How can Cambridge University's Centre of Governance and Human Rights ensure the sustainability of its innovative public opinion analytics platform, that leverages the ubiquity of mobile and radio in Africa to reach the least heard voices?
2. Africa's Voices started as an applied research project
from Cambridge University...
2
• Africa’s Voices leverages radio’s ubiquity, with over 90% of Africans listening
regularly in a variety of languages, and prodigious growth in mobile telephony.
• Stations ask questions and prompt debate on air, and Africa’s Voices gathers
listener SMS responses and audience demographics.
• This data gives direct and immediate access to a social discussion space in
which shared meanings and ideas – and their change over time - can be
identified.
• Africa’s Voices uses proprietary analytical methods to derive insight from the
data and provide an easy to access view of the responses.
• Africa’s Voices began at Cambridge University as an applied research project
that explored optimal modalities for enabling and interpreting hard-to-reach
voices in Africa in development and governance discussions.
• Challenging conventional survey-based approaches, researchers have
developed new tools for interpreting public opinion as shared and contested
ideas expressed in local conversational language.
• Over twoyears, with pilots in radio stations in eight countries, the project’s
findings give confidence of the enterprise’s basic architecture and relevance.
Background on Africa's Voices
Challenges Recommendations Analysis Roadmap Appendix
3. Key Challenges
• Can Africa’s Voices become a viable self-
sustaining social enterprise?
• What would it need to do and become to
achieve it?
…but its future depends on establishing itself on more
independent foundations in a challenging environment...
3
Kenya is primed for economic growth which could change viable use cases in the
mid-term
• Pan-African GDP has increased at 4.9% annually between 2000 and 2008
• Between 2005 and 2012 Kenyan GDP almost doubled from 18.739 to 34.059
billion USD
Technology penetration in Kenya favours Africa’s Voices’ platform at this point in
time:
• Mobile phone usage has increased from 13.8% to 71.17% between 2005 and
2012
• 78.1% of rural and 83.1% of urban Kenyans owned a radio (2013)
• 93% of Kenyans surveyed reported listening to radio in the past week (2011)
The growing availability of cheap 3G technology is a factor that Africa’s Voices
needs to be mindful of for future
• Internet usage in Kenya increased from 3.1% in 2005 to 32.1% in 2012
Context
Challenges Recommendations Analysis Roadmap Appendix
Progress to date
• Project initiated mid-2012
• 9 Radio Stations engaged in applied pilot
across Africa
• High level business case created in November
2013
• Value chain investigation undertaken
• Discussions with IBM underway as potential
partner or analytics provider
4. …however, there is real opportunity in the radio enabled
discussion space for the right enterprise
Vision
To be the primary platform that is used to engage, facilitate
and understand voices within civil society throughout Sub-
Saharan Africa.
4
1. Develop the platform with delivery of information as aid, scoping, M&E of aid and service
delivery as its initial primary purpose
2. Subsidise the non-market* research and community discussion space using the M&E and
info as aid market
3. After reaching a viable size, the market research space becomes viable
4. An upfront investment of $80,000 will be required to build and pilot the platform
5. Ongoing expected monthly costs of between $13,000 and $17,000 should be able to be
covered without external funding
Recommendations and Conclusions
Challenges Recommendations Analysis Roadmap Appendix
Mission
To bring listeners, radio stations and 3rd parties together
through useable technology, mutually beneficial
relationships and innovative insight, using advanced analytic
capabilities developed in Cambridge to interpret data for
both practical and pure research purposes.
Values
• To encourage, facilitate and amplify a breadth of voices
in public discussion
• To support meaningful social change through
information
• To analyse real social discussions to deliver public
opinion insight
• To innovate through technology and robust analytics
• To deliver knowledge products faster than comparable
approaches
• To differentiate from conventional polling enterprises
Platform
AV
Radio
Buyers
Current
* Market research is defined as corporate-led marketing mix research for profit
5. AV could be used to deliver aid into hard to reach areas
and measure its effectiveness…
5
Challenges Recommendations Analysis Roadmap Appendix
Potential for AV as an aid delivery platform Disaster Relief
Potential Collaborator Analysis
Delivery of information as aid via sponsored shows
is not a new idea. However, the AV platform could
facilitate it through:
• A user interface that allows specific radio
stations to be identified and targeted quickly
• A detailed background on the socio
demographic information for each station
• The speed which programmes can be set up
• Cross location comparisons
• Messaging that can be changed over time
to fit feedback or demographic
• Providing a platform for sharing of
community knowledge about the issue and
for public discussion thereon
• Detailed reporting that will be available for
key stakeholders from the platform
Legend
1. Geospatial Location
2. SMS data collection
3. Interactive Voice System
4. Incentivized Responses
5. Information outgoing
6. Cloud-Based
7. Demographics Collected
8. Social Media Presence
1
2
3
4
5
6
7
8
Farm Radio International
1
2
3
4
5
6
7
8
Text to Change
Farm Radio International
is a platform that
disseminates farming
information via radio
partners throughout
Africa in addition to
providing a platform for
sharing community
farming information.
Text to Change is a
mobile phone platform
that facilitates the sending
and receipt of information
via telephony between
respondents and partners
of Text to Change like
UNICEF.
Following natural disasters such as
the hurricane that hit the
Philippines, radio is often the only
communication mechanism still
running. Rapid response using
interactive radio has been
demonstrated to be highly effective
in this environment.
6. AV
…it can also be used to measure and evaluate wider aid
delivery…
6
High Level Financials
Challenges Recommendations Analysis Roadmap Appendix
Example M&E Partnership Process
Define Initial
Scope
Formative
Research
Aid Delivery
Evaluation
Corroboration
Evaluation
Measure
Effectiveness
1. Aid Spend in Kenya: $2.5bn
2. M&E % of Project: 5%
3. Calculated M&E Spend in
Kenya: $125m
4. Total running cost of AV in
first year in Kenya after
startup: $84k
5. % of total M&E spend to fund
AV: 0.06%
Potential for AV platform within M&E Benchmarking
The potential in the scoping and M&E space for
the AV platform is significant:
• NGOs and aid agencies will have broad near
instant access to ask questions
• Trends, benchmarking and analytics will
facilitate decision making
• Quantitative and qualitative data will allow for a
richer feedback mechanism
• Multiple radio stations can be engaged at the
same time with the same question
• Effectiveness can be demonstrated through
the analytics and trend analysis
• Performance Reporting will be available for
key stakeholders from the platform
demonstrating engagement
Aid
Agency
7. Legend
1. Geospatial Location
2. No-Cost SMS
3. Interactive Voice System
4. Incentivized Responses
5. SMS information outgoing
6. Cloud-Based
7. Demographics Collected
8. Social Media Presence
…this is an area that aid agencies and service delivery
firms have found challenging…
7
Challenges Recommendations Analysis Roadmap Appendix
Potential Collaborator Analysis
Result, effectiveness or impact tracking is
something that NGOs struggle with:
• Often M&E budgets are squeezed too
tight
• M&E is often considered as an
afterthought
• Little continuous improvement using
M&E
NGOs Bilateral Agencies Multilateral agencies
Multilateral agencies, such as the World
Bank and UNESCO, assign KPIs to their
projects:
• Measuring these KPIs has been
difficult for the program teams
• P4R has highlighted the debate on
quantifiable feedback metrics
• Metrics are driving spend
Bilateral aid agencies find M&E
challenging and engage 3rd parties to
undertake the evaluation and analysis of
projects:
• Can be anywhere from 10 to 20 days
of time
• Used to demonstrate external
evaluation to governments
1
2
3
4
5
6
7
8
Text to Change
Text to Change is a
mobile phone platform
that facilitates the sending
and receipt of information
via telephony between
respondents and partners
of Text to Change like
UNICEF.
GeoPoll is a mobile
platform that allows
customers to survey their
110 million respondents
via SMS, the web, or
voice.
1
2
3
4
5
6
7
8
GeoPoll
8. …and there are specific examples where AV would have
been beneficial in this space
8
Challenges Recommendations Analysis Roadmap Appendix
UNESCO Challenges InfoAsAidChallenges
• The reliability of data obtained through traditional surveys
• The ability to measure intangible benefits
• The institutionalised methodology that doesn't directly
engage the bottom of the pyramid
• Research fatigue on the part of the responders
• The breadth of surveys over sub-Saharan Africa as a
whole
• Reliable and robust methodology that underpins the
analysis
• The ability to measure intangible benefits of aid delivery
• Engage recipients at the bottom of the pyramid in a free
and open space
• Needs to be applicable to either a single small scale site or
on a pan-country scale as required
• Engagement to make a difference to overcome the
research fatigue
• Speed of implementation to be able to react to demands
• Implementation difficulties during both World Vision Kenya
and Save the Children projects
• Technology issues with FrontlineSMS not already
recording the SMS
• Measuring the benefits of the project was difficult due to
the lack of analysable data
• Training of staff on the ground was very time consuming
• Very tight timescales
• Rapid scoping of environments on the ground
• The limited data that was collected was not analysed and
used in either a monitoring or evaluation context
• Ability to quickly roll out the technology
• Easy to use and understand with limited ongoing training
• Metrics available without the user being present
• Measurement of results based on reliable and robust
methodology
InfoAsAidRequirements
UNESCO Requirements
Miranda Eeles, Project
Manager - InfoAsAidJaco Du Toit, Advisor for communications and
infortmation - UNESCO
9. Building on this foundation, other areas of interest can be
targeted and insight gained from the platform…
9
High Level Financials
Challenges Recommendations Analysis Roadmap Appendix
Example ResearchPartnership Process
Potential for AV as a Research and Engagement platform Benchmarking
From a community engagement
standpoint, the platform delivers the following
benefits:
• Open but optionally anonymous discussion
space
• Allows facilitators information on what
other communities are doing
• Socio demographic information will allow
for targeted engagement
As annon-market research platform, AV will give
several benefits over standard methods:
• The feedback and analytics should be near
instant
• Using radio and SMS, data can be collected
from hard to reach places using the web
• The ability to rapidly iterate hypotheses should
be significantly increased
Theory
Research
design
Hypothesis
Measure
and collect
data
Data
Analysis
Findings
AV
1. Estimated survey spend per
question per response: $1 -
$3
2. Estimated travel cost: 25%-
40%
3. Travel savings by using AV:
$0.25 – $1.2
4. Number of question or
responses required to fund
AV: 9,300 a month
10. 1
2
3
4
5
6
7
8
Jana
…with research and community engagement being
important for a platform with social impact
10
Challenges Recommendations Analysis Roadmap Appendix
Potential Collaborator Analysis
Research Community Engagement
Legend
1. Geospatial Location
2. SMS data collection
3. Interactive Voice System
4. Incentivized Responses
5. Information outgoing
6. Cloud-Based
7. Demographics Collected
8. Social Media Presence
1
2
3
4
5
6
7
8
Trac FM
• Kenya has been a popular location for development studies in
Africa. There is fierce competition in the research space, yet
getting reliable data is challenging
• Non-market research actors still employ traditional
methodologies. They are time consuming and not always cost
effective. However there is a growing trend towards innovative
technology use in this space
• Real time data gathering and analysis are difficult to achieve
• Citizen engagement is important for empowering policy
discussions in the region. However, it has been challenging to
get citizens heard or to participate in public discussions
• Unlike the other three areas there will be no buyer for this use
case. The community would stimulate the discussion as
opposed to third parties. A long code would be used
• Open community discourse is a difficult topic and needs to be
handled with care
Jana is a mobile web-
based application that
provides customers with a
database of respondents
willing to answer surveys
in exchange for airtime
credits
Trac FM is a platform that
uses radio broadcasts to
promote public
engagement while also
capturing opinions from
participants via SMS
11. Once significant scale has been achieved, the market
research space can be targeted…
11
High Level Financials
Challenges Recommendations Analysis Roadmap Appendix
Example Market Research Partnership Process
Potential for AV as a Market Research platform Benchmarking
Market research has historically focused on the
creation of representative panels to garner views
and opinions. However, AV could be used to
potentially add value in this space through:
• Allowing self-selecting panels to voice their
opinions and demonstrate engagement
• Being able to understand and interpret the
rationale behind views
• Cross location comparisons
• Demographics and geographies that can be
targeted through identifying stations
• A wide reach to different areas that are not
currently serviced by market research firms
• An ability to generate a sentiment index in a
similar way to that of social media
1. Assuming $0.5 per response
2. Total running cost of AV in
Kenya after startup: $84k
3. Number of responses per
day required to fund AV: 600
4. To get to this
potential, significant critical
mass would need to be
reached
Define
objective &
problem
Determine
research
design
Design
Research
Instrument
Sampling &
Data
Collection
Analyse
Data
Visualise
results
AV
12. 1
2
3
4
5
6
7
8
Jana
…care should be taken however to ensure the platform is
still considered to be for social impact and change
12
Challenges Recommendations Analysis Roadmap Appendix
Potential Collaborator Analysis
Catering to MNCs could conflict with the
core values of AV:
• MNCs are primarily interested in
profits and not with social impact
• If listeners perceive conflict of
interest there could be listener
disengagement
Corporate Responsibility Focus Ethical Questions
Market research is only one part of AV’s
core activities:
• AV needs to continue to both
passively monitor AND facilitate
public discourse
• How do you incentivize people if
purely market research?
Thorny ethical challenges are likely to
arise:
• Should information collected to
facilitate social change be monetized?
• What can legitimately be done with
this information?
• An expectation of privacy?
Legend
1. Geospatial Location
2. SMS data collection
3. Interactive Voice System
4. Incentivized Responses
5. Information outgoing
6. Cloud-Based
7. Demographics Collected
8. Social Media Presence
GeoPoll is a mobile
platform that allows
their customers to
survey respondents
via SMS, the web, or
voice.
Jana is a mobile web-
based application that
provides customers
with a database of
respondents willing to
answer surveys in
exchange for airtime
credits
1
2
3
4
5
6
7
8
GeoPoll
13. The success or failure of Africa's Voices will be in part
determined by two key drivers…
13
Challenges Recommendations Analysis Roadmap Appendix
Critical Success Drivers and Barriers
In this quadrant, the uptake by
radio stations will not be met by
sufficient demand from the
buyers. This will cause
significant funding issues.
In this rapid growth
quadrant, significant investment
in people will be required to
keep pace with demand. Being
able to react quickly to
demands by both radio stations
and buyers will be key.
Although uptake in this
quadrant is slow, it is even
between stations and buyers.
This will result in slow but
sustainable growth. Sufficient
uptake will be required to break
even however.
In this quadrant, the uptake by
radio stations will be slow. This
will mean that buyers do not
have the breadth of reach
required and so will not use the
platform to its full potential
Potential Barriers to success:
• Legislation covering data residency
• Branding
• Communication of USP
• Vernacular language translation
• University of Cambridge's participation and
role in the enterprise
• Demonstration of capability
• response rates to generated content
Enablers:
• Analytics capability
• Internet proliferation in radio stations
• Incentivisation of radio stations
• Staffing of the enterprise
LOW - Uptake of Platform by the buyers - HIGH
LOW-UptakeofPlatformbytheradiostns-HIGH
14. 9000
11000
13000
15000
17000
1 2 3 4 5 6 7 8 9 10 11 12
Fast Growth
Moderate Growth
…and getting to a successfulposition will not be
straightforward…
14
Challenges Recommendations Analysis Roadmap Appendix
AV Roadmap and Timeline
Month 1 - 3 Month 4-5 Month 6 Month 7
Develop wire
frame and
data model
Proof of
concept
development
Engage, map and subscribe
radio stations to the platform
Platformlaunch
ed
Three applied
pilots
conducted
First commercial
pilot
All community
radio stations
engaged
Go / No Go
Decision to take
platform forward
Analysis of market and current buyer
uptake to determine if and where to
expand, in which order
Uptake of Platform by the radio stns
is critical during this period
Uptake of Platform by the buyers
is critical during this period
Month 8
Begin engagement of radio stations in 2nd
country
Month 9-12
Follow staged on-
boarding of radio
stations in 2nd country
Year 2
Leveraging lessons
learned and
reputational
momentum, onboard
next 3 countries
An acute understanding of the M&E market and which countries hold
the most value for buyers is critical during this period
12 MonthBurn Rate Analysis
Build Period
Go/No Go
Decision
Minimum Viable
Product
Applied Pilot period and
Case Study DevelopmentBuild Commenced
Following Go Decision, Enterprise use /
revenue generation starts
US$
US$
15. …and does not come without significant risks and
uncertainties…
15
Challenges Recommendations Analysis Roadmap Appendix
Mitigation Measures
Risks and uncertainties
Unknown
leadership and
structure slows
development
Engagement of
radio stations
more difficult than
expected
Envelopment by
existing platforms
orNewCo
Cost base
significantly
higher than
expected
Willingness to pay
lower than
expected for
buyers
Funding not
available to cover
build costs
Cambridge
University spin out
difficulties
Listeners cease to
engage with
directed radio
shows
Turmoil causes
development to
no longer be
viable
3G uptake faster
than expected
Legislation
changes mean
storing SMS data
more difficult
Other African
countries too
dissimilar
Upfront engagement of
radio stations and
audiences to demonstrate
potential and prevent
envelopment
Strong team in place from
very beginning
Engage with a small
number of key partners
before build to ensure
funding and continued
success
Ensure capability for
social media and internet
space is designed in
from the start
Engage with the
university to demonstrate
the value of AV as a
standalone enterprise
16. …but the benefits of being able to target, listen and
engage this space are clear to see
16
Challenges Recommendations Analysis Roadmap Appendix
Thank you, questions?
17. Appendices Removed from this
version
17
Challenges Recommendations Analysis Roadmap AppendixAppendix
18. Disclaimer
18
Challenges Recommendations Analysis Roadmap Appendix
“This work has been undertaken as part of a student educational project and the material should be
viewed in this context. The work does not constitute professional advice and no warranties are
made regarding the information presented. The Authors, Cambridge Judge Business School and its
Faculty do not accept any liability for the consequences of any action taken a result of the work or
any recommendations made or inferred.”
Appendix