I've long been a proponent of validating app ideas before investing too much time in them. This comes from many failed attempts to launch products that I had not validated. It comes from reading the advice you'll find across all the indie hacker forums, VC blogs, and startup books. It comes from working at startups where one wrong move could kill the company.

But maybe you don't need to validate before you build. Maybe you can just build something and launch it and that's all the validation you need. This might sound like the self-justification talk of a developer who just wants to write code, but I'm not a developer. At least, that's not my primary persona. I'm a product person. I'm Head of Product at Pinata, I was a product manager at ClickUp, and I was the only product person on the team at my last startup. Validating ideas by talking to customers, building prototypes, testing, and learning is ingrained in me.

And yet, for many successful projects, I'm learning that the founders did not validate their ideas first. This thought first occurred to me as I read Lenny Rachitsky's seven-part series on building a B2B SaaS business. In part two, he asked the founders he interviewed how they validated their ideas.

A final, and surprisingly common, path to validating an idea is simply to launch it and see how the market responds. This path is most appropriate for two situations:

1. Founders who are getting desperate and need to just try stuff out (e.g. Segment, Loom)

2. Founders who have a crystal clear vision of what needs to be built (e.g. Canva, Databricks)

This was interesting because these weren't just indie hackers he was interviewing. These were the founders of some of the biggest and most interesting tech startups out there. But Lenny limited this path to the two situations above, and that made sense until I was listening to an indie hacker podcast called Startups For The Rest Of Us.

In the episode, Rob Walling speaks with Grant McConnaughey, founder of Postpone. Rob Walling is famous for his "stair-step approach" to building a SaaS, and for teaching indie hackers how to validate their ideas before building them. So, it was no surprise when he asked Grant (whose company is now making six figures and is more than supporting him) how he validated his idea. I expected Grant to respond with some form of "I built a landing page and did customer interviews and this and that." But he didn't.

He told Rob that he didn't validate his idea. He had the idea, he put it out there, and he got a tiny bit of traction. That was enough for him to keep working on it and slowly build momentum. Together, as Grant and Rob discuss this approach, they land on what I would consider a third situation in which founders should probably just launch their idea and see what happens.

If it doesn't take much time to get a working prototype to market, just launch it.

In 2017, I built Graphite Docs in my spare time while learning to code. I didn't validate it because it was already validated in my mind. I was building it for myself, scratching my own itch as Paul Graham likes to call it. I launched it and eventually had tens of thousands of users.

Compare that with my next startup, SimpleID. This was a massive technology-intensive project. My co-founders and I did not simply launch and see what happens. We interviewed potential customers for weeks and quickly learned that our original idea would have fallen flat on its face. From there, we pivoted to a new version of the idea and conducted more interviews. In total, it took over 3 months of interviews and iteration before we actually started building the product. The product took another 3 months to build.

When I look at my side projects, I realize that I probably wasted a lot of time with validation on many of them. These are projects that took a couple of weeks at most to code and launch. Spending time validating by interviewing people is probably not nearly as helpful when you can just launch something and get it into people's hands.

Jason Fried, co-founder of 37Signals, does not believe you can validate a product before it is launched. When he first wrote that, I thought he was wrong. But the more I consider his position, and the more I examine the success of products in the wild, I realize many times validation is just a delay tactic. If you can build quickly and test the market with a real product, you should.

I did that with Static (which is what I'm using to write this blog post, by the way), and I'm doing that again with a new project. I'm fortunate to be able to build apps quickly. The trick, now, will be to follow through with them once they hit the market and get into the hands of potential customers.