Building the wrong things faster

A developer on our team built a business dashboard in under an hour. Beautiful graphs. Clean UI. Everyone in the room was impressed. And honestly, it was impressive.

But the data feeding it wasn't flowing in real time. Some of the numbers didn't match. And three teams were still tracking their own versions in separate spreadsheets.

The dashboard was great. The problem was everything underneath it.

This is the pattern I keep seeing in 2026. Building has never been easier. And that's making it dangerously easy to build the wrong things.

Something caught my eye this week.

Linear, the issue tracker that's been steadily eating Jira's lunch, put out a post about their future direction. They said the issue tracker as we know it was built for handoffs. PMs write specs. Engineers pick them up. Work gets passed along a chain.

Their argument: that model is dying. The lines between PM, designer, and engineer are blurring so fast that the handoff doesn't make sense anymore. So Linear is moving toward something different - less "task management" and more "context to execution."

Meanwhile, Atlassian's stock keeps dropping. Lovable, which grew fast on vibe coding, just pivoted to become a general app builder because building random apps alone doesn't sustain revenue.

The tools are changing underneath us. And I don't think most businesses have caught up to what that means.

Here's what I think it means.

Everyone is talking about AI transformation. But what I'm actually seeing on the ground, working inside businesses, is that AI is the easy part. The hard part is the business underneath.

A growing company we work with is hitting this wall right now. When they were small, everything worked because the team could hold it all in their heads. Suppliers managed through personal relationships. Pricing set by feel. Workarounds layered on workarounds, each one reasonable on its own.

Now they're scaling. And all those workarounds are cracking at the same time.

No tool fixes that. You can't automate a process that was never defined. You can't build a system on top of data that three teams track differently. You can't walk in with a playbook from your last client, because every business has its own logic, its own relationships, its own reasons for doing things the way they do.

The only way through is to start small. Build with the people who are going to use it. Solve a real pain point first - not the flashiest one, the most trusted one. Get adoption. Then expand from there.

It's the same pattern that built every great tech product. Ship small, learn, improve, repeat. Not a big bang. Not a transformation roadmap. Just continuous, embedded improvement.

The real shift

My business partner, who is not a developer, built himself a full personal operating system over the weekend. Database, CRM, project management, analytics dashboards, GitHub integration. All wired together with AI tools. Replaced Notion, replaced his project tracker, replaced almost everything.

Took him a few hours.

So yes, building is getting absurdly accessible. But harnessing that power to actually change how a business operates? That requires something the tools don't give you. It requires knowing the people, the processes, the decisions, the business rules. Knowing where technology fits, and where it doesn't. Knowing how to get a team to actually trust and use what you build.

The biggest risk right now isn't being slow to adopt AI.

It's using AI to make the wrong things faster. Prettier dashboards over broken data. Automated workflows over undefined processes. Shiny demos that get applause in meetings but die quietly in the real world.

The businesses that win won't be the ones that built the most. They'll be the ones that built the right things. And kept improving them.

---

Weekly notes from conversations with my business partner, where we share what we're seeing on the ground across clients and the market - not the hype, the real shifts.