Thanks for reading this series on Finding PMF! If you missed earlier articles, here is the start of the series.
If you have magic, habit, discovery and a great growth loop working, getting you to an “A” on my PMF report card, it’s time to raise Series A. You should probably be in the $200k MRR range at this point, which is solid for raising Series A. You are in the top 1% of seed-stage companies and should be able to raise a great round. Here are the things to work on before you raise Series A. They are all negotiable with VCs if your growth and organic leads are awesome. Work on them now anyway because solving these issues along with strong PMF is what makes very big companies. You are going to get asked about them in Series A, so give each some serious thought.
- Your unit model needs to work. Principally, COGS should be a maximum 50% of your Revenue, and LTV/CAC ideally is 6:1 or better. The latter usually happens if you have 80% organics unless your paid is atrociously expensive. The COGS item is important, because you really just need to net out COGS when you think of revenue. That means that R&D, Admin, and CAC need to come out of the remainder. 50% is just a rule of thumb on the maximum COGS.
For example, if Joe is paying a teacher $20 for a half hour one on one class, and charging a student $30 for it, then his gross margin is only 30%. That’s not awesome. The reason to deal with unit economics now is that early, you can do all kinds of unscalable things because you didn’t have many users. But now you need to get yourself set up to scale. PMF achieved by giving away dollar bills for 95 cents is not PMF.
I don’t have a hard and fast rule about how things have to work here, other than your net customer lifetime value (LTV) should be 6 times your customer acquisition cost (CAC). That word ‘net’ is incredibly important. You need to only count revenue after COGS for the purposes of this calculation. Once you do that, you have lots of comparables with software companies that have 90%+ gross margins. Yes, you can survive at a 3:1 LTV/CAC ratio, but life’s too short. If you have 80% organic and a 6:1, you can pump so much money into product and your growth loops that no one will stop you.
2. Your total available market (TAM) for your initial focus should be quite small (facebook’s GTM focused on students at Harvard) but for Series A you will have to make the case about how you can move to adjacencies without changing your product. In general, a bottoms-up market analysis where you take your customer LTV and multiple by a target group of users and get to $1b is fine for now.
3. You need to be thinking about moat now.
It’s just a fact that great ideas get copied. If there is a reason like a network effect that your business is harder to copy, that’s a very very good thing. Moat is the kind of thing that starts to get pounded by Series C, but is extremely useful to think about now, because you are based on organic leads. Since the network you have created is arguably the best form of moat, thinking about it at the same time can be quite useful.
I see a lot of companies think nothing about how they are going to compete if others come into their marketplace until much later in their company development. I don’t think it’s ever too early to figure out what makes the company stronger as it gets bigger. What makes it far more likely someone will choose them than a competitor. A lot of times that can be as simple as brand and awareness. But watch out for this as your sole defense.
For example, if Joe’s math school was awesome but a competitor came along with a great experience, you could imagine the price of the two services falling, which would be terrible. If instead, Joe had a pair problem solving part of his school where he matched advanced students to work together on hard problems, he would start to build stickiness around the community, and not just to himself. You need to think about the moat in seed stage. You may not be able to deliver on it yet, but have some hypotheses and start to test that out. As with everything, read everything you can about moats and start thinking about what you can do,
4. Figure out the things that are toughest to scale and come up with solutions.
If you are fortunate to have a pure software business, then likely the above are the main things you have to do. Many of the companies I work with have a human component to them (the school thing), so a huge part of seed is to figure out how to scale the humans. I’ve written posts on some of the innovative ways my companies have done that, but in general I’ve come to believe that some sort of ‘hack’ is necessary to get scalability. Brute force hiring is just too difficult without some sort of positive feedback loop. For example, lambda school uses current students as TAs to work with a small group of students daily for a few months of their time at lambda. This means the more students they have, the more TAs they have, a nice loop.
5. Double Down on Product
Now that you have PMF in your first niche in the market, it’s time to really understand what they love, and to understand who you aren’t reaching yet. The best way I have seen to do that is with the Superhuman test.You have enough customers to really understand what differentiates core from non-core users. By understanding both what your core users love about your product and what the gaps are, you will be able to focus development time on those areas. It’s really worth it to move these another ten or twenty percent towards what users want. Often that allows you to think more clearly about the adjacencies from the group of users who initially got you to product market fit.
For example, let’s say that Joe got amazing PMF with homeschool parents. What they loved was how his team helped their children reason about high school math. Their two or three areas that they love most are expertise and the range of questions they ask the student. However, there are a large number of users who try joes service a few times but churn out. When he surveys these non core users, they also appreciate the instructors and questions, but they have two main areas that would make them a lot more excited. First, their kids are middle schoolers, not high schoolers. The topics joe is teaching aren’t applicable to middle schoolers and his teachers aren’t great at working with middle schoolers. Second, they don’t have the time to schedule classes with joe. They just want to be able to have their children come on the system when they had questions and work with someone great.
Given this information, the challenge becomes how to double down on the things that the core loves while devoting development resources and teacher recruiting to middle school and scheduling to create a larger audience.
The best metric is the percentage of users who would be very disappointed if you pull your product. This metric is too manipulable for investor presentations, but it is very useful internally. Ironically I have found that a lower number here initially is better, because it shows that there is so much demand for your product that even people who it doesn’t really satisfy are willing to respond to your survey.
I know this sounds like a long journey, and in fact there are a lot of things to get done. I hope that having concrete goals for your company performance will help make decision making and focus easier as you build your company!