Business Canvas: Week 6

There were two main changes to our business model this week.

  1. The idea of ‘communities’ was moved from “Customer Relationships” to a “Value Proposition” due to the customer discovery session proposing the idea of “Yelp for Diabetics”. Lots of interesting potential here. We need to test this hard.
  2. ‘Channel to doctor’ as a value prop may not be what we thought. However, Oren’s suggestion of finding out if physicians would pay for a “Mint-style” analysis is another area we need to test.

…in summation, lots of testing to be done.


This slideshow requires JavaScript.


Phone Interview: Brittany T.

This morning we had the pleasure of speaking with a very experienced Type 1 from southern California, Brittany. She graciously agreed to speak with us despite having to go into work on a Saturday morning. For background, Brittany intensely works out 5 days/week for 1+ hour sessions, mostly cooks at home (vegetarian), and eats 5-6 small meals evenly spaced throughout the day. In other words, she is very active and healthy. She was diagnosed with Type 1 over 18 years ago and has been at the stage where she can essentially pin point her glucose levels without the need for testing. However, she still tests six times per day. Her understanding of diabetes is also very high. Her college major was in Nutrition and she is still studying to become a certified diabetic educator.

Despite being a data-driven person, she finds her largest challenge with diabetes to be the time it takes to properly manage it. Her main reasoning for not using applications to help her manage is the amount of time it took to enter in all the information, and the lack of a personalized database (i.e. app has calorie tracker, but it has no selection for a cup of broccoli; only premade dishes).

Another interesting piece she identified was that when she would go to her endocrinologist with several months of intensive data collection, he would say, “Whoa, I don’t need to see that!”. After which, she figured “what was the point…”. We will contact a few endos to see if the reasoning behind this is because the data is too large to go over in one session (i.e would it be better if we made the feedback loop shorter), or they just have no interest in it at all. One contradicting piece about exercise she provided (with respect to Dave’s response) was that she gauged the intensity of her workout qualitatively and found that metric to be the most important to her.

Thank you to Brittany for spending her Saturday morning with us, and for graciously agreeing to further discussions!

AdWords Campaign

A couple of days ago we finalized the details of our Google AdWords campaign. We decided we wanted to test the differences between “Diabetes Management” and “Diabetic Community”. After hearing so many ideas about the social aspect of diabetes, and having our most recent customer discovery session involve this social aspect, we decided it was important enough to test as a core value proposition against the analytical management of diabetes.

Our AdWords credit will not come in for a week, so we will have a waiting period until we have data to analyze and report on.

In the meantime, we will be testing this dynamic directly through our customer outreach.

Customer Discovery Session: Dave Palmer

Happy Belated Lead Day! This past Wednesday (2/29) we were granted a special Lead Day gift of having the opportunity to hold another customer discovery session. Our guest of honor was Dave Palmer, a middle-aged Type 2 diabetic who works for the U.S. Treasury and SF Police Department. His meal of choice was Arinell’s Pizza, loaded with meat (our customers love pizza!).

Aside from the difference in demographics, the key difference this time around was that we had a pre-written script of questions. This was to prevent us from proposing “leading” questions that could affect our customers responses. The beginning was just “ice breaker” questions to get to know Dave better. Going into the session, we knew that he was a passionate individual (he had been exchanging emails all week about interesting ideas he had – awesome!). The most interesting fact we learned was that he not only attended a gym 3-5 times a week, but he enjoyed exercising. He told us how his body would actually feel bad if he did not workout for multiple days. This lead to his one strong piece of advice for any diabetics, “join a gym”. For him, dietary changes only did so much, but the changes he made in his activity level helped him lose significant weight and control his diabetes much better.

For brevity, here is a list of some the most interesting insights Dave provided:

  • Optimal glucose levels are different for everyone
  • Possesses internal gauge of how his glucose levels are (can guess within 10 mg/dl)
  • Likes the idea of “checking in/out” to workouts
  • Visual representations > text (wrt how to input information)
  • Lack of diabetic support group / community
  • Yelp for Diabetics
    • See items/restaurants that previous diabetics have dubbed “Diabetes Friendly”
    • “how to attack the Indian buffet bar”
  • Desires correlation between foods/workouts and glucose levels (further validation of recommendation engine)

Dave’s good friend, Eugene, was unable to make it this week. Therefore, we will set up a new session with him next week. Either way, Dave is a passionate user that seems to fit the earlyvangelist persona. However, we must be sure to account for his data-driven persona (which may not be the norm).


A big thanks to Dave for spending nearly 3 hours with us!

New Hypothesis: Social Avenues

This is one of several belated posts from the past week.

The recurring theme of advice, or new potential markets to consider, from the class this past Monday (2/27) was that of social networks. Now we realize that it is futile to try and create the next big social network. However, the use of a niche social community to act as a means of increasing user engagement and retention, and to enhance the core value (increased happiness/awareness), was an interesting idea. Although we had already incorporated certain social aspects through game-mechanics (social pressure), we did not consider the larger effect of connecting diabetics.

Therefore, in addition to testing the revenue streams, we will also be testing the idea of a diabetic community to exchange knowledge and experiences.

Business Canvas: Week 5

We have found that our value proposition for the diabetics customer segment needed to be broken down so we could test each individually. In addition, the large response we received on the frustrations from manual input, we added “normal” glucometer makers as a potential Key Partner. Both the invasive and non-invasive makers as partners are big guesses we need to test.

This slideshow requires JavaScript.

Kunal: Customer Discovery Session Recap

As previously mentioned, we had a customer discovery session with Kunal yesterday. He is a Berkeley graduate student in Civil Engineering. We shared had a very informative discussion over some delicious Zachary’s Chicago Pizza (he chose).

For the majority of the session, we took Steve’s advice on running these types of sessions (as opposed to focus groups). So we tried to ask questions objectively, and let Kunal steer the conversation. What we found by listening to each other is that it was very difficult to remain objective and not state our opinions, thereby potentially affecting Kunal’s responses. Despite this, Kunal had some fantastic insights that we never would have thought of on our own. For brevity, here is a summary of the most surprising points:

  • Top 3 Frustrations at beginning
  1. simple reminder to take night time insulin
  2. manual input (excel) → usually doesn’t track
  3. seeing a high reading and thinking “why the hell did that happen, I thought I did it right?”
  • The activity of exercise matter (not just duration and intensity)
  • Leaderboard: “knowing someone was doing better than me would kind of piss me off”
    • don’t show ranks or scores, but people above and below in your category would be helpful
    • ability to click on others to see what they are doing that is working for them
  • Top 3 Frustrations at end
    1. providig feedback/recommendations – remove thinking process (avg. stdev, user sets target of HA1c, show potential adjustments i.e carb ratio from 12-10)
    2. just remember to take my damn night time shot
    3. LifeScore and objectives – some sort of feedback that shows immediate feedback of how they are doing, and if they want they can go further (to make changes)

Check out the slideshow of some pics from the session.

This slideshow requires JavaScript.

American Diabetes Association (ADA)

Meeting people in person is way more effective than email!

Dropped into the ADA office on 9th Street in Oakland unexpectedly.  Met two very nice ladies who were very enthusiastic with our project and what we are trying to accomplish!  They are providing a gold mine of bay area sources which will enable us to increase our contact pool more effectively and much faster!

I showed Ka-Ryn, the associate director, a sample of our app screens – basically the kind of stuff we want to test (see previous post).  She recommended we make the fonts of the app customizable as many diabetics suffer from retinopathy and have a difficult time viewing small fonts – something I doubt we would have come up with on our own!  She was extremely excited about the complete product – both hardware and software, wants to connect us with local diabetic non profits in the bay area for testing our hypothesis’, and thinks local VCs will eat it up.  Sounds good to me! 🙂

Also spammed a number of blogs with survey monkey links.  One had us removed for violating their soliciting policy within minutes!

Customer Discovery Zone

Highlight this week:

Toward the end of our meeting with Kunal, a fellow student and type 1 diabetic, we showed him our low-fi app to see if it satisfied his wants/needs.  After seeing our “Lifescore” he made a few remarks about how it might carry negative connotation (high scores which are good vs high sugar levels which are not good) and how we might improve it.  After hearing his suggestions I asked what he would think if replaced it altogether with an experience bar and level instead.  To that he smiled, and while deep in thought replied “That would be cool.”.

Though we are targeting diabetics, this does not really relate to the disease, so what do you think? A or B?

“Just Fit a Curve To It”

The title was the theme of the past several hours. In the midst of programming, we took a slight detour to devise the algorithm that would be determining the scores for our leaderboard. Initially, we only wanted to perform this exercise to determine what values we had to keep on our server, and how to organize them. Then it slowly turned into chaotic scribblings of math. Looking at it after the fact, we are not quite sure how we even formulated certain parts. However, this exercise was very useful as it cleared up certain parts of the application that were debating over and paved the way for some potential changes in how we structure the usefulness of the application. Tomorrow night we will be holding a “customer discovery session” with a few diabetics. We are hosting a dinner to talk with them in person about their greatest frustrations with diabetes, top items on their wishlist, and iff (if and only if) they mention it we will show them some mockups of the application to gauge the proximity their needs.

This is very exciting to finally have a long (several hour) time frame with some diabetics to get some hopefully eye-opening feedback to solidify/modify our business canvas.

And here’s a brief look into some of the whiteboard mayhem.

This slideshow requires JavaScript.