Brian Reath

The business owners who struggle the most are the ones with companies that can only function when they’re in the room. That’s why we run on EOS at Headway, and the most valuable thing it gave us was the discipline to build systems that run the company without the leadership team becoming the bottleneck. But beyond systems, who you hire is arguably more important. With the wrong people in your team, your systems don’t really matter all that much. If people want to flee from responsibility, they’ll find a way to make it happen. Instead, I’d argue we should be building teams that embrace ownership. It’s hard to screen for someone like that in an interview environment, but once something goes wrong, and something always does, you’ll know whether or not you’ve made the right decision. Which dramatically helps the leadership team to the point where they’re no longer the bottleneck.

Every company I talk to right now is racing to figure out how to use AI. Yet many of them are building on top of a data model that was never designed to support it. In my opinion, the infrastructure question is the one that doesn't get enough attention in these conversations. Adding an AI layer on top of a messy, inconsistent data architecture doesn't produce a solid AI-powered product or system within your company. AI multiplies your circumstances. For better or for worse. And if the underlying data model isn’t documented well, siloed, or inconsistent, the output reflects that. The teams that capture the real advantage from AI are the ones that did the infrastructure work upfront, while everyone else was focused on the surface-level adoption race. Implementing a slightly better LLM, trying the latest agent, that sort of thing. I've been in enough product conversations in the last year to know where most companies actually are. And most of us need to sort out our infrastructure first, before scaling with AI. Easier said than done, but entirely worth it.

Most product team job descriptions I’m seeing will look outdated within a few years. The assumption behind most of them is that depth in one lane is the most valuable thing a person can bring: design, engineering, product, pick one and stay there. The teams I'm watching build something interesting are organized differently. They're hiring people who can move across disciplines without losing judgment in any of them, who understand the business problem well enough to evaluate whether the technology is actually solving it, and who can direct AI doing the specialist work rather than compete with it. I've been calling this profile the "specialized generalist" because I haven't found a better name for it yet. Maybe “M-shaped generalists” is another good term? I’m unsure. But hiring these types of people will have outsized returns in the near future.