How to find your people: A brief guide to crowdtesting
At the thin, early edge of lean startups, you’ll find a hypomanic, fast-talking guy named Justin Wilcox. He’s the one with the mohawk, black tie and beard. Wilcox leads workshops all over the world teaching aspiring entrepreneurs how to figure out who will buy the dreams they spin. The way to find out, generally, is to talk to strangers -- lots of them.
Wilcox is an itinerant preacher in the church that lean built. He travels as far away as Serbia and Brazil to share the gospel that entrepreneurial fantasy should schedule an appointment with the public as soon as possible. As Steve Blank and Bob Dorf, the grandfathers of the lean movement, wrote: “No business plan survives first contact with customers.”
That's because most plans are bad. The trick is to realize they're bad fast and move on. Wilcox had some students in Belgrade who thought the world needed ready-made Tequila shots -- lime, salt and booze all packaged together. The world didn't. But what they found were a bunch of Serbs who could never find a place to party outside their own homes, and voila, they pivoted toward a sort of Balkan Partybnb. It took under two hours.
"I come in and I say: 'You have 90 minutes to build a website -- go.' After that we kick them onto the streets to interview people face to face," Wilcox said. "When they come back, they say: 'No one has the problem we thought they had.' They're pissed at me. They want to know why we made them spend 90 minutes building a website. And I ask them: 'Isn't that better than spending 90 days? Or 90 weeks?' And they get it. They're like, 'Oh sh*t, I'll never write code again.'"
There's something slightly absurd about founders looking for a business model. It sounds like characters in search of an author. Hollywood would have us believe great visionaries nurse their dreams in isolation with only genius as their guide.
"We all have this idea that being an entrepreneur means having an immaculate vision and building on it," said Edward Sullivan, who founded the property management startup Dreamlease and attended a Wilcox workshop in December. "The real process is very messy, and it involves talking to a lot of different people. If you listen well, you can hear their pain points."
In fact, finding a business model is like conducting a job search. People who want salaried work send out resumes; people who want to run a business seek customers with a common problem -- the salaried will have one employer, the founder will have many.
The only problem that lean startup principles run into is human nature.
Our brains are biologically predisposed against lean. We're programmed to respond emotionally to something first, and then rationalize. It’s an irony that startup people, whose role in contemporary life appears to be disruption, would be so attached to fixed ideas. And a further irony that this psychological quirk is called "status quo bias."
"In phase one, before you hear about lean, you go build an amazing product that no one wants," Wilcox said. "That's exactly what I did. After that, if you're still interested in entrepreneurship at all, you look at Lean Startup."
In phase two, people learn a new way of building a business, and they still go out and build something nobody wants. "It's so much more fun and so much easier to go build something than it is to learn about someone else and their problems, and then to go solve those problems," Wilcox said.
Finally, they reach phase three, known outside of Silicon Valley as rock bottom. "When people don't want to fail any more, when they get tired of it and get hardcore, that's when they really start doing what they should be doing."
Wilcox had his road-to-Damascus moment after spending six months on a million-dollar idea that ended up not covering the rent. It was an app that translated foreign road signs into English, and to someone who's travelled, it may sound like a pretty useful tool. But people didn't buy -- and no matter how good they think you're idea is, if they don't buy, you might as well start polishing your LinkedIn profile. A mailing list means interest. A few pre-orders suggest demand. Lots of pre-orders mean you have a market.
"After that, I thought I should probably invest some time finding out what something is worth before I build it," Wilcox said.
When machines learn, they're actually engaged in error minimization: they measure how close their predictions come to a benchmark. Crowdtesting is a kind of error minimizer for humans. The benchmark is what people want; the theoretical model is what you made for them.
There's also a larger truth here. The crowd economy is bringing entrepreneur and consumer closer together. Makers and users are converging to solve acute, niche problems for small, zealous groups of backers. Retail has been disintermediated, to use an obfuscation of finance. You can hook supply up to demand with a hose, and the hose doesn't have to run through Wal-Mart anymore.
While there are many benefits to convergence, it leads to some peculiar relationships. Successful crowdfunders really identify with their customers. They feel a quasi-pathological empathy, an extreme and constant interrogation, asking themselves how buyers will react, testing those suppositions and circling back. Yes, it's codependence, but codependence for business, and it's quite an achievement.
That's because talking to strangers is hard, and it’s harder when you don’t like people all that much. Brogrammers may represent the cool, backslapping, snowboarding face of software engineering, but core of the profession will always be nerds.
If there’s a definition of nerd, it’s someone shy and smart whose feelings of superiority and social unease combine to isolate them. That’s where a lot of hackers are coming from, and telling them to take directions from random strangers as they refine their dreams is a hard slog.
"It's incredibly difficult to talk to customers," Wilcox said. "It's scary. And there are a lot of great techniques like validation and empathic interviewing skills that a developer won't know intuitively."
While they're trying to anticipate and please, they're also trying to figure out how much money you'll pay. "Crowdfunding can answer if people can pay, but not how much," Wilcox said. "You can't A/B test on Kickstarter. If you ask people what they would pay, they don’t know, but if you ask them to get out their credit card, you’re guaranteed to know they would pay that price."
One thing Wilcox noticed was that most demand curves didn't behave like he thought. "With new products that solve compelling problems, price is an indicator of quality, just like social proof or design. It's the opposite of what founders think. They think the lower-priced something is, the more units you will move. But pricing something higher doesn't just raise the revenue per unit, it raises the number of units sold."
What's the takeaway? The official name of all this is customer discovery. It's easier and cheaper than ever before, and it's about as close to the scientific method as business gets. And it's a chance for Silicon Valley and entrepreneurs everywhere to stumble on some really good ideas.
- Selfstarter lets you build your own crowdfunding site and keep the money until you figure out if your idea is even worth the time and stress of applying to Kickstarter, which takes 5 percent of the funds you raise, and also requires a video, which can cost thousands of dollars.
- Pre-order software.
- Squadhelp is one way to brainstorm on a company name. Finding available domain names is rough, but you can crowdsource it. There's poetic justice to that, if you're planning on crowdfunding your idea. identify domain names that are available.
- Mechanical Turk is a TaskRabbit for remote, digital work. One of its stranger uses is to recruit survey takers. You pay people in your target demographic to pretend as though they're customers, and then find out what it would take for them to pay you.
- Unbounce helps A/B test landing pages for customer conversion, and doesn't require programming skills. It essentially enables nontech founders to run their own validation tests.
- Powtoon creates animated videos and presentations for free. You can make an explainer video to show your value proposition. You can produce video quickly without futzing around and spending a bunch of money. It's a barrier eliminator.
- Streak is a gmail add-on that helps you manage how you communicate with potential customers, automate some messages, and track where each target is in the pipeline of acquisition.
- Online post-it notes help you parse what your customers are telling you. Justin groups their problems in clouds. Those clouds help make qualitative assessments quanitative.
- Evernote helps you collect notes and recordings from your customer interviews.
- Greasemonkey: Greasemonkey is a Firefox add-on that allows software developers to customize how any website appears on their laptop. If you thought America wanted to read status updates of all their friends’ purchases, you could test that out by tweaking how Facebook appeared in your browser. (Spoiler: They don't.)
Here are a few tools that Wilcox and his peers are sharing with "wantrepreneurs," those founders in search of a business:
[Image via Where's Waldo]