The Empty-State Opportunity Most Sites Waste

There’s a category of page that almost every site has and almost no one designs: the page with nothing on it. The dashboard of a brand-new account. The search that matched nothing. The empty cart. The inbox before the first message, the project list before the first project, the report before there’s any data. Designers call these empty states, and the name itself hints at the problem. They’re defined by what’s absent, so they get treated as an absence, a nothing, a screen you pass through on the way to the real product.

But look at who’s standing in front of them. The person seeing an empty dashboard is a user who signed up minutes ago, which is to say the most fragile user you will ever have. They did the thing every landing page begs visitors to do. They clicked the button, filled the form, maybe verified an email. Their reward is a blank screen, sometimes with the words “No data yet.” All the persuasion that got them here, the careful headline, the testimonials, the promise of what the product would do for them, hands off to a page that says, in effect, nothing to see.

This is a strange allocation of effort when you think about it. Teams will argue for a week about the shade of the signup button and then ship an empty state that a framework generated. The economics are backwards. The marketing page is competing for the attention of strangers, most of whom were never going to convert. The empty state has the full attention of someone who already converted and is now deciding, in the next ninety seconds, whether the decision was right. Effort spent there acts on the most persuadable audience you have, at the most decisive moment, and it’s the page that gets none.

The reason is visibility. The people who build a product stopped seeing its empty states long ago. Their accounts are full of test data. Their searches match things. Every demo runs against a seeded database, because a demo of blank screens would be embarrassing, which should itself be a clue. The empty state is the one part of the product that its makers structurally never experience, and the one part that every single new user experiences first. It’s a blind spot with perfect coverage.

What should an empty state do instead? The principle is that an empty state is not a report of absence, it’s the start of a path. The screen knows exactly where the user is: at the beginning. So it should do what a good beginning does. Say what will appear here once things are working, so the user can picture the payoff. Show the single next action that gets them there, one button, not a menu of options, because a new user given six choices researches none of them. And where possible, shrink the distance to the first real result, with sample data, an import, a template, anything that lets the user see the product being the product before they’ve done all the setup themselves. The empty state’s job is to get itself deleted as fast as possible.

The zero-results search deserves its own mention because it’s the empty state with the clearest money attached. Someone who searched your store or your docs just told you, in their own words, exactly what they wanted. A page that answers “no results found” and stops has taken the strongest intent signal a visitor ever emits and returned nothing. It could loosen the query and show near matches. It could suggest the closest category. It could at minimum keep the search box filled and prominent so trying again is effortless. Instead the standard treatment is a dead end and a sad icon, presented to precisely the visitor who came closest to converting.

I started paying attention to this through building GazeSite, in an oblique way. The tool audits pages as a stranger sees them, screenshot and markup together, and grades whether a skeptic would know what to do next. Pages fail that test most spectacularly when they have the least on them, because with no content there’s nothing to infer from; whatever the page explicitly says is all the guidance the visitor gets. A rich page can survive mediocre design because the content carries it. An empty page is design and nothing else. It’s the purest test of whether anyone thought about the person standing in front of it.

Here’s the exercise, then. Make a list of every screen in your product that can be empty, new account, no results, cleared cart, completed queue, expired data, and go look at each one the way a first-day user does. For each, ask one question: does this screen tell the person what to do next, or does it just describe the void? You’ll likely find that your product’s real onboarding isn’t the tour or the checklist you built. It’s this scattered collection of blank screens nobody owns. They were always the first thing new users saw. The only question is whether anyone on your side has seen them lately.