Solaris

AIX

Windows XP/Vista

Windows NT

Windows 2000

Windows Server 2003

Windows Server 2008

Page 30: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 28

UX (22.1 percent).

The share held by IBM proprietary operating systems (OS/400 and z/OS) wasjust 0.7 percent. Only WebFOCUS had a significant proportion — 8.4 percent— of its sites running on IBM proprietary platforms.

Input data volumes have a large impact on platform shares. The Unix(including Linux) share rose from 8.8 percent in sites with less than 200 MBof input data to 47 percent in sites with over a terabyte of input data.

Customer size also affects the platform mix. Here, the Unix (including Linux)share rose from 8.2 percent in organizations with <$100m revenue to 42percent in organizations with more than $15bn revenue.

All the sites using eight of the 25 products analyzed were on Windows.Overall, 48.6 percent of Oracle (including Hyperion), 37.2 percent of SAP(including Business Objects) sites and 14.4 percent of IBM Cognos were onUnix. Only 3.5 percent of the IBM Cognos sites used AIX as a server platform.

Platform choice seems to have very little impact on business success.Windows sites reported slightly fewer business benefits but higher goalachievement rates than Unix sites.

An increasing number of BI applications are now deployed in 64-bit mode,rising from 16.9 percent of sites with less than 200 MB of input data to 40.3percent in sites with over 1 TB. Over 70 percent of the SAP BI/BWdeployments were 64-bit, followed by QlikView and Oracle BIEE with about50 percent. No other product had more than 44 percent running in 64-bitmode.

3.12 Data sources In earlier years, Oracle databases were by far the most popular data source for

OLAP applications, followed by Microsoft. However, the gap had beenclosing, and in 2005, Microsoft squeezed ahead of Oracle. The lead has grownsince then. In 2008, 55.5 percent of sites (up from 52.7 percent in 2007) citedMicrosoft databases as a source, followed by Oracle (46.7, down slightly from47.8 percent in 2007), flat files (24.5 percent), IBM (21.5 percent), manualdata input (14 percent), MySQL (5.8 percent), Teradata (5.2 percent), Sybase(3.1 percent), Netezza (1.5 percent) and PostgreSQL (1.1 percent). These are,of course, based on site numbers and do not reflect the volumes of data,which are likely to be very much higher from, say, Teradata and Netezza than,say, Microsoft databases.

Oracle was the most common data source among all sites with more than500GB of data, but Microsoft was ahead in all smaller bands. IBM databaseswere also more prevalent in the larger sites, but still well behind Microsoftand Oracle in all brackets. Teradata nudged ahead of IBM to third place inmulti-terabyte sites.

Page 31: The BI Survey 8 participant summary - CXP

Complimentary participant summary

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher

All rights reserved. Copyright ©2009

Figure 20 –

0%

10%

20%

30%

40%

50%

60%

70%

<20

0M

B(2

59

)

20

0M

Bto

1G

B(2

20

)

1G

Bto

10

GB

(29

8)

Perc

enta

ge

of

site

slo

adin

gdata

from

each

sourc

e

Complimentary participant summary The BI Survey

permission from the publishers

copied without formal permission from the author and publisher

rights reserved. Copyright ©2009 Business Application Research Center

Data sources by BI input data volumes

(29

8)

10

GB

to1

00

GB

(22

3)

10

0to

50

0G

B(1

50

)

50

0to

10

00

GB

(10

2)

1to

2TB

(95

)

2TB

or

mo

re(9

4)

The BI Survey 8

copied without formal permission from the author and publisher.

Business Application Research Center 29

Oracle

Microsoft

Teradata

IBM

Flat files

Netezza

Manualdata input

MySQL

Sybase

Others

Page 32: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 30

Oracle databases were most often encountered in sites using BusinessObjects,Cognos, Essbase, SAP BI/BW and of course Oracle BI tools.

But Oracle BI tools do not seem to be widely used in Oracle database sites. Infact, Crystal Reports, BusinessObjects, Microsoft AS, MicroStrategy, CognosReporting and SAP BI/BW were the most often used products in Oracledatabase sites. At a vendor level, Oracle BI tools are less widely used in Oracledatabase sites than those from Business Objects, Microsoft, Cognos andMicroStrategy.

Much less surprisingly, Microsoft Analysis Services, Reporting Services andExcel PivotTables are the top three products in Microsoft database sites, nodoubt helped by the fact that the first two are bundled with SQL Server.Crystal Reports is just behind.

Prior to its purchase of Cognos, IBM had no BI tools of its own, so it is notsurprising that Crystal, Microsoft AS and MicroStrategy are the top three BItools in use in IBM database sites. The highest ranked Cognos product is insixth place.

Among sites using open source databases, Microsoft AS, Crystal and QlikVieware the top three BI tools.

Teradata sites were different to the others: MicroStrategy was dominant,followed at some distance by BusinessObjects, Microsoft AS and CrystalReports.

In sites performing manual data entry, Infor PM OLAP, Microsoft AS andSAP BI/BW were the top three BI tools. Surprisingly, TM1 and Essbase weresome way behind, in ninth and tenth places.

3.13 Data volumes The median input data volume was about 6.8GB, which is up from 4.93 GB in

2007, 4.31GB in 2006, 3.8GB in 2005, 4.4GB in 2004 and 5.85 GB in 2003.Though there has been a clear increase since 2005, there is no evidence of‘exploding data volumes’. In fact, median data volumes have fluctuated in anarrow band between 3.8 and 6.8GB, which are surprisingly low figures,since we first started measuring them back in 2002.

Page 33: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 31

Input data volume band Percentage of those answering Analysis band

Less than 10 MB 3.7%<200MB:

18.5%10 to less than 50 MB 6.6%

50 to less than 200 MB 8.1%

200 to less than 500 MB 7.6% 200MB to 1GB:15.2%500 to less than 1000 MB 7.6%

1 to less than 5 GB 14% 1GB to 10GB:20.7%5 to less than 10 GB 6.7%

10 to less than 20 GB 5.2%10GB to

100GB: 15.4%20 to less than 50 GB 5%

50 to less than 100 GB 5.2%

100 to less than 250 GB 5.4%100GB to 1TB:

17.3%250 to less than 500 GB 4.9%

500 to less than 1000 GB 7%

1 to less than 2 TB 6.5%

>1TB: 13%2 to less than 5 TB 3.1%

5 TB or more 3.4%

Figure 21 – Reported input data volumes(Based on responses from 1467 respondents)

As in previous years, MicroStrategy users report much higher data volumesthan most other products, with a median figure of 492GB. But, for the firsttime since the survey began, it did not have the highest median figure, beingjust beaten by Oracle BIEE, with 500 GB. But the Oracle BIEE mean figure of1.19TB is much less than MicroStrategy’s mean volume of 1.56TB, soMicroStrategy’s customers were probably generally accessing more data thanOracle’s.

BusinessObjects with a median of 400GB and SAP BI/BW with 150GB werein distant third and fourth positions. No other product had a median datavolume of even 20GB.

QlikView and Infor PM OLAP, with in-memory databases, had median inputdata volumes of less than a gigabyte.

Applications running in 64-bit mode had a median input volume of 24.2 GB,compared to 4.1GB for those running in 32-bit mode.

3.14 Implementation and rollout On average, 1.8 different implementation resources (in-house, consultants,

outsourcers) were used. By far the most commonly used implementers werein-house IT resources, with almost 69 percent of sites using them to someextent.

The most commonly used external consultants were from software vendors(33.8 percent). Few sites used large general-purpose consulting firms (5.0percent) or outsourcers (or 3.5 percent) for their BI implementations.

Page 34: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 32

Two-thirds of projects were in-house led, much more often by IT people thanby business users. Vendor consultants and specialist BI consulting firms werealmost equally likely to be the primary resource in externally led projects(about 13 percent each).

All resources 2003 2004 2005 2006 2007 2008

In-house IT resources 79.4% 69.0% 65.3% 65.6% 69.7% 68.9%

Vendor consultants 26.9% 32.2% 32.4% 33.8% 35.8% 33.8%

Business end-users 18.4% 27.6% 26.6% 29.0% 26.6% 28.4%

Specialist BI consultants 19.9% 22.5% 24.5% 26.3% 23.3% 24.9%

Independent contractors 13.8% 12.6% 13.2% 11.2% 11.1% 10.1%

Large general-purpose consultants 8.3% 7.0% 5.1% 5.9% 5.1% 5.0%

Outsourcer 3.2% 3.5%

Others 2.1% 0.9% 1.1% 2.1% 1.9% 2.2%

Total 166.7% 171.0% 167.0% 173.9% 176.7% 176.9%

Figure 22 – All the implementation resources used

The mix differed significantly by product. For example, in-house IT peopleled 75 percent of Microsoft Reporting Services projects, but only six percentof MIK projects. Business end-users led 52 percent of MIK projects, but onlyfour percent of Panorama projects. Specialist BI consulting forms led morethan 25 percent of TM1 and Oracle BIEE projects, but less than five percent ofActuate, MIK and WebFOCUS projects. Vendor consultants led more than 25percent of Bissantz, Board, Cubeware, Infor PM OLAP, MIK and Targitprojects, but less than five percent of BusinessObjects, Essbase, MicrosoftSQL Server and Oracle BIEE projects.

SAP BI/BW sites had median consulting fees of $500k, followed by CognosReporting sites with $127k and BusinessObjects with $115k. At the otherextreme, a QlikView customer only spent $8.5k on external consulting fees.

Sites with less than 200MB of input data had median consulting fees of$18.9k, rising to $165.6k in sites with over a terabyte of input data.

There was some correlation between project success and consulting spend.The Business Benefits Index (BBI) peaked for projects with an externalconsulting spend of less than $5k. It fluctuated above that, but the overalltrend is for the BBI to decline in projects with higher implementation costs.This is probably a reflection of the increasing problem rates as projects getlarger.

3.15 Deployment issues and problems 31 percent of sites reported no significant technical, data or people-related

problems with their applications. However, this good fortune will probablychange for some of these trouble-free sites when they have been using their BIapplications for longer: the sites most likely to report no problems are thosethat had been using the software for less than three months. Those with themost complaints had been using it for between three and five years.

As in the previous four years, the single most common problem was slowquery performance, reported by 18.1 percent of respondents. The next two

Page 35: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 33

were poor quality data (16.4 percent) and company politics (13.8 percent).Before 2004, one or other of these had always been the worst problem, butthese latter two problems have become less common, while complaints ofpoor performance have remained ahead.

Other common people-related problems included lack of interest frombusiness users and failure to agree on or changing business requirements.

All

use

rsan

dco

nsu

ltan

ts(2

049)

Users

(18

01)

Bu

sin

es

su

se

rs(4

62)

Tech

nic

al

use

rs(5

80)

Bo

thte

ch

nic

al

an

db

usin

ess

users

(74

2)

Co

nsu

ltan

t(2

48)

Ven

do

rp

erc

ep

tio

ns

(437)

Query performance too slow 18.1% 18.9% 19.9% 21.6% 16.3% 11.7% 11.2%

Poor quality data 16.4% 15.2% 13.2% 14.5% 17.0% 25.0% 33.0%

Company politics 13.8% 13.7% 11.3% 15.5% 13.3% 14.9% 25.6%

Administrative problems (people-related) 13.7% 14.0% 14.5% 14.7% 13.1% 11.7% 16.5%

Could not agree on requirements 12.0% 11.7% 10.0% 11.7% 12.7% 14.1% 19.0%

Lack of interest from business users 12.0% 12.2% 12.3% 12.1% 11.9% 10.5% 13.3%

Unreliable software (bugs, etc.) 10.1% 9.9% 10.2% 11.7% 8.4% 11.3% 2.7%

Requirements changed before project was completed 9.4% 8.4% 6.7% 10.5% 7.8% 16.1% 17.6%

Unable to get data from some systems 7.7% 8.0% 8.0% 8.6% 7.4% 5.2% 10.8%

Missing key product features 7.6% 7.7% 6.7% 8.3% 8.0% 6.9% 3.4%

Product could not handle the data volumes 7.2% 7.4% 5.4% 9.8% 7.0% 5.2% 4.3%

Software too hard to use 6.4% 6.7% 7.6% 7.1% 5.7% 4.8% 6.9%

Security limitations in the product 3.8% 3.9% 3.5% 4.7% 3.8% 2.8% 2.3%

Product could not handle large numbers of users 2.3% 2.2% 1.9% 3.1% 1.8% 3.2% 1.4%

Other 7.4% 7.3% 6.9% 8.3% 6.9% 7.7% 4.3%

Total problems 1.48 1.47 1.38 1.62 1.41 1.51 1.72

No significant problems 31.0% 31.1% 32.7% 26.0% 34.4% 29.8% 26.1%

Figure 23 – Reported problem rates by respondent type

Cognos Reporting users had the most product complaints, including queryperformance and unreliability. Microsoft Excel PivotTables was almost asbad, with data and user scalability being the biggest problem areas. SAPBI/BW was just behind, with the biggest problem areas being queryperformance, unreliability and being too hard to use.

By suite, Oracle classic customers reported the most product problems andMicrosoft SQL Server the fewest.

Bissantz and QlikView sites reported the fewest technical problems.

Bissantz, Panorama and WebFOCUS sites were the most satisfied with thequality of product support they received, while BusinessObjects and Essbasesites were the least satisfied by far. In general, customers of large vendorsreported much worse support quality than those using small and mediumsized BI vendors.

Page 36: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 34

Software cost and implementation/support were the biggest deterrents towider deployment. Slow query performance was the third highest.

By product, SAP BI/BW sites saw the most deterrents to wider deployment,particularly poor performance, high implementation/support costs andadministrative and maintenance complexity. Panorama, MIK and MicrosoftAS sites reported the fewest deterrents to wider deployment.

3.16 Performance issues Not only was query performance the most severe product-related problem, as

in all previous Surveys, but it was again the single most common problem ofall. Furthermore, sites that used query performance as a selection criterionwere more successful in business terms than those that did not.

Query performance is closely linked to data volumes. In previous editions ofThe BI/OLAP Survey, median query times have risen or fallen in closeassociation with that year’s median input data volumes, though that trendwas broken this year, when a small rise in data volumes was accompanied bya small improvement in query times. But there is still no evidence that thehuge improvements in hardware performance have improved queryperformance very much.

Query times inseconds

1stquartile

Median3rd

quartile

All (1976) 3.26 7.53 28.79 7.5

Input data volumes

Less than 10MB (53) 2.17 5.19 12.14 5.2

10 to 50MB (94) 2.75 5.25 16.00 5.3

50 to 200MB (115) 2.94 5.94 19.25 5.9

200 to 500MB (111) 2.79 5.98 26.82 6

500 to 1000MB (110) 3.35 6.90 22.67 6.9

1 to 5GB (199) 2.27 4.64 16.59 4.6

5 to 10GB (94) 3.20 7.27 25.33 7.3

10 to 20GB (73) 2.97 6.03 17.31 6

20 to 50GB (70) 3.92 8.67 36.82 8.7

50 to 100GB (72) 3.36 9.50 35.00 9.5

100 to 250GB (77) 3.77 8.68 33.21 8.7

250 to 500GB (70) 4.07 9.12 48.33 9.1

500 to 1000GB (101) 3.47 8.83 29.77 8.8

1 to 2TB (90) 4.57 14.55 53.57 14.5

2 to 5TB (43) 6.72 25.71 152.73 25.7

5TB or more (44) 7.14 30.00 300.00 30

Figure 24 – Quartile analysis of query times by inputdata volume

There was a sharp decline in business benefits and goal achievements asquery times lengthened.

Page 37: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 35

QlikView users (with very small data volumes) had the fewest performancecomplaints, and SAP BI/BW and Cognos Reporting customers, with largedata volumes, the most.

Overall, the median reported query time was 7.53 seconds, varying from 3.08seconds for QlikView to 40.9 seconds for MicroStrategy. Although OracleBIEE customers reported similar data volumes to MicroStrategy sites, it had amuch faster median query time of 5.6 seconds.

Applications with less than 100MB of input data had a median query time of5.2 seconds, rising to 30 seconds in sites with 5TB or more.

Adjusting for varying data volumes, Essbase, QlikView and TM1 had thefastest query times. MicroStrategy and Cognos Reporting have slow querytimes, even allowing for the larger data volumes that they handle.

Probably because of the different product mix, Windows sites reported muchfaster query performance than Unix sites, even after adjusting for the latter’slarger data volumes.

Several vendors are promoting real-time BI, meaning that the time betweennew data becoming available from source systems and its availability for BIqueries is minimized. But the benefits analysis in The BI Survey 8 shows thatthis reduced latency only has a limited effect on the Business Benefits Index(BBI). The BBI and goals do decline as latency increases, but the curvesfluctuate and there is no inflexion point.

Overall, the median latency is 47.6 minutes, which can hardly be described asreal-time, though it is a little better than the 51 minutes reported in 2007.

The products that do not use pre-built aggregates (TM1 and Infor PM OLAP)had, not surprisingly, by far the fastest data refresh times (ie, lowest datalatency). Taking into account data volumes, Crystal, Cubeware andWebFOCUS also performed very well, while Cognos Analysis is about 2½times slower than expected for the data volumes it handles. Cognos Reportingis almost twice as slow as its data volumes justify.

3.17 Web BI We have consistently found that respondents expected a faster move to Web

deployment than has actually occurred. There was an increase in Webdeployments in 2008 compared to previous years, but it remains lower thanpredicted by participants in the previous edition of the Survey. It seems thatthe growth in Web usage continues to be slower than most people expect,though the gap narrowed in 2008.

North American sites had significantly higher Web deployments thanaverage, while German and Italian sites were the least likely to be Web-deployed.

Web deployment was much higher in larger organizations (particularly thosewith more than $15bn revenues), but very low in organizations with fewerthan 500 staff.

The highest rate of Web deployment was in arcplan sites, with a median rateof 99.3 percent. Actuate and WebFOCUS were just behind. Board, Cubewareand Bissantz sites had almost no Web usage.

Page 38: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 36

Among multi-product vendors, Cognos and Oracle BI sites had the highestrates of Web deployment. Hyperion, Microsoft and Business Objects weresignificantly lower.

Organizations that had performed a formal, multi-product evaluation weremuch more likely to have deployed their BI applications via the Web thanthose that only considered one product.

The most popular Web technologies were .NET (33 percent) and J2EE (20percent), followed by Web Services (14.5 percent) and AJAX (13 percent). Butmore respondents expressed no strong preference than preferred any singletechnology apart from .NET.

Firefox penetration increased to 22.5 percent, though Microsoft InternetExplorer remained dominant with 97.3 percent. Firefox was used in morethan 35 percent of Actuate, Cubeware and Essbase sites, but even in sitesusing Microsoft Reporting Services, 15.4 percent used Firefox. Fewer than tenpercent of TM1 sites said any of their users accessed the product via Firefox.

The highest rates of business success were reported by sites with between 40and 50 percent of seats Web-deployed. Sites with no or 100 percent Webdeployment had the lowest level of business benefit achievement.

Extranet deployment rates remain stalled at 16.4 percent (down from 17.1percent in 2007) compared to the 28.3 percent predicted by last year’ssample. Respondents invariably predict a big rise – this year to 38 percentwithin two years – which seems never to materialize.

Actuate (56 percent) and WebFOCUS (46 percent) sites were the most likelyto have an extranet deployed, with Bissantz, Cubeware, Infor PM OLAP andTM1 the least likely. Among multi-product vendors, Oracle BI sites were themost likely to have extranets deployed.

US organizations were more than four times as likely to have a BI extranetthan Italian organizations.

Customers and partners were the most popular extranet target users, withvery few giving access to the general public. Actuate and Microsoft RS siteswere the most likely to provide access to customers and Crystal,BusinessObjects and Panorama sites to partners.

Page 39: The BI Survey 8 participant summary - CXP

Complimentary participant summary The BI Survey 8

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher.

All rights reserved. Copyright ©2009 Business Application Research Center 37

Figure 25 – Extranet plans

16.4%

4.4%

6.3%

10.9%11.7%

50.3%

Already provideInternet access

to BI data tooutsiders

Within the next6 months

Within the next6 to 12 months

Within the next1 to 2 years

In more than 2years time

Will neverprovide

outsiders withInternet access

to BI data

Page 40: The BI Survey 8 participant summary - CXP

Complimentary participant summary

Not to be quoted without permission from the publishers

Not to be quoted or copied without formal permission from the author and publisher

All rights reserved. Copyright ©2009

4 Next stepsFor more information or to order onwww.BI-Survey.com

Or if you have any queries

Michael SchumacherTel: NoRest of world:mschumacher@bi

Complimentary participant summary The BI Survey

permission from the publishers

copied without formal permission from the author and publisher

rights reserved. Copyright ©2009 Business Application Research Center

For more information or to order on-line please visit:Survey.com

Or if you have any queries, please contact:

Michael SchumacherTel: North America: 1-866-274-9720Rest of world: +44 (0) 20 8879 [email protected]

The BI Survey 8

copied without formal permission from the author and publisher.

Business Application Research Center 38


Recommended