Lesson: Finding Customers with Cindy Alvarez

Step #3 Hypothesis: Common mistakes

So for the first thing that I tend to tell people to start off with is forming a hypothesis. And there's a lot of assumptions that go into your hypothesis and I think that's probably the first good place is to say, "My hypothesis is" and one of the formats I usually recommend is, "I believe this kind of person has this kind of problem doing this kind of activity." Or "I believe this kind of person has this kind of problem that could be better with this high level type of improvement." A lot of times I recommend that people do hypothesis generation as a team.

What's been really useful I've been doing at workshops recently is having everyone on a team write down what they think the hypothesis is and then everyone reads it out loud. And even among teams who are incredibly in sync, who might be working in the same building and thinking about this together all the time, there'll be gaps. It's like, "Oh, that's your assumption? That's your hypothesis?" And between all of them you can say, "Well, whose do we feel the most strongly about?" And then inherent in that we say, "Let's challenge that. Let's all look what are the assumptions inherent in this? Is there an assumption that people have access to a certain kind of technology? Is there an assumption that will magically solve the distribution problem? Is there an assumption that this kind of person is a target market?"

There's a lot of things in there and one thing that's been helpful I find is giving people examples of where assumptions could take people down the wrong path. For example, I was working with the startup, whose hypothesis said something about our assumption is that parents are having this kind of problem. It turns out that parents weren't their primary customer. It had to do with buying age appropriate toys and so parents know what 9-month olds play with and what 2-year olds play with.

The people who don't remember that are the aunts and the uncles and the grandparents which is actually a bigger market. I said, "This is exciting. You have more people to go after who have this need." But the first order assumption was who cares about kids? Who buys kids presents? Well, it's parents. Well, no, it's actually these other people who definitely want to buy presents, who definitely have money to spend but don't actually remember. I have a 5-year old and a 9-year old. I don't remember what 3-year olds do at this point. It falls from your head so quickly. So at the hypothesis stage, that's your first, "Okay, let's see some assumptions."

And then as you start asking questions of customers, this is where again I say it's great to bring a smart person who's not you. We do customer interviews quite often in pairs on my team. So one person will lead the interview, they'll the ask the questions, they'll be making the eye contact. The second person is just taking notes but they're also listening because that second person is going to notice when you ask a leading question. They're going to notice that the customer said yes but their body language said no. And that second person afterwards can often come out and say, "You know what, you said this and I think you're leading the witness" or "You said this and I think that's assuming that this person has this kind of problem." And so they answered as if but we didn't really ascertain if they do or do not have the problem.

There are a handful of pitfalls that people fall into, common traps. One of them is starting out too narrow. So I would kind of say this is cheating at customer development where you have a product, you're pretty sure that you want to build this thing and so you just kind of back form your hypothesis to basically say, "I want to build widget X. Therefore, my hypothesis is that customers need widget X." And that's not actually going through the process.

That's trying to game the system and that's trying to like line everything up so that you get the results you're hoping for. And so that's very common. It's actually very difficult for people to abstract it up a layer and say, "Okay, the reason that I want to build widget X is that I believe that there is a problem. I believe that there's a certain kind of person who has this problem and that they have the resources to want to solve it." And that's what your hypothesis is, that there's a problem space in a customer. But if we go in saying, "I think the world needs this," you're going to get a false-positive or a false-negative. You're not going to really learn. That's one of them.

Another is going in with kind of leading the witness in biased questions and this is very hard to do because you are very enthusiastic. And most of your life as an entrepreneur you're trying to sell stuff, you're trying to sell people on your beliefs and your visions. And so you have this idea and the first thing you want to do is say, "Isn't this a great idea? Don't you want this?" And that again will get you false-positives most of the time because people want to be polite. Or if you are a compelling seller of a vision people will start agreeing with that without even really thinking about how it fits to their life.

What used to be very common in user research and I think thankfully is becoming less so is showing people mock-ups of a product and basically saying, "Would you use this?" And what happens then is that you shaped people's answer based on that. They're now looking through a product sized keyhole and trying to figure out, "Okay, of the thing you've showed me, let me think about how this would fit into my life," as opposed to, "Let me tell you about my life." And then you as a good product person or entrepreneur can pick out what thing that would meet those needs. If you go into a store you're going to make buying decisions based on what's in the store, whereas someone asked you what you really needed you would probably come up with something very different.