How do YOU test a startup / product hypothesis?
I'm curious how you guys go about proving (or disproving) a hypothesis? What have you found is most effective in regards to time and quality of testing?
I'm curious how you guys go about proving (or disproving) a hypothesis? What have you found is most effective in regards to time and quality of testing?
Sometimes I feel like this is more of an art than a science.
Most great ideas are great precisely because they are disguised as bad ideas - and if you were to do market research on them or ask people about them you might get negative results.
Thats why best products where built not for the market, but built based on the standards of the leaders of the companies, and people just found that they liked it too.
If you ask Elon musk how much market research he does, he will say zero. He builds products he likes with the idea that hopefully (he knows they will) the market likes it too.
In fact it might be entirely counter productive to build products that you don't like, just because you've done some studies and that suggest that someone else will like it.
People don't know what the fuck they want most of the time, they think they do but they really don't.
Ford: If I would have done market research on the transport market, they would have asked for a faster horse.
Its your job as a creator to make something you know people will love even if they don't realize it. and have the tenacity to work through it even when you know its right, but everyone thinks your wrong.
This is very typical, and why I said its like an art is because you could actually be wrong, and be tenaciously wrong. And from the outside or objectively its hard to tell which you are - you simply don't know - you have to feel strongly about it, and if you feel strongly about it there must be some grain of value there and you could pivot to it once you realize what it is.
so really, just work on projects you like and feel are important, that way if your wrong, well at least you spent your life doing something you liked to do.
Very well said. I think it's a point of view that should be talked about more.
Excellent advice.
This comment was deleted 7 years ago
I'm glad someone is asking these questions. When you start, there is like a dogmatic approach to testing, which consists, basically on these steps:
However, imagine you pitch to 10 friends and 1 gets it and likes your idea. Is that enough to keep going? Imagine you have an MVP and 1000 people sign up, is that enough? What if 10.000.000 visited your website and only 1000 registered.
The biggest issue is that most people following a lean approach, don't have clear expectations before performing a test. What are your go's/no-go's? And these should be established before you start with any testing, because it will make clear what you are actually trying to test.
I am in the process of validating my ideas, however, I am in the scientific hardware business, and thus time-scales, etc. are different. I do believe, however, that the best feedback quality is in interviews after interaction with my MVP. They use, they make mistakes, they see results, then I check. And normally the interviews focus on clear, specific aspects, such as the software interaction, the lightning, the sample preparation, etc. Building an MVP in our case was a leap of faith, since we have no clue if there is a market for it or not. It is new tech, meaning that no one knows how well it performs, if the results are better than with existing devices. Thus, we need to prove it on a case by case basis and start building the market ourselves.
If you are thinking about software, you have the advantage that your MVP can be very small. This means, you can build a simple solution to the problem you have found, and then you can always start by siting next to a user, and see how they use your tool, or meeting/calling them after 1 week to see what they did, etc. You can also grow the feedback base by asking to respond to questionnaires, etc.
If you are planning a user-centric approach, then you should focus on your user when testing. At the beginning, nothing will be better than real user feedback. Validating the MVP, I believe, should be done quickly and at the beginning, because then you will have to validate the business model.
Depends on the idea. If it’s an issue I’m encountering and thought “I wish I had {insert product} to solve my problem”, then I may validate with some friends before deciding on moving forward with the idea.