Building Got Easier. Selling Didn't.
AI makes it easier to build software. It does not make customers care.
That distinction is going to matter a lot over the next few years, and I don't think the market has fully priced it in yet.
Small teams can now ship products that would have taken much larger teams not long ago. Solo operators can prototype ideas in a weekend. Domain experts can turn a workflow into a tool before a traditional product team finishes its first planning doc. More people should be able to make things, and now more of them can.
But lower creation costs create a new problem: the market fills up with technically impressive products nobody asked for.
A working product is not a painful problem. A slick demo is not urgency.
I have spent 15 years in GTM roles across companies like Square, Mailchimp, and Zapier. The products that struggled were rarely the ones that didn't work. They were the ones that worked fine but couldn't answer a simple question: why would someone change what they're doing right now to use this?
I watched this play out at Mailchimp when the SMS space started heating up. Competitors had solid products — good UX, fair pricing, reasonable feature parity. But email marketers didn't wake up desperate to add another channel. The teams who gained traction understood that the real barrier wasn't capability. It was workflow inertia and the fear of breaking something that was already working. Technical quality was table stakes. Distribution and narrative were the actual work.
That shapes how I think about early-stage investing now. I'm not a VC. I'm an operator who makes concentrated bets alongside my day job, looking for founders who understand their customer from lived experience rather than research. When I look at a deal, the product is almost never the question. The question is whether this team can get someone to change their behavior.
A few filters I keep coming back to:
- Is the problem painful enough that someone is already spending money or time on a workaround?
- Does the founder understand the customer from lived experience, or just from research?
- Is there a believable path to reach buyers without novelty doing the heavy lifting?
The best AI companies will still need taste and distribution. When building gets cheap, the middle fills with products that are fine but forgettable. Novelty wears off. What remains is whether the thing actually changes someone's day.
The question shifts from "can this be built?" to "who needs this badly enough to pay for it, and why will they trust this team?"
That is a harder question. It is also the right one.
