Finding Product Market Fit 3/6 — Habit
Welcome to installment three of this series on product market fit. Here is the previous piece. I’m John Danner, Managing Director of Dunce Capital. In the last four years, I have made over two dozen pre-seed investments in the space including Lambda School, where I am on the board, Outschool, Prenda, and Contra. Before Dunce, I founded three companies, NetGravity (IPO), Rocketship Public Schools (12k students) and Zeal (sold 2018).This is what I work on fulltime with founders in my portfolio, so I hope it adds some value for you! If you like this, you will probably like my substack as well.
The Goal to Exit This Phase — 65% of your new users should retain d30.
Get Ready — Measurement
You have enough users now that you can get quantitative results for your experiments, which helps a lot at guarding against founder gut being the decision maker. Now is the time to put a strong measurement system in place. You may have to play with a few different measurement systems to get one you love. Some use heap.io or mixpanel, other people love segment, and a new one is being invented every day, so whatever works for you is fine. Put strong measurement in now, because it gets much harder to do later when you have lots of experiments running, and you are beginning to have enough users that quantitative results will be useful, even if not statistically valid yet. You need three things out of your measurement system:
1 — Lead attribution so that you can understand your referrals. For example, if a parent of a student in Joe’s class shares their class video on Facebook, when their friends click it, you know where they came from.
2 — Funnel analysis showing where new users drop and retain through your stages of activating a new user.
3 — Retention cohorts of your users by week so that you can see their engagement (logins fine at this stage though actions crucial later) and correlate that with their retention.
Creating Habit Intentionally
Users must form a habit around using your product. Hooked is the best book about this, so read it if you haven’t. When people use your product several times per week, it becomes something they do automatically.
Someone has seen the magic and is psyched about your product. That doesn’t mean they will become a habitual user. Don’t leave this to chance! My favorite example of this is the meditation app Calm, and their feature called ‘Daily Calm’. Before Daily Calm, I had to go into the app every day and find a session I wanted to do. That slight amount of friction was enough to keep me from forming a daily habit. When ‘Daily Calm’ came out, it made it so simple that I easily became habitual.
Depending on your app, you will have to decide how many interactions a user needs to have and what is meaningful to form a habit. Figure out how your app is going to improve their life every time they use it. It’s awfully difficult to form habit if someone isn’t using your product at least two or three times per week, although daily (or hourly) is much better.
It will be obvious to some, but if you can retain a user into month two, your user lifetime value (LTV) doubles :) Measure this with weekly retention cohorts so that you can isolate the effect of your experiments from past user behavior.
Back to Joe
Joe started charging after the first session and had the vast majority of students pay. Now he needs to make sure this sticks. He realizes that the best way to create habit is by associating himself in a student’s mind any time they are frustrated with math. The general habit pattern is Cue -> User Response -> System Reward. In Joe’s case, he wants to remove friction between frustration and being able to help the student (reward), so instead of teaching scheduled math classes with students, his team is available all of the time to help them. That way, when they are frustrated, they come to Joe, he helps them, and it strengthens the habit. The simple act of removing the friction between the cue and what a user needs to do to pick Joe has a lot of value.
In experimental language, Joe’s hypothesis is that frustration is a student’s main motivation for working with him (as opposed to being a great mathematician for example). Based on that, he tries a bunch of experiments about satisfying that frustration (like being able to text message between classes) before finally moving to on-demand help for students.
As with magic, my opinion is that until you get habit to 65% d30 retention, you are wasting your time worrying about later stages, because so many people fall out of your funnel that growth doesn’t really work. More specifically, target 80% d7 retention retention and 80% of those users retain d30. That puts proper focus on getting habit working in the first week.
Usually you need to achieve both habit and the next stage Discovery to get to a $100k MRR company (10k users at $10/mo for example) and raise your seed round, so I will discuss it at the end of the Discovery section.
Click here for my fourth piece in the series, on Discovery!