If execution keeps getting cheaper, what's actually scarce? The answer is decision quality. For most of my career, the constraint was throughput. Could you design it, build it, and ship it fast enough to matter? That bottleneck has been loosening for years, and the teams I watch now aren't limited by how much they can produce. Instead, they’re limited by knowing what deserves to be produced. Which means the skills that add up are different than they used to be. Understanding the customer deeply. Framing the problem correctly. Knowing what to leave out. Shipping more has stopped being the differentiator. Choosing better is the whole game now.

The tools your internal team uses should closely resemble the tools your customers use. But many companies treat those as separate products with separate standards for some reason. Customer-facing gets the design attention. Internal gets whatever ships. Then a support rep is trying to help someone through a workflow they've never actually seen, using an interface that behaves differently from the one the customer is looking at. Work with any distributor or brokerage, and you see the cost of that difference immediately. But when both sides share the same patterns, your team can reason about a customer's experience without translating. Onboarding gets faster. Support gets more accurate. You made the decision once, and both audiences benefited.

The best outcome of a software discovery process is sometimes learning you shouldn't build anything. That's a difficult thing for a product firm to admit sometimes. But plenty of problems that arrive framed as software problems turn out to be process problems, staffing problems, or something an existing tool already handles adequately. Building custom software is expensive to create and more expensive to maintain - even in the AI era. Every feature becomes something someone owns forever. The question worth asking early is whether the difference between what exists and what you need actually justifies that. Sometimes it clearly does, and that's when the work is worth doing. Sometimes the honest answer saves a client a year and a budget.

We used to spend weeks synthesizing user research on a FigJam board. Dozens of customer interviews, and the work of finding patterns across them was almost entirely manual. Sometimes hours, sometimes weeks, depending on the volume. That specific task has collapsed to literally just hours. Sifting large amounts of qualitative data and surfacing patterns is genuinely one of the things these tools are good at. But the human work still remains. Our attention moved to whether the right patterns got pulled out, which ones actually matter, and what they mean for the product. That's the shape of a use case that works. One specific point of pain, clearly defined, with people still owning the judgment.

Success also means being present for the people waiting at home, alongside the conferences and the time on the road. The goal is to be intentional about travel and make the most of it, spend fewer days away, and make more room for family and the moments that matter.

FOMU beats FOMO. Fear of missing out is about losses that happen when you do nothing. Fear of messing up is about losses that happen when you do something. The second one is the better place to operate from. FOMO is borrowed. It's a reaction to what everyone else appears to be doing, and it drives teams to adopt tools and ship features because a competitor did, not because a customer needed it. FOMU means you're already in motion. You've made a real decision, and perhaps it didn’t work out. But now you have the information necessary, thanks to that failure, to pivot and try again. This time, with a higher likelihood of success. Action produces information. Information you can’t find anywhere else. Because your circumstances are unique to you. No LLM or internet forum can answer for you.  So acknowledge your fears and take action anyway. Nobody else will do it for you.

The grass always looks greener when you're looking at someone else's business from the outside, but every business model comes with its own challenges. As AI makes it easier to automate more of the customer journey, the bigger question is whether we're simplifying for efficiency... or removing the human element that makes the relationship valuable.

Systems thinking and design systems go hand in hand… but there’s a lot more to the story. A design system is bigger than a component library. But the useful part is that nobody has to debate that button again. The same principle applies to far more than UI. One clear ICP means one sales deck, one go-to-market plan, and one set of places to show up. One reliable customer feedback loop means the team isn't reinventing how it learns all the time. These are all small decisions, and small decisions get discounted because they look small in isolation. But when the decision becomes a standard, it compounds through every person and every workflow that follows. That's what I mean by "one decision that makes a thousand." The best systems remove work people never needed to be doing in the first place.

Startup advice assumes information is the bottleneck. It isn't anymore. Every framework and case study a founder could want is findable in an afternoon. What isn't findable is the specific person who already solved the exact problem in front of you, in a similar market, at a similar stage. I've helped run the startup community in Wisconsin for years, and I see this over and over again. Founders show up wanting to talk on stage. What actually makes the event worthwhile is the conversations they have with people after the fact. That's what the community is for. There are things you can’t possibly learn without the help of others. Just the way it is.

The most valuable thing a partner can tell you is that you're wrong. But most service relationships aren't built for that. The incentive runs toward agreement, because agreement keeps the engagement comfortable. So a client proposes a direction, the team has real doubts, but nothing gets said. Six months later everyone can see the problem, and by then the cost of changing course is enormous. I'd rather have the uncomfortable conversation in week two. When that happens, the work gets dramatically better. I think just about everyone knows this, but so few are actually willing to go out of their comfort zone. Eat the frog.

Years ago I worked on a project where 3 teams had each built their own version of the same workflow. Funny enough… none of them knew the others existed. Each had solved the problem competently. Each had made slightly different choices. All three were now in production, and customers were hitting whichever one their entry point happened to route them to. Nobody was careless. Everyone was doing their job well. The failure was communication. That's the argument for systems thinking that I find most convincing, and it has nothing to do with efficiency. It's that individually good decisions, made in isolation, reliably add up to a product nobody designed.

The more complex the problem, the more important it is to slow down, define the problem clearly, and understand what success looks like. AI can dramatically accelerate focused work like analyzing user research and identifying patterns. But without customer context, clear goals, and meaningful metrics, it can't replace the human judgment needed to reach the right outcome.

This couldn’t be more true: “The only way to win is to learn faster than anyone else.” ~Eric Ries Software, business, doesn’t matter. This applies everywhere.  Everyone’s learning. It’s just a question of what they’re learning, and how fast. The easiest way to learn is by taking action, making mistakes, and paying others for information. Find a reasonable way to pull that off over time, and you’ll eclipse everyone else.

High-performing teams are built on mutual care, direct feedback, and a shared commitment to getting better. When people address issues openly instead of whispering behind the scenes, accountability becomes a tool for trust, growth, and collective success.

Wild loophole to have to put 13” laptop away but a 13” iPad with case (heavier) is all good during takeoff and landing.

The worst sports teams I played on weren't short on talent. Someone would say one thing in front of the group, then say something different in a one-on-one afterward.  People whispering in the shadows. Nothing brought into the light. That's what actually breaks a team, and it looks nothing like a skill issue. Without real accountability and a willingness to be fully transparent, you get people playing a political game for their own position. Everyone has something to gain in this world. What separates people is what they do when they could take the easier route and choose the disciplined one instead. That choice tends to tell the rest of someone's story.

My high school colors were red.  College was purple. Indoor football was green. Then I got past all of that and had nothing dictating what I wore. So I picked black, and now everything I buy is black. The remaining decision is just weather and context.  Short sleeve or long. Hoodie or not. Charcoal blazer if it's a business event. White sneaker or leather boot depending on how much I'm dressing it up. I've been drawn to reducing decisions since long before I had a word for it. If you enjoy deciding what to wear, cool. But if you don't, here’s your reminder that you don’t have to.

The easiest thing to do with AI is hand off everything and see what happens. You’ll find the problems with that idea real quick. The tasks people offload first are usually the ones that need the most context, not the least.  What the company is trying to do. What's been tried before. What failure looks like. That's the same context you'd give a person before delegating anything meaningful. With complex problems, the highest-leverage move is to slow down, define the problem, and define what success actually means. Which, ironically enough, is exactly what teams should’ve been doing for humans all along.