Building Polar’s Product/Market Fit Engine
For the past month or so I’ve been revamping Polar’s product-market-fit engine to help understand our users and to prioritize feature requests.
More importantly, it’s helping me build a product for the long-term based on science rather than whim.
Wouldn’t it be great if we could build a engine that could systematically build a product roadmap but also help ensure that we’re building something that users actually want?
The Field of Dreams Fallacy
In the movie Field of Dreams, (spoiler alert…) the protagonist hears a voice that tells him to build a baseball field.
Later in the movie, after the field is built (and spending all of his family’s money in the process), thousands of people arrive out of nowhere, bringing in money to help save his farm.
In reality, this almost never happens. Apps that explode and take off virally are very rare and the most successful viral apps are actually the 4th or 5th attempt by the founders to get something to take off.
The truth of the matter is that it’s very difficult to build a highly successful app and when you do there might not actually be a customer acquisition model that actually works well.
Many startups and open source projects have an amazing product but can’t find users because they don’t have an affordable customer acquisition channel.
So, how do we solve this problem?
Product/Market fit (PMF) is the idea that you’ve both found a market (a pool of users that want to buy and use your product) as well as a product that resonates with them and actually solves their problem.
This is the holy grail. If you can nail this you’re going to have happy customers/users. They’re going to tell their friends, you’re going to find new users just by word of mouth, etc.
More importantly, they won’t churn. Specifically, they’re actually going to use your product and keep coming back.
Churn is the idea that people check out your product, then never return.
The best way to reduce churn is to build something that people actually want.
This is where the PMF engine comes into play. It provides a scientific framework for listening to customers and bootstrapping to PMF.
Using Science and Statistics to Hit Product/Market Fit.
I’m a scientist and it’s how I see the world. If I can’t quantify something with statistics I can’t improve it over time.
A scientific approach to actually running a company/startup/project is going to be profoundly superior than just throwing darts at a board.
The Lean Startup was the first book that I read that really put this at the forefront.
Recently, Superhuman pushed this idea further with the idea of actually using user feedback in a more scientific manner to nail PMF.
The central focus is to improve your features set for users who already love (and preferably pay for) your product.
Make Something People Love
Paul Graham’s observation that you should first make a small group of users love you is a very important insight:
Ideally you want to make large numbers of users love you, but you can’t expect to hit that right away. Initially you have to choose between satisfying all the needs of a subset of potential users, or satisfying a subset of the needs of all potential users. Take the first. It’s easier to expand userwise than satisfactionwise. And perhaps more importantly, it’s harder to lie to yourself. If you think you’re 85% of the way to a great product, how do you know it’s not 70%? Or 10%? Whereas it’s easy to know how many users you have.
Users that are insanely happy with your product are more likely to recommend it to others, give you money, give you good reviews, etc.
Users are who are only moderately interested in your product are going to provide you significantly less value.
They’re significantly less likely to upgrade, less likely to be loyal if your product has a small issue, more likely to stop using your product, etc.
Talk to your Users
Build out a meaningful set of user interview questions to gain important insight about your app.
Actually talk to your users. No email. Talk to them.
I then setup a script to send off 10 emails at a time to our users inviting them to time that works for them.
I find that since only about 10% agree to a user interview I just keep sending out emails until my calendar has enough user interviews scheduled.
Try to target about 3-5x per week.
OK. I know I said to talk to your users. But don’t JUST talk to them. Some don’t have time but still have valuable feedback. Take a modified version of your user interview questions and create a Typeform that users can complete in less than 1 minute.
Use a lot of open ended questions. Ask for data points too.
Avoiding Survivor Bias
One of the main issues is that most of the people who reply to our survey’s tend to be people that really like Polar.
The problem is that this doesn’t help me solve issues that are causing potential users to churn.
Find this subset of users and reach out to them directly and talk to them. Pay them if you have too.
This can be valuable insight.
Here’s where the main value comes from and that’s customized feature priority to help you nail PMF and make your users insanely happen.
Here’s the most important part though - do not worry about making everyone happy!
Some of your users aren’t going to like the app. Maybe you just don’t work well on their favorite platform. Maybe you’re missing a key feature that’s not really on your roadmap.
It doesn’t matter. If they don’t value your product don’t value their features request and feedback as highly as other users.
If you do you’re going to be optimizing for building a general product and we need to focus on landing a beachhead.
We need to figure out some subset of users that are insanely addicted to your product.
Not all features are equal in terms of complexity. Some you can bang out in 1-2 hours and others require months of planning and work.
This needs to be factored into our planning or we might be working on the wrong things.
Now score the customer demand but segment based on how they feel about your product. There are a few ways to do this.
The easiest is to segment. FIXME..
This is slightly controversial but sometimes it’s important to ignore your customers and implement a feature that will have greater long term impact on the company.
For example, if you have two features, but one of those features causes dramatically increased user retention, go with this second feature.
These type of features will yield higher revenue in the long term which means you can have the financial resources to build out the rest of your feature set.