This last section focuses on tactical applications. But first, here’s a quick recap of where we’ve been in Chapter 1:
Most founders start with building; they assume their idea is fact, jump straight to delivery, and rack up enormous opportunity costs building the wrong thing. Then they “release and pray,” hoping the product is valuable. Most of the time, it’s not. Sometimes this kills the startup.
Smart founders start with discovery and learning about customers. Rather than Build → Learn → Measure, they Learn → Measure → Build. They recognize their idea is based on assumptions, and they form hypotheses to test those assumptions. They especially analyze value. In software “is this valuable to customers?” is, almost always, your biggest risk. And discovery is the process you use to de-risk big assumptions, reduce opportunity costs, and place a smart bet on what you build.
Companies who do this type of learning and discovery well are customer-driven. They’re the ones who gather rich information about customers, leverage that information to build value delivery systems, and build profitable businesses. They’re the companies who not only survive—they thrive. And they do it by learning and applying what they learn. (Think Amazon, Intercom, Casper, and Drift.)
Here’s how you can transfer this to your startup.
You’ll want to start where smart founders start: recognizing your biggest assumptions.
There are all kinds of assumptions baked into your product idea, including:
And many, many others. So how do you decide which one to focus on? You sort your assumptions by risk; some assumptions matter much more than others. For example, assuming your target market will pay money to eliminate a pain point is a much riskier assumption than assuming you can reach your audience via Twitter. Sure, crazy high social engagement is nice, but only if the problem is significant enough for people to pull out their credit cards. Otherwise,1,000 likes are about as useful as 1,000 thumbs up stickers collecting dust on your desk.
Here’s a list of assumptions to get you started:
(To create a version you can manipulate with your own data, hit the “Copy Base” button in the top right corner.)
Filter your biggest, riskiest assumptions (hint: they’ll deal with value). Then, from that high-risk list, highlight the ones you have zero evidence for. The ones where, if a friend said, “prove it!” you’d have absolutely nothing to show them. These are the ones you need to test ASAP with a hypothesis.
A hypothesis puts you in an experimentation mindset; it keeps you focused on learning as opposed to building. As Hiten Shah points out: “Ideas are just a description of the solution to a problem. Converting your idea to a problem hypothesis enables you to stay focused on validating the idea instead of building the solution.”
Use this formula to create your hypothesis: We believe doing x will result in outcome y for the customer which will have z business results.
Then use this Airtable to track your assumptions and hypotheses:
The table is specifically designed to be something that sticks with you—a living database. It’s a format you can come back to each time you test an assumption, or whenever you launch a new thing. And as you document core assumptions, test hypotheses, and identify risks, you’ll compound learning and build a learning machine.
This table is also designed to help you start thinking about what you’ll do with your hypothesis. Depending on your constraints (time, resources, and budget), there are a variety of discovery methods you can use to prove or disprove your hypothesis. I’ve added a list of those, plus some related tools, to help. If this piece feels overwhelming, don’t panic. I’ll cover how to test your hypothesis, and what to do with the data you get, in later chapters.
Once you have your initial product hypothesis. Run it through this gut check checklist. It’ll not only force you to think carefully about the hypothesis you’ve created, it’ll encourage you to think about how you’ll test it, how you’ll know if it’s valid/invalid, and who can give you the data you need.
If you can check each box, it’s go time.
So you know your startup needs to be a learning machine. And you know effective discovery, fueled by a strong hypothesis, is how you start learning. Good! You’re already ahead of the game, and you’ve already saved several weeks of time and thousands of dollars.
Now, you need to know how to keep making progress. How to take your hypothesis, prove or disprove it, and do something with that information.
In other words, you need to understand customer research.