DFW Startup Week 2026

I got back Monday from a road trip through Broken Bow, Oklahoma up to Fayetteville, Arkansas, visiting friends and walking around state parks. Tuesday I spent entirely indoors, in auditoriums and conference rooms, on a panel and meeting founders. The two days had nothing in common and I enjoyed both.

It started with a MassChallenge breakfast, pulling together the Dallas community they’ve spent a few years curating. Familiar faces, a few new ones. Startup week mostly works as an excuse to get everybody in town at the same time, and the conversations it produces aren’t the ones anybody planned. Nothing has replaced standing in a room with people.

The Boring Work Got Cheap First

In the afternoon I went to the DFW Startup Week conference at SMU and sat on a VC panel put together by Capital One. The subject was AI and what it’s doing to early-stage company formation, which is the subject of every panel right now.

The first question was about the most interesting thing I’d seen a lean team pull off in the last year. I’m a practical guy, so my answer was probably the most boring one on the stage. What actually changed is that the boring work got cheap first.

Founders are using AI to cut overhead that used to be unavoidable. They’re drafting their own legal documents and running their own basic marketing, work that used to mean paying somebody by the hour before there was any revenue to pay them with. That removes most of the early-stage professional services layer. Starting a company now costs roughly what a serious hobby costs, so the financial penalty for being wrong has dropped, and more people are willing to be wrong.

It feels like an edge and it won’t stay one. Every founder has the same tools and the same falling costs, so the advantage gets competed away as soon as enough people notice it. When the price of finding an answer goes to zero, the scarce input becomes knowing which question was worth answering. That skill hasn’t gotten cheaper, and I can’t think of the mechanism that would make it cheaper, because it mostly comes from having been wrong before and nobody has figured out how to sell that.

The panel went fine. I meant to leave early for the next event and spent an hour and a half standing around talking to people instead. A few of those conversations were great and a couple were alarming, and I’ll write about them separately.

The Judgment Layer Is Uneven

The next stop was the North Texas Startup Demo Day at Richardson IQ. I got there late and missed some of the presentations. What I saw included a couple of apps with real potential and some genuinely cool science projects that won’t have a product for another decade. The ideas were mostly fine, and almost nobody had a clear sense of what they were actually asking an investor for.

Some founders still don’t understand that venture capital is a specific, high-risk product with a narrow set of buyers, rather than a general-purpose form of money. A $20 million company is a good outcome by almost any measure a normal person would apply. To a fund that needs one or two positions to return the whole thing, that same company is a rounding error that consumed a board seat for eight years. Both readings are correct at once. A founder who doesn’t know which room they’re standing in will spend a year pitching people who were never going to say yes, and that year costs time and equity, with the equity part being permanent.

Cheap tools make this worse, which nobody says out loud. When it took $500,000 to find out whether an idea worked, the market killed bad framing early and cheaply. Now a founder can build the product, get it in front of users, and stay alive for two years on almost nothing while still aimed at the wrong buyer. Falling costs extended the runway for good companies and confused ones by the same amount.

High Ceiling, Low Floor

Dallas has a high ceiling and a low floor for founder quality. The ceiling is high because the best people building here are as good as the best people building anywhere, and the cost of living means a failed attempt doesn’t wipe them out. The floor is low because nobody’s having the hard conversation early enough. In a denser city you pick this up from a peer standing next to you, free, six months before it matters. Here you usually learn it when an investor passes, which is later and considerably more expensive.

More capital doesn’t fix that and neither do better tools. Dallas has plenty of capital. What it doesn’t have is enough people who already failed at this and will say so out loud to somebody a few years behind them. The specific hole is first checks written by former operators. The dollar amount matters less than whose name is on it. The person writing that check has personally made the mistake the founder across the table is about to make, and will name it in the first meeting instead of the fifth.

That makes it an exit problem more than a funding problem. The population gets created when a local company sells and five or six people walk away with enough money to write checks and a recent enough memory to write them well. Dallas has cleared the question of whether you can build a company here. The open question is whether the people who win here stay and fund the next group. Starting more funds doesn’t get you there, since a manager who’s never operated doesn’t have the lesson to pass on in the first place.

The Pitch Nobody Here Is Making

The other habit worth dropping is running San Francisco playbooks. Founders here would do better looking at who’s already within driving distance. This region sits on enormous procurement budgets in energy, logistics, hospital systems, and the defense supply chain, and those buyers are far easier to reach from Richardson than from Palo Alto. An AI company selling into industrial or clinical operations has a structural advantage here that it wouldn’t have in California. I almost never hear anybody make that pitch.

The silence is the same gap showing up somewhere else. The tools got cheap and the information got cheap, and what’s still missing is a founder asking which advantages they already hold and aren’t using. It’s the same error as pitching a $20 million business to a power-law fund. In both cases somebody is executing well inside a frame nobody stopped to check, and better tools only help them do it faster.

Which is the argument for a week like this one. The panel content matters a lot less than the ninety minutes after it. Judgment moves between people in person, in unstructured time, usually from somebody who’s been wrong before to somebody who’s about to be. Putting a few hundred of those people in the same buildings for a week is the cheapest method anyone has found for speeding that transfer up. Everything else about starting a company got cheaper this year. The hour and a half in the hallway held its value. I spent it giving away advice and getting a much clearer look at what this city is still missing.