Post on 06-Jun-2020
transcript
Posters used to promote the Amplify IVR prototype are hung in Gurgaon.
WHAT IS IVR?
Integrated Voice Response (IVR) is a flexible
phone-based technology that can be deployed
in communities with low literacy and limited
or no Internet usage as a method for creating
alternative social networks and disseminating
information. Facilitators of these systems (usually
local nonprofit organizations) post information
to a central server and allow the public to call
into a local number to hear announcements,
ask questions, and participate in asynchronous
(not live) discussions. All information is recorded
and saved into the IVR system with the aid
of a live moderator, who monitors incoming
calls for quality control. IVR systems are low-
cost and can be rapidly deployed using simple
technology accessible to many low-income
communities around the world.
BENEFITS TO AMPLIFY
Amplify seeks to redefine international aid by
directly including low-income communities
around the world—traditionally recipients of
aid—in the process of designing solutions to
a range of global issues. Amplify engages its
audience through the website OpenIDEO.com,
which has a user base of over 50,000 people.
This website, however, relies on access to high-
speed Internet as a prerequisite for effective
participation. As many communities in our
target population do not have regular Internet
access, they have no route to participation on
OpenIDEO. IVR offers a means of expanding
the Amplify audience base to anyone with a
basic feature phone. The system can also be
programmed through a “missed call” system to
ensure that all calls into the system are free of
cost for users.
IVR has the potential to radically lower the
barriers to participation for low-income
communities with lived experience on the
issues Amplify is tackling, and to create a
diverse base of users across countries that
OpenIDEO has so far been unable to reach in
large numbers.
AMPLIFY: CHALLENGE ONE
WWW.IDEO.ORG/AMPLIFY 2
EXTENDING THE REACH OF AMPLIFY
In April 2014, the Amplify program explored the
use of Integrated Voice Response (IVR) as a means
of engaging local communities in the collaborative
innovation process. The experiment included
explorations of various IVR technologies and two live
prototypes in India.
AMPLIFY CHALLENGE ONE: IVR Prototyping
Woman in Gurgaon shares her thoughts on the GKA radio programming.
HOW THE IVR SYSTEM WORKS
IVR systems function much like customer
service phone trees. With a missed-call system
in place, a caller calls a local phone number,
and the server hangs up on the caller after one
ring (thereby avoiding any cost to the caller).
The telephone server then returns the call and
plays a pre-recorded introduction to the IVR
system. The user is then given the choice of
several menu options. Basic options include a
way to listen to entries other users have posted
and a way to record and post a new entry
to the system. Each entry is monitored by a
moderator who uploads all usable entries to the
system for callers to hear (while also eliminating
entries which are unintelligible). Each published
entry is then added to the queue of recordings
that users hear when calling in. Callers are
instructed to press a certain number to skip
between entries, and another number to leave
a comment on something they hear. On more
sophisticated systems, callers can receive SMS
notifications if someone comments on their
entry, and those comments are the first things
a caller hears the next time they call in to the
system.
PROMOTION AND ENGAGEMENT
IVR systems are paired with a variety of
promotional tools that attract potential users,
including:
Radio: By airing promotional ads,
dramatizations, call-in programs, or other
relevant content, radio can provide context
and funnel users to an IVR system. Community
radio stations are particularly well-suited for
this type of partnership, as they have a more
restricted geographic focus and are more
visible in the community (due to local outreach
programs) than commercial stations. Air time
is also vastly less expensive on community
stations and their programming tends to be
more flexible.
Mobile Outreach: SMS blasts or direct mobile
phone outreach can be very effective, and has
the benefit of using the same technology from
a user’s perspective as the IVR system itself (a
mobile phone).
Analog Communications: Community
meetings, promotional posters, and murals are
common methods of exchanging information
across the developing world.
AMPLIFY: CHALLENGE ONE
WWW.IDEO.ORG/AMPLIFY 3
EXTENDING THE REACH OF AMPLIFY
A female driver is interviewed on a radio show promoting the Gurgaon campaign.
PROTOTYPE ONE: GURGAON
Primary Partner: Gurgaon ki Avaaz (GKA) GKA
is a community radio station serving the peri-
urban population of Gurgaon—roughly 2 mil-
lion people. They have an active listener base of
low-income residents, with programming that
includes traditional music, comedy, and call-in
shows on topics relevant to the community
(e.g. health).
Tech Implementer: GramvaaniGramvaani is a technology company special-
izing in IVR and mobile phone-based applica-
tions, with a strong presence throughout India.
Gramvaani has also built IVR systems in other
South Asian and African countries.
“IT IS IMPORTANT TO CHANGE THE COLOR OF ‘WOMENS’S ONLY’ TEMPOS [SHARED VEHICLES]...THIS WAY IT BECOMES EASIER TO DIFFERENTIATE EVEN IN THE DARK.”
– NIRANJAN Gurgaon | April 23
Campaign Promotion:• Four 1-minute commercials broadcast
throughout the campaign on GKA, each depict-
ing a common safety issue and encouraging
listeners to call the IVR system
• Three 45-minute live shows on GKA (with
three rebroadcasts), each featuring live guests
and inviting listeners to share their views on-air.
• 19 posters distributed throughout one of Gur-
gaon’s informal slums
Results: • 45 calls, of which 17 were of usable quality
• Overall call quality: Moderate
• Callers were very passionate about the topic,
but gave little detail in responses and seemed
confused by the task of giving feedback.
• Callers were nearly all male
AMPLIFY: CHALLENGE ONE
WWW.IDEO.ORG/AMPLIFY 4
LIVE PROTOTYPING
LIVE PROTOTYPING
We ran two live IVR prototypes, testing various
methods of attracting users to the IVR system. The
two systems were nearly identical, each playing a
series of 2-minute summaries of ideas submitted to
the OpenIDEO platform and prompting callers to leave
comments or share their own ideas.
Women’s group in Gurgaon discusses the challenges of accessing the IVR
“IN MY OPINION, THE NUMBER OF FEMALE POLICE OFFICERS SHOULD BE INCREASED IN ALL STATIONS.”
– NITIN Delhi | May 2
PROTOTYPE TWO: DELHI
Primary Partner: Breakthrough A nonprofit media organization based in Delhi,
Breakthrough produces multimedia campaigns
to raise awareness about violence against
women around the world.
Tech Implementer: GramvaaniSee description for Prototype #1.
Campaign Promotion (week one):• A 30-second radio jingle aired on BigFM,
one of Delhi’s most popular commercial radio
staions (especially with commuters).
• Social media posts by Breakthrough
Results: • 15 calls, none of which were usable quality
• Commercial radio proved to be a highly
ineffective tool for promoting IVR. Drivers are
not typically able to make a call in the moment
or capture the number, and the incentive to
participate is lowered without the community
tie-in.
Campaign Promotion (week two):• A series of SMS messages sent to 8,000
people in the Delhi area on Breakthrough’s
contact list, via SMS GupShup, an India-based
mobile technology company.
Results: • 179 calls, of which 11 were of usable quality
• Overall call quality: High
• Callers were more inclined to provide their
own ideas than to leave feedback on others’
posts.
AMPLIFY: CHALLENGE ONE
WWW.IDEO.ORG/AMPLIFY 5
LIVE PROTOTYPING
SETTING UP THE SYSTEM
• IVR technology is easy and inexpensive to
set up quickly, but requires a partner who can
establish a local phone line to manage the
system—ideally an NGO based in the target
country.
• It is critical to have a team on the ground to
work directly with local partners in preparation
for launch because systems are difficult to
test remotely; a local presence is crucial
for galvanizing those partners to action and
troubleshooting.
• 3-4 days on the ground for pre-launch work
is insufficient to handle all of the necessary
preparation. A block of 8-10 pre-launch days
would be more appropriate.
• While working with local partners who share
a passion and mission for the specific IVR
subject matter is important, it is equally crucial
to find partners who have the bandwidth and
enthusiasm to take on the project. Subject-
matter expertise is less important than a reliable
technical and/or promotional partner.
• Lack of a direct connection with
implementing organizations like radio stations
and tech companies can create bottlenecks
for changes to the system. Partnering directly
with a tech implementer and promotional
partner (like a community radio station) is most
efficient.
• Slow progression on contracts is a risk to
the project as it can take time to negotiate
contracts with international IVR partners.
It’s best to start this process early and with
sufficient lead time that it doesn’t interfere with
the prototype.
MANAGING THE SYSTEM
• IVR providers can give real-time data on
participation and usage, making it fairly easy
to monitor techical successes and make
changes in the early days of a campaign. Once
established, IVR systems can be monitored
remotely with relative ease.
• Using a live moderator is crucial to success
of an IVR system linked back to OpenIDEO for
both quality control and effective transcription/
translation.
• High call volume for future prototypes will
limit the amount of content that gets heard; it
may be useful to categorize the comments to
personalize the experience for each user.
AMPLIFY: CHALLENGE ONE
WWW.IDEO.ORG/AMPLIFY 6
INSIGHTS
INSIGHTS FROM LIVE PROTOTYPES
Based on the experience and results of our two
prototypes, we gained the following insights into
the use of IVR as it relates to the Amplify program.
• Gramvaani currently has no way to automate
the IVR interface so that pre-recorded ideas
or comments are the first thing callers hear;
callers instead hear all comments in the order
they were posted unless a moderator manually
moves the pre-recordings to the top of the
queue. This limitation makes it difficult to use
IVR to get feedback on specific pre-recorded
items, since they will quickly fall to the bottom
of the queue and will not be heard.
• Mobile networks in target countries often
have low call fidelity and frequent loss of
service mid-call. We should explore system
capabilities to do multiple call-backs if the
network is not working.
ENGAGING USERS
• Callers into the system are engaged and eager
to talk, and want to find a way to help. The IVR
system can be valuable not just for capturing
content initially through recordings, but also
for identifying engaged community members.
This finding echoes previous learnings from the
IDEO.org #crapmap project in Ghana.
• Callers are used to the format of IVR as a way
to lend their voices to a social network; the
specific subject of the IVR campaign is often of
secondary importance. A single IVR line could
be used across several Amplify challenges
to build repeat participation, or linked to
another organization’s existing IVR system (e.g.
Jharkand Mobile Vaani).
• Live radio shows that dig into relevant content
are more successful at engaging users and
driving call-in behavior than promotional ads.
• Motivation for people to use the IVR system
to critique others’ ideas isn’t very strong—this
kind of feedback takes coaching and in-depth
conversation. An easier value proposition would
be asking callers to share personal stories
related to the topic (e.g. during the Research
phase of a challenge).
• Callers have little incentive to remain engaged
with the IVR system over time without a clear
two-way feedback loop. The ultimate goal of
the IVR system should be to create a true social
network, with the ability for callers to share
their thoughts with a broader community and
receive responses.
• Listening to the radio and calling into an IVR
system is a leisure activity. IVR prompts should
feel less like a survey or chore, and more like
an active social community you’d want to visit
repeatedly.
• Callers are unlikely to listen to lengthy
90-second recordings. All future content
should be in clips of 30 seconds or less.
REACH & AUDIENCE
• Women in many low-income areas of India
are unable to call in because they have no
credit on their mobile phones. We should
test the effect of offering small credits to
prospective female callers.
• Although we initially believed commercial
radio would provide visibility for the IVR system
to a much larger population, most of those
listeners are drivers who can’t immediately
capture a phone number or make calls.
Community radio, which has smaller but
more engaged audiences, is more effective in
spreading the word.
• Online social networks (Facebook and Twitter)
are ineffective in garnering participation, as
many of those users are already have access to
OpenIDEO directly.
Migrant families, especially women, generally
have no home radio; men often listen to the
radio while they work as taxi or rickshaw
drivers.
AMPLIFY: CHALLENGE ONE
WWW.IDEO.ORG/AMPLIFY 7
INSIGHTS
CHALLENGE TWO: ROADMAP
PROPOSED STEPS
• Daily SMS digest of the system, which
highlights activity on their stories
• Moderator to review calls
• A custom SMS when a comment is left
on that user’s content
Other levers to test:
• Provide small amounts of mobile credit
to female target users as a way to
ensure accessibility of the system
• Try hiring a small subset of local
“super users” who encourage community
members to participate in the system
• Experiment with small-scale translation
and transcription program with a group
of active OpenIDEO users
• Test simple methods of getting IVR
responses onto OpenIDEO (text-only
posts of transcriptions to start)
Create a marketing plan with promotional
partner, with a mix of spoken and written
media (live radio shows, 30-second
recorded promos, and simple posters).
Co-create campaign content.
Work with partners to ensure local
relevance.
Arrive in country 8-10 days prior to launch
to test all aspects of the campaign on the
ground with partners.
Record any remaining prompts and identify
a live moderator to assist with quality
control, translation, and transcription.
Monitor placement of written promotions
around relevant community public spaces.
1.
2.
3.
4.
5.
6.
7.
8.
Find a local technology partner.
Begin contract work early to ensure
timeliness. See Appendix 1 for a list of
potential technology partners in Kenya, India
and Tanzania.
Recruit community radio station with an
active listener base as promotional partner.
See Appendix 1.
Work with tech partner to identify an
IVR system that maximizes two-way
communication.
If necessary, find a secondary partner
to build in SMS functionality for direct
feedback.
Our goal is to create a social network that
incentivizes repeat use. An ideal system
would utilize an existing phone number
with an active community. In lieue of
this, a system should use return calls and
targeted SMS blasts to encourage repeat
engagement.
Features of the ideal ground-up IVR system:
• Simple IVR call-in number
• Missed call system with double call back
to assist with bad networks
• Menu with the following choices:
1. Share your story
2. Listen to other peoples’ stories
3. Comment on someone’s story
4. “Like” someone else’s story
• Simple beeps before and after recording
stories, and ability to review one’s
idea before posting
• SMS follow-up to thank caller
AMPLIFY: CHALLENGE ONE
WWW.IDEO.ORG/AMPLIFY 8