October 29, 2014 by Matt Chanoff .
“Science is what we do to keep us from lying to ourselves.”
- Richard Feynman
“A man sees what he wants to see and disregards the rest.”
- Paul Simon, The Boxer
Who are your customers? Entrepreneurs have a bias toward action, and you can’t act – you can’t marshal resources, make products, decide on features, develop marketing plans, or succeed as a business, without knowing something about your customers. So for lots of startups, knowing something about your customers is an obstacle, a hurdle you have to jump over in order to get on with the business of building your business.
The primitive approach, abhorred by experts today, is to simply hurdle this obstacle yourself. You look around, see a pattern, and come up with something like this: “My customers are millennials who need to manage their student loans.” This sounds pretty good as target markets go. There are lots of millennials. You can find plenty of research showing that they have a lot of student loans, and that they’re having trouble managing these loans. Together with that story, if you have a record of some success and domain expertise, you can probably raise some seed funding.
And there you are. Time to develop the product.
The more sophisticated approach says “No no no! That’s just a fantasy market. You have to get out of the building and talk to people in order to validate it.” Practitioners will then deploy surveys, with questions like “How much of your income is taken up by paying off student loans?” and “Would you be interested in a product that helped you better manage your loans?” You accost students on campus: “What are your top three worries after you leave college?” “Is your choice of majors impacted by the loans you need to pay off?”
The student loan problem will probably show up in your validation process as a big deal, exactly as you expected. (If not, you’ll pivot and come up with another idea.) If so, you’ll try to further validate, asking people to sign up to learn more about managing their loans, then showing them minimum viable product versions of student loan management systems. And again, you’ll be off and running.
Why doesn’t flashpoint follow this approach?
Look at these pictures:
[image face2] [image face3] [image face1]
Good luck not seeing faces in these pictures. Identifying patterns is what we are innately, extraordinarily, good at. So good, in fact, that once we’ve seen them, it’s pretty much impossible to un-see them. Look at any of those pictures again and try not to see a face.
Those faces are like your market idea. As an entrepreneur, once you’ve seen the pattern – lots of millennials…lots of student loan debt…help required in managing the loans. It’s very hard to stop seeing it.
But there are no faces in those pictures. Seeing a market doesn’t mean it’s there. Going out into the world to validate your hypothesis blinds you to most of what’s really in front of you.
Here’s a video that asks you to go out into the world and solve a problem. Take a minute and 43 seconds to watch it.
Seriously, how many of the 21 things did you notice? No shame – it’s just how our brains are built. I noticed one.
Here’s another way our brains are built: Once you have an idea or see a pattern, you will naturally discount and easily forget evidence against it, and at the same time, become better at at remembering, and less critical of, evidence that supports it. This is known as “confirmation bias.” Suppose you go out into the world on behalf of your startup, and your research discovers the following fact:
Eighty percent of college students click on the adword phrase “help with student loans” at least ten times in their senior year.
If you’re looking to validate your business idea, that’s pretty great news. Do I have a market? Well, of course I do, lots of people are trying hard to find what I’m selling!
But wait a minute. We’re assuming a correlation between people who search on this phrase and people who will buy a solution to their student loan problems. This sounds plausible – if people are searching for something, and they find it, at least most of them may then go out and get it – all I have to do now is get the product right.
This scenario is exactly like the pictures of faces that weren’t faces. It sounds plausible. In fact, it sounds unarguable, because the fact fits the pattern you’ve started out with. But maybe students click on those words because their advisor told them to, or the phrase shows up prominently on their college website, or because Google has profiled them and the phrase just shows up a lot on their screens, or they’ve heard other students complain about their loans, or they have problems getting student loans.
Since the student loan business idea is now planted firmly in your mind, these alternative explanations probably sound like a stretch. But if we had started with a business idea about providing student loans and come across the same factoid, we would have seen it as validation of that idea.
Ok, so maybe all that clicking doesn’t mean that that students will buy your product. It’s got to mean something, right? They’re interested in something related to student loans, otherwise why would they click on the phrase so much? Maybe helping manage the loan burden won’t work, but we can use that evidence to pivot to the “getting student loans” idea.
Actually, it could easily mean nothing. Suppose you were to see a list of phrases college students clicked on, and it looked like this:
Phrase Click frequency per semester
Dinner schedule 420
Activities this weekend 370
Electric cars 190
Study guides 190
Spinach facts 50
Help with student loans 10
Still think you’re looking at a market?
When you go out looking for facts, or looking to validate hypotheses that support your conclusion, you’re very likely to find examples that fit the bill; or at least seem to.
That’s why science operates the opposite way. There are a lot of methods held up as the “scientific method.” What the authentic ones have in common is that they’re designed to try to reduce or eliminate errors. Measuring things rather than judging their size by eye is one such method. Seeking to invalidate hypotheses rather than validate them is another. Without a mindful commitment to using these sorts of methods, scientists, and entrepreneurs default to easier methods that are prone to error. Just like we can’t help noticing the faces in the pictures above, just like we can’t help not noticing the details in the video, we can’t help that our business ideas, our ideas of who the market is, inevitably bias us toward confirming the theory we already believe. This is confirmation bias. If you want to find a market that actually is a market, you need to work against it.
Hat tip to Kevin DeLaPlante at criticalthinkeracademy.com for some of the ideas in this post, and to AGNeuropsychologie for the fantastic video.