Ask a founder to walk you through their landing page and they’ll have an opinion on every word. Ask them what their app shows when a payment fails, and most will have to go check.
That gap is backwards. A landing page gets seen by a stranger deciding whether to trust you. An error page, an empty state, a “no results found” screen - these get seen by someone who already trusts you enough to be a customer, and is now stuck. That moment carries more weight, not less.
Where the attention actually is
The visitor reading your homepage copy is skimming. The user staring at “Something went wrong. Please try again later.” after a failed transaction is not skimming. They are fully present, mildly anxious, and reading every word you give them. It is one of the highest-attention moments in the entire product experience, and it is usually the one part nobody at the company chose to write.
Most of these screens are the default string a framework ships with, or a line an engineer typed in five seconds to unblock a pull request. Nobody is angry about that - it’s a reasonable place for priorities to land during a sprint. The problem is that it stays that way permanently, because “fix the error copy” never outranks anything else on a roadmap.
What it costs
An empty state that just says “No items” reads as indifference. A failed-payment screen with no next step reads as a company that didn’t think this far. Neither is a technical failure - the feature usually works fine on retry - but the user doesn’t know that, because nothing on the screen told them.
Compare that to an error screen that says what happened, what to do next, and how to reach a human if it keeps happening. Same underlying bug. Completely different signal about who they’re dealing with.
The fix is cheap and nobody does it
This isn’t a redesign. It’s a pass through the ten or fifteen screens a user sees when something isn’t going right - 404, empty cart, empty search results, failed upload, session expired, payment declined - and writing each one like a person who cares wrote it. An afternoon, not a sprint.
The reason it doesn’t happen isn’t difficulty. It’s that these screens don’t show up in a demo, a deck, or a design review. They only show up in production, to a user who is already having a bad five minutes. Which is exactly why they’re worth the afternoon.