I was very excited for class this week, because it meant I got to hear results from the first round of hypothesis testing! Validated or invalidated hypotheses awaited.
News and announcements
Class started with news items I wanted to get quickly out of the way. Namely: updating blogs each week by Monday night, completing surveys on Fridays, document your interviews and experiments immediately after they happen, running at least one experiment each week (program is built on a weekly cadence), Demo Day page up (http://ouaccelerator.eventbrite.com), mentor relationships (make sure to keep in touch), and legal updates (need to file to get seed money). I’m also waiting on the debt contract – this will be a good thing to have done in advance of class!
Mentor dinner followup
Then we had a brief recap of the mentor dinner with Elaine Hamm. I asked the students what they thought about the dinner and they also pointed to Elaine’s question about “What do you fear most?” I think this is a really great question, since entrepreneurs are being trained to ignore all warning signs and carry on ahead. The main focus of the accelerator is to test assumptions methodically and critically and only by knowing what you fear can you do that.
Learning Point for order of class
We then started the conversation about the Udacity videos they watched from Steve Blank on value prop. Next week I’m going to start with a presentation of the experiments the teams ran, to close the book on the previous week, and then start the new week’s topic.
Conversation on Udacity videos
I highlighted the following topics I wanted to make sure to cover:
- fact reporting vs insight – here’s what happened vs here’s what it means
- a company is not your invention – it’s the business model. we’re going for product market fit, which means paying customers who will be upset if you go out of business
- what is value prop? what is your value prop?
The main point we circled around was the difference between a value prop and a feature. Value props are what are important to your user. They are more “realtors need better pictures and movies from more angles to better sell their homes” than “realtors need drones”. The key is to not pre-suppose the solution. With a little practice, we got all of our value props into the language a customer would use, which is step 1 towards getting them validated.
Business Model Canvas (on big paper!)
One comment from week 1 was that we filled out the BMC on 8.5 x 11 paper. So one of the team member’s went to Kinko’s and printed out a medium sized version and large version. My test was to see which paper the students took to actually work on and the jumbo sized one won hands down! It’s much better for sticky notes, which means rather than literally rewriting the canvas each week, just the sticky notes can be updated.
Last Week’s experimentw
To being our rework of the canvas, particularly to focus on value prop, each team presented their experiment from the previous week. In each case, the test did not pass. Sidebar: I hate using the word failure here, because a failed test is emphatically not a failure! In fact, it most likely means you will save time and money by not chasing certain ideas any further.
The model for presenting is Hypothesis – Experiment – Results – Action.
A few important ideas came up from discussing the tests. One was that even though the test didn’t pass, the team wasn’t sure if they had gotten good information back. That is purely a test design issue and I expect the teams will get more adept at designing tests as we go along. I always ask what a team would do differently depending on whether the test comes back yes or no. If the team would do the same thing next, the test isn’t good.
Another ideas was the concept of guardrail to guardrail, borrowed liberally from Nail It and Scale It. We had learned in week 1 about how to interview customers without leading them down your path, since you’re trying to uncover frustrations. While everyone can improve that skill forever, the issue was the difference between the customer interview (to uncover frustrations) and an experiment trying to sell something to a customer.
People are pretty good at telling you what they are frustrated with, and they are also pretty good at telling you if they would buy what you are selling. But people are terrible at proposing ideas to their own problems that they would then buy once the solution is built! That’s the catch, and where the entrepreneur does the bulk of the work. I think about it as using customer interviews to learn about problems, then the entrepreneur comes up with a solution that she thinks would solve the customer’s problem, and then the entrepreneur tries to sell the customer the solution. The feedback from the last step, where the customer either say yes (score!) or no (why?) is where all the real learning happens.
Adding in value prop
The teams then went back to the proverbial drawing board to update their business model canvas with both their learnings from their experiments and the value prop discussion. Just in week 2 I saw a few pivots to chase after better customers and to getting people to commit to being beta testers. That’s a win in my book! Early adopters are key to any startups initial growth.
Big ideas I didn’t get to
- Waterfall vs. agile
- what’s a MVP?
- are you assuming the customer problem? or validating?
- Value Prop vs features (painkiller, gain creators, problems/needs)
Steve Blank recommends discussing these ideas. After presenting experiments, discussing value prop, reworking the business model canvas and designing new experiments, we didn’t get to the first two. The teams decided to discuss agile in the startup skills class the following day, so there would be plenty of time for that topic. The point about a MVP being different from a beta is very important, and has come up in side conversations and office hours. Since I’m so close to the experiments, I feel comfortable that teams understand this idea.
What do you want to do tomorrow?
- Practice scrumm and agile?
- Look at different experiments
- marketing, branding?
- talking to customers
This is the list of what I thought the teams might like to do in the startup skills class. I put scrumm and agile at the top because I was hearing that the teams didn’t know what to do next. The accelerator was uncovering lots of assumptions to work on and we were doing a good job identifying the most critical ones. But how to organize teams and day-to-day tasks to maintain forward momentum is a good thing to work on.