Information and Communication Technologies for Development

http://www.youtube.com/watch?v=7JW8BIoo_y4&feature=related

Sorry for not posting this earlier but do check out this revamped version starring our very own Guy and Edison! (Kudos to them too for the creation of the demo.)

Most of us are signing off for the summer, except for Brian who will be continuing the project with Jonathan at Nokia. We look forward to hearing Jonathan and Brian's updates in the fall.

Hope everyone has a fun and fruitful summer wherever you are -- Cambridge, Denver, Kuala Lumpur, Mexico City, Nairobi or San Diego!

What fruitful pre and post-class discussions today! In light of two new components to our final deliverables (especially the documentation, we redistributed our duties after class.

Guy and Edison
- New demo video + subtitles
- Make sure prototype is working
- Demo to Luis before May 22nd

Akani
- Write up survey findings and testing feedback sections
- Develop financial model (incl. assumptions) and success metrics
- Summarize the problem description and existing solutions sections

Brian
- Write up system architecture + diagrams
- Write up documentation/tutorial and license

SuiLin
- Write up marketing, content-building and revenue strategies, future development sections
- Edit (add references, check links) and convert presentation to a pdf
- Create cover sheet about contents

SuiLin will do the final compilation of the report and Guy will burn everything (report, video, presentation, source code) into the DVD to drop off at the Media Lab by 5pm on Thursday.

It's the home stretch. Team Mosoko, we can do it!

Kudos to everyone for making today's presentation a success!

Below is our game plan for the final lapse for the semester:

Writeup:
* Akani - problem overview, existing solutions, survey findings
* Brian - system architecture + technology, testing feedback, success metrics
* SuiLin - sustainability strategy, future development, editing

New video and perfection of prototype:
* Guy and Edison

We're nearly there!

Do check it out!
http://youtube.com/watch?v=4uT9xobdX1o

Credits to Guy and Edison for the filming and subtitles.

Sorry for not posting earlier. The past two weeks have been crazy on my side with my honors thesis due on Tuesday (April 29) but it's done. Yay! Only an oral defense left on May 12.

We're going to have a busy week ahead. By divided our responsibilities, we should be able to get things done well and not overwork anyone (hopefully) :)

Video of demo, prototype update and testing - Guy and Edison
Printing of Poster, sustainability analysis and translation of final survey - Brian
1st survey analysis and creation of final survey - Akani
Compilation of presentation slides, report and sustainability analysis - SuiLin

6 slides for the presentation + Demo
1. Market/Problem
2. Existing solutions
3. Our solution + system design [Brian]
4. Survey Results / Customer Needs + chart/table [Akani]
5. Testing Feedback [Guy/Edison]
6. Sustainability Strategy [Brian/SuiLin]

May 8th
12-1 Brian, SuiLin
2:30-3 All of us
4:20-5 Akani, Guy

After a quick look it seems that there are some interesting findings:

0. Most common type of goods: clothes, electronics, home appliances
1. there is a market for Gas Cilinders :)
2. people sell/give away many cell phones
3. people tend to buy more through market or shop
4. people tend to sell more through their close network (friends and relatives)
5. people look for jobs and recruit employees through close network
6. same for house hunting
7. same for house selling, but a little bit more presence of classifieds
8. people like to buy cars going to a market
9. they sell cars through classifieds and friends

Anyway, more details to come in the following weeks

Today Jonathan, Edison and I met with Nick Sullivan, the author of the book,"You Can Hear Me Now" on Grameenphone's story.

Some of his insights include:
1. Leverage existing structures
e.g. merchants/self-employed workers for content building; phone credit top-up outlets for marketing; M-Pesa for transactions

2. Focus
- identify target group (e.g. merchants? students?), roll out with the most preferred category to test market then expand into other categories

3. User experience
- Nick prefers directed search (asking users questions "What car brand are you looking for" and "What phone model are you looking for?" ) to searching by keywords because users may not be familiar with the concept of searching and "hand-holding" might help to reduce negative initial experiences. The likelihood of relevant search results might be higher too.

Coming up next...
- testing of prototype among African student community (Guy and Edison)
- analyzing findings of preliminary survey and developing closed-end one (Akani)
- translating surveys into Swahilli (Brian)
- doing research on Cellbazaar and M-Pesa (SuiLin)

The 2 updates:
(I) Our survey is currently being disseminated in Kenya! Many thanks to Professor Tavneet Suri and her research team's help as well Rachel Hall-Clifford's feedback. We look forward to the findings.

(II) Kudos to Guy, we have a demo up and running! It's still work-in-progress but we've cleared the first milestone we set for ourselves.

The 2 themes during this week's meeting:
(I) Safety - My Kenyan friends in the States kindly took our survey. Their main concern was how we would regulate the authencity of the posts. We decided that we would offer users guidelines of some "best practices", e.g. meet at a public place, bring a friend with you etc. Since we are unlike e-Bay that acts as a medium for market clearance, we don't need a similar buyer protection scheme. Instead we will adopt "terms and conditions" and "disclaimers" similar to Craigslist.

We have finalized our survey and have sent a copy to Rachel Hall-Clifford for her comments. The next thing we are brainstorming is Mosoko's roll-out strategy.

The rollout strategy is two-fold: (i) marketing (ii) user education. Our current idea is to use micro-blogging by voice/text message, as it'll be easier to move someone from "advertising" themselves and searching for their friends' status updates, to advertising and searching for goods/services.

We are still figuring out a revenue model that will make this initial "micro-blogging" application sustainable, e.g.
- costs for posting/retrieving status update?
- ad-free but subscription fee vs. free but with ads?
- post 1 update, get 3 free retrievals?
- charge for more than 10 friends?

If you have any ideas, please let us know!

Some sites we are currently studying are
- Dodgeball
- Twitter
- Jaiku
- Pownce

Here are areas on the business side of Mosoko that we will be focusing on:

Survey creation – SuiLin
Survey translation – Brian
Revenue strategies – Akani and Brian
System Analysis – Brian
Partnerships with content providers – SuiLin
Cultural context – SuiLin
Rollout and long-term strategy – Akani

Guy and Edison met with Jonathan after class regarding the technology side of the project. We look forward to hearing their updates!

This week Jonathan and I met with Professor Tavneet Suri regarding our market research survey. Professor Suri kindly offered to help us disseminate the survey in Kenya while Jonathan has contacted Nokia Nairobi for Institutional Review Board permission. As I formulated and revised the survey questions and options, I kept questioning: "Can we ask this? If yes, how?"

I had two concerns: 1. Trust 2. Patience

1. Trust
- How familiar are our survey respondents with surveys? Would they be willing to share personal information, such as their monthly incomes, business and social networks with us? This information is essential to estimate their willingness to pay and learn how they currently buy, sell, rent, hire or find jobs.
=> Respondents will be able to select from various ranges. The more personal questions will be left till the end of the questionnaire.
- What is the users preferred mode of communication: voice or text messaging in lieu of face-to-face communication, especially with strangers and business transactions?
- Respondents may entertain the idea of being able to search and post ads through cell phones but how do we gauge their confidence in the Mosoko database's authenticity that they embrace this seemingly "disruptive innovation"?
=> Hopefully the Nokia brand and partnerships with widely-circulated newspapers will help. We will assess this during the prototype testing.

We made great progress with week. We now have a more detailed picture about how Mosoko will actually work. It will have an hybrid interface with key pad based browsing and voice based posting and searching. One key feature we added is tagging. We will ask people to tag their posting, these tags will then be used to match with people searches.
The survey has been finished and with the help of Prof. Tavneet Suri we will be able to collect 30 responses in Kenya
The next targets are to finish the first prototype after we come back from spring break and to continue refining the sustainability/business model.

Added to the various modules currently under construction and testing, Mosoko has built some consensus around a possible way on how the product shall function. Please find attached... Feel free to add comments and suggestions on improvements that may be incorporated into this initial draft.

By far the most successful form of classifieds in Kenya are paper based and more particularly those posted by the local leading newspapers. One of the leading local papers actually has its own weekly advertiser devoted purely to classifieds. These are also posted online but currently there are no mobile applications for collection and dissemination of such content.

Nevertheless, there are numerous value added texting services dedicated to a specific classifieds category most popularly jobs and social networking (dating).

Attending Dave Castell’s talk made me ponder how Mosoko competes with the status quo – existing sources of classifieds, where people can buy/sell goods and find/post jobs. This week I studied five non-Kenyan companies to see how Mosoko could differentiate itself. (Brian studied the local Kenyan options.) The companies I analyzed are:
- Cellbazaar (Bangladesh)
- Craigslist
- YouPost (South Africa)
- Kijiji (E-Bay owned Craigslist-like classified operating in 30 countries)
- Babajob (India's Linked-In for maids/drivers)

We believe that Mosoko’s value proposition is
1. it is more inclusive
~ To cater to illiterate/semi-literate users who don’t understand English, Mosoko uses an interactive voice recognition system for both Swahili and English speakers. Shared cell phone users can also use Mosoko because they can have individualized IDs to access their search results and conceal their cell phone number if they don’t want to reveal their identity.

2. it provides users the convenience of “anywhere, anytime”
~ Similar to Cellbazaar, Mosoko is accessible both via the web and cell phones. Users are not disadvantaged if they are not covered by the current classifieds, e.g. those who do not have frequent internet connectivity or are not within the circulation of the printed classifieds. They can post ads real-time through voice-recorded or text messages that will be converted into database entries that are searchable by voice and text.

Yesterday we had an interesting discussion at our weekly team meeting.
As we are getting into more detail about how Mosoko interface will actually work. We debated what are the implications in terms of cost, programming complexity and customer experience, the key question from our discussion was:

- How do we balance the browsing vs searching capability of the interface.

On one extreme case would be to ask people right up front what they are looking for, conduct a search in the postings database and get back to them with the query results. The main drawback from this approach is that the query will have way too many matches, that would make the search results irrelevant. Although it would be very neat from the user perspective just to call Mosoko and mention 3-4 key words to conduct the search, this means a short call time and lower air time cost for Mosoko.

Akani - Market research on Kenya demographics (income, occupations, telecom market etc)

Brian - Case study of similar projects' experiences in Kenya & Write-up (due on March 13th)

Edison - Formulate survey (produce initial survey and we will revise as a group)

Guy - Research technology / interfaces used by comparable telecom/internet applications - capabilities, limitations etc.

SuiLin - Research business models (revenue, marketing etc) of comparable telecom/internet companies