Most projects arrive as a request: a new website, an app, a redesign, a logo, a feature. The request is not dishonest. It is simply incomplete. It describes the object someone believes they need, not the problem that made the object feel necessary.

That distinction matters. If you build the object and the underlying problem remains, you have spent time and money on a more polished version of the same frustration.

Requests are translations

When a founder says “we need an app,” they are often translating something more specific: customers are dropping off, the team cannot explain the offer, or a workflow is too slow to survive contact with the real world. The app is a guess at the remedy.

When a business says “we need a redesign,” they may mean the brand no longer matches the quality of the work, the site does not convert, or nobody on the team trusts the current pages enough to send people there. A new look can hide those issues. It rarely removes them.

The cost of skipping the problem

Teams skip this step because it feels slower. In practice it is the opposite. A misframed project creates extra rounds, extra features, extra pages, and extra vendors. Each one is trying to compensate for a question that was never answered: what, exactly, should be true when this is finished?

I have watched organizations spend more money coordinating a strategist, a designer, and a developer than it would have cost to keep the original objective intact from the first conversation. Handoffs do not only lose time. They lose the problem.

A better first question

Before asking what to build, ask what is failing. Is it understanding? Trust? Conversion? Operations? A moment in a customer’s day that currently feels awkward or incomplete?

Once that is clear, the solution often becomes smaller, more obvious, and more useful. Sometimes it is still a website. Sometimes it is a prototype. Sometimes it is a targeted rescue. Occasionally it is the decision not to build the thing that was requested.

That last option is not a refusal to help. It is the work.

More insights · Start a project