I was chatting with a friend the other day about his current and past App businesses and I wanted to share parts of our conversation with a broader audience. I’ve made a few edits to obfuscate his identity but the spirit of the conversation remains intact.
Andrew: Random question, what’s the core hypothesis you set out to test when you started with your app?
Friend: I’m not sure I can phrase it quite like that. I was testing whether people would engage in substantial ecommerce from their phone. Whether spending less time shopping was worth the tradeoffs.
Andrew: Was there another model – aside from hypothesis testing – that you thought about when you were in the very early stages?
Friend:
I’m looking for a “boom”
Andrew: Go on.
Friend: The first version of anything is so full of hypotheses. I think that’s the issue with these concepts and with the Lean Startup.
It’s the idea and the design and the marketing and the execution. I tried to fix some variables by carefully explaining the details to people of how it worked, but even still it’s a hot mess.
I’m looking for the market to give a strong indication that they like the idea (i.e. a “boom”).
Lean Startup is for funded companies, by the way. That’s my conclusion. It takes lots and lots of money to do these kinds of experiments.
Eric Reis (author of the Lean Startup) used to do it acquire new users for $10 / day, but the market has changed. It’s now $1000 / day. At least for apps. I can’t drive 100 people a day to my app for $10. That’s simply not doable.
So all the experiments fail to yield conclusions.
Andrew: Right – so do you think a different/more specific/smaller/etc hypothesis could yield clearer conclusions? Or is the entire hypothesis-test model no good for these circumstances?
Friend: I don’t really know honestly. I left the process with a lot of learnings but no clear picture.
The “boom” idea is that you put out a whole package as one big fat unarticulated hypothesis.
And you try your best to learn from it by seeing what people do and talking to people and using it and whatever. It’s messy as shit.
Andrew: Yeah, sounds like a mine-field the way you’re describing it… except that you WANT to accidentally step on a mine.
Friend: Yeah, maybe you can be more disciplined about it.
I think the reality is that if you really want to be disciplined, you start with very high quality interviews and slowly work your way towards software.
You don’t start with software at all. Unless you yourself are the customer, in which case you just build what you want, and the only hypotheses are:
- whether you are unique
- whether you can reach people like you
- whether the pain is enough for people to care
- and, whether they’ll pay for it
Andrew: Right – which might be why we hear that advice so often: build for yourself first, because it limits a huge range of hypotheses
Bottom Line: I don’t think there’s a right or wrong answer here, but I wanted to get this out there and have a conversation with folks.
Are you a hypothesis-tester or a boom-seeker? Let me know in the comments.