Assignment 5: Refinement



Create a minimum viable service. It must measure aspects of customers behavior to test and refine your key hypotheses.  A 15-minute presentation and a (15-25 page) report should add detail to the description from earlier assignments and include:

      the value hypothesis, the growth hypothesis, and supporting hypotheses specific to your service,

      a use case design of the service,

      an implementation a path by which a new user discovers your service,

      an implementation of the service itself, and

      the outcome of measurements and analysis.

Learning Objectives

      Formulating testable market hypotheses

      Understanding the process of acquiring a user

      Designing web/mobile sites

      Implementing web/mobile sites

      Analyzing results

Grading Criteria

Refinement Phase

Quality of hypotheses (relevant, clear, testable):  20%

Use Case Design of minimum viable service: 30%

Implementation of marketing funnel and the minimum viable service: 45%

Style of presentation and report 5%

Overall Promise

The course syllabus says that 10% of the final grade is based on promise, so your final presentation should summarize the case for your business being a good one.


Presentation Outline

Since these presentations will be reviewed (offline) by some venture capitalists, use Art Bonis ten-slide outline. Most of this material can be taken from previous assignments.


o   Project Name

o   Team Description

o   Contact Email


o   What need(s) are addressed?

o   How are they met today?

o   What is missing?


o   Describe service and/or

o   show the video sketch.

      Business Model

o   Show the value Proposition


o   List your team and their skills.


o   Describe the implementation.


o   Describe the possible markets.

o   Report on what was learned from surveys and the minimum viable service.


o   Describe best competing services.


o   Estimate revenues and costs.

      Current Status

o   Show current implementation.


Write what you will say orally in the ppt notes section so that absent readers can get the full picture.

Report Outline


The report need only describe the work specific to this phase.

      The value hypothesis, the growth hypothesis, and supporting hypotheses specific to your service,

      A use case design of the service.

      A description of your advertising method.

      A description of your implementation.

      The outcomes of measurements and analysis.


Process hints


      Define the hypotheses to be tested:

o   the value proposition, which you have been working on already, and

o   the growth hypothesis: how you get users and customers and how to make their number grow


      Design your service on paper as use cases with wireframes. Read Cockburn for the general idea. Write the use cases abstractly, leaving unspecified how its implemented—phone app, web page, email, tweets, written notes, phone calls, etc. See the template and example for a good way to present them. Your design should be viable in the sense that many users can get enough value from it to keep using it. You might not implement all of it during this phase, but all then needed parts of the use case should be there.


      Design the channel(s) for acquiring customers. Experience with the surveys might help, e.g. use reddit. A typical channel would be Google Ads, leading to a site or the Google Play store for Android (costs $25). You can actually include your video sketch in the information submitted to Google Play. Less obvious channels are blogs, Facebook, paper posters, etc. ; but the web ads and stores usual provide analytics. Use your expense account.


      Implement the minimal viable service. Eventually your service should be supplied through many channels—web, phone, email, etc.—and scale to millions of people. I recommend Heroku and Ruby on Rails for that. But now, with only three weeks, you have to take some drastic shortcuts:

      You can supply the service by hand! Each of you can take turns receiving and replying to customer requests. If it gets to be too many, you will have collected enough data.

      You could perform all the communication via email and texts rather than apps.

      You could use one of a wide variety of site creation tools, e.g. WebSitesIn5, Drupal, or Balsamiq. Some of these cost money but are within your $100 budget.

      If your service is to be used on a phone, it is easiest to use the web browser on the phone.  You might have to modify the design to make it fit. There are also some services that help this process. WebSitesIn5 includes a mobile page option.  There are sites that will convert your web site into a mobile app. If you need customers to register and log in, the free site is very helpful. You could make it a Facebook app that also handles logins. You could even try disguising a Google form as an app.

      If you want to create a mobile-only site you could try,, or whatever else you can find by Googling mobile app creator.  Also, suggestions are coming in from my Quora query: This might be more difficult than the web site approach.

      In any case, you need to build in analytics that tell you what each user does from the moment they access your service. Some of the creation sites, e.g. WebSiteIn5, help with this. Google Analytics can be used.


      Your goal in for the implementation is to get as much information from potential users as possible. Since the most likely action of any user at any stage is to just go away, you should prioritize earlier stages. Therefore,

1.     Find as many good places to advertise the service as possible, make the description appealing, and make it easy to find the service. Try to get thousands of views. Less than a few per cent of viewers will go further.

2.     Make the initial display in the service as appealing and clear as you can, so they might continue.

3.     Try to get them to take one or two steps past registering. Getting more than a hundred here is plenty.

4.     Things that they wouldnt get to during the early stages can be left unimplemented.

5.     At every stage collect as much information about what they do as you can.


      Launch the service. When you have enough data, or time runs out, shut it down somehow so people stop coming.


      Analyze the results. Conclude which hypotheses are true, false, or undetermined.


      You are not required to base your thoughts on the reading or lectures, but they should give you an idea of how I grade when its subjective.


      Any number of you can make the presentation, but rehearse multiple times so that you are consistent and not redundant.



Place the items (presentation, pointer to service, report) in the teams Dropbox folder by midnight, 5/2. Also fill out any surveys you are mailed.