I'm Betting on Teams, Not Tools

When I evaluate a new tool or product for my stack, I’ve noticed I don’t actually spend that much time on the feature list.

I spend it trying to figure out the team.

Features change. Pricing changes. The interface gets redesigned. What changes much more slowly is how a team thinks. Whether they’re ahead of the industry or just maintaining. Whether they have a clear vision of where software is going, or they’re just shipping what users ask for.

That difference matters more than almost anything else on the comparison spreadsheet.

Reactive isn’t always bad, but it’s not enough on its own

Let me be clear about something: listening to users and being reactive to feature requests is not a flaw. For a lot of products, especially open source tools and infrastructure software, that’s exactly the right approach. Langfuse is a good example. It’s one of the most actively community-driven tools in the AI observability space. They listen, they ship what people need, and the product is better for it.

The question isn’t reactive vs. proactive. It’s whether the team has a point of view beyond what users are asking for today.

The teams I trust most do both. They’re responsive to their users and they have a vision that pulls them forward. They’re not just solving the problems you have. They’re solving the ones you’re about to have.

The teams that shape the industry

There are some teams that just keep showing up a little earlier than everyone else. Not because they’re lucky, but because they have a stronger point of view on where things are going.

Cursor is one of them. They introduced Cloud Agents in late 2025 (btw that’s just 5 months ago, the pace we are experiencing here is actually crazy), then shipped Computer Use for cloud agents, so agents could use the software they created to test changes and demo their work. Then came Automations, and self-hosted cloud agents, which is also a good sign that they are listening closely to what serious teams need. Vision and responsiveness, both.

Linear feels similar. Their Linear Agent announcement wasn’t just “here’s our AI feature.” It was a pretty clear statement about where product development is heading: more context-aware systems, more automation, less manual stitching together of knowledge.

Even Anthropic could be a good example. What gives them an edge in my mind is not just Claude Code. It’s the combination of product, protocol, and research. They introduced MCP, which is now becoming a standard. They keep pushing Claude Code forward at a ridiculous pace. And their research keeps shaping how people across the industry think about agents, safety, and how these systems should actually be used.

That’s the kind of signal I pay attention to. Not just whether a team ships fast, but whether they are helping define the direction of the field while they do it.

Why this matters more than it seems

When you adopt a tool, you’re not just buying what it does today. You’re entering a relationship with the team that builds it.

Their roadmap becomes part of your roadmap. Their bets become your bets. If they’re building toward a future where AI agents are first-class participants in your workflow, and you’re using their product, you’ll get pulled in that direction whether you planned it or not.

And switching costs are real. Integrations, workflows, institutional habits, team knowledge. Changing tools is expensive and disruptive. So you’re not choosing a product for this month. You’re choosing a partner for the next few years.

That changes the calculus entirely.

The question I actually ask

I’ve started evaluating tools with a different set of questions. Not just “what does it do?” but:

• Do they ship proactively or only reactively? Are they solving problems you have, or problems you’re about to have?
• Do they have a clear point of view on their space, or are they just executing a feature backlog?
• When the industry shifts, do they show up early or scramble to catch up?
• Do they have a track record of making good bets?
• Do they publish their thinking? A research blog, a design system post, an engineering deep-dive, anything that shows you how they reason?
• Are the people building it active in the community? Do they write, share, build in public? Can you follow how they think on social, at conferences, in open source?

That last point matters a lot to me. A team that builds in public, that shares the thinking behind their decisions, gives you far more signal than a polished marketing page. Following the people behind a product on Twitter (or X, for me it’s still Twitter :)) ), reading their blog posts, watching their talks… it tells you whether their instincts line up with yours before you’re two years deep into the integration.

The best tools hand you the future

The tools I’ve gotten the most value from aren’t the ones with the longest feature lists. They’re the ones built by teams who had a clear vision of where things were going and built toward it consistently.

When you find a team like that, the product almost doesn’t matter as much. You trust that whatever they ship next will be worth using. And that trust compounds over time.

So when you’re next evaluating a tool, spend less time on the feature matrix. Spend more time asking: do I believe in how these people think?