Segment order matters more than TAM slides

We keep making one sequencing mistake in early-stage companies: trying to sell the “best” customer before the product has earned the right to do that work. The order I trust is usually boring. Start with the customer who feels the pain every day, can say yes without a committee, and will tolerate a rough edge if the outcome is real. Only after that do I move to the segment with better ACV, slower procurement, and more internal stakeholders. I learned that the hard way on one product where we spent 8 weeks chasing the company we wanted on the cap table instead of the company that would actually use the thing every week. My current filter is simple: if a segment needs the product to be “done” before it can be useful, that segment is not first. If the buyer can use it while it still looks unfinished, that segment is first. That usually means owner-led teams, then small ops-heavy teams, then larger companies with a process owner. I am not saying the larger companies are bad. I am saying they are usually later. What changes the order for me is not enthusiasm. It is repetition. I want to see the same use case land 3 times in a row without a founder sitting in the middle of every step. On one product, the signal that moved us from “not yet” to “now” was not revenue. It was when 4 different customers, all in the same segment, asked for the same workflow tweak after using the same core feature for 2 weeks. That told me the problem was real enough to spread, and narrow enough to package. Before that point, I stay disciplined about saying no. We had a stretch where a few bigger logos wanted the product, but only if we added custom reporting, a compliance review, and a workflow exception that would have broken the simpler segment. We passed. Not because those accounts were unimportant, but because they would have changed the build order too early. Every startup has a graveyard of “good” enterprise ideas that arrived before the product had a spine. The trigger for moving up-market is usually some combination of three things: the core workflow is stable, support questions are becoming predictable, and the team can explain the product in one sentence without context. If I cannot watch a new salesperson or operator describe it cleanly after 20 minutes of training, I do not think the product is ready for a bigger buyer. That is not a market issue. It is a sequencing issue. I also think founders over-index on the segment that sounds most impressive. In one category, the first 12 users came from tiny teams with ugly processes. That was perfect. They exposed the sharp edges fast and gave us language we could actually sell. We did not chase the fancier buyer until those users stopped surprising us. Once the product was getting the same objections in the same places, then we had a real reason to move. The signal that matters most is not “someone said they love it.” It is “the same type of customer keeps pulling the product into the same shape.” That is when I know the order can change. I am suspicious of teams that start with the hardest segment because they want the logo. Usually what they are really buying is slower learning. The easier segment is not a consolation prize. It is often the shortest path to knowing what the product actually is. For me, the cleanest sequencing question is this: which customer can create the strongest proof with the least explanation? Start there. Then wait for the product to force its own next market.