You probably do not need a new website

She is stopped at a light on the way to the second job of the morning, and she does the thing she does most weeks: pulls up her own site on her phone. The header still bothers her. The photo behind it is one the old designer picked, the tagline is a sentence she has never said out loud to a customer, and the page looks like it was made a while ago, because it was. She has stared at that header so long it has become the thing she sees instead of the site.

What she does not notice, because she has known the number by heart since before there was a website, is that her phone number is part of an image. It has never been tappable. Nobody has ever pressed it and had their phone start dialing. They would have to memorize it, switch apps, and type it in — or, more likely, back out to the search results and pick the shop underneath hers, whose number lights up blue.

That gap is what this is about. The thing you cannot stop seeing on your own site is almost never the thing losing you work.

you are the only person who has looked at it twice

You have seen your homepage hundreds of times, on a laptop, with good internet and no errand, in the particular mood of someone judging their own storefront. That is not a customer. That has never been a customer.

The visitor sees it once, briefly, needing something specific: is this the kind of shop that handles my problem, do they come to my part of town, and how do I reach them right now. Research on how people use local business sites keeps arriving at the same unglamorous list — the phone number, the service area, the hours, whether you deal with the exact thing they have. Nobody on that list is evaluating your header. They are running an errand, already annoyed, because something in their house is broken.

That asymmetry is the whole problem. You see an artifact with your name on it. They see a vending machine.

So the first honest question is not does my site look current. It is does the errand complete.

the ten-minute version

Here is a diagnostic you can run tonight without buying anything from anyone, including us.

Use your phone, not your laptop. Turn wifi off first. Search for the thing you do plus your town, find yourself in the results, and tap through the way a stranger would — from the search listing, not a bookmark. Then check four things.

Is the phone number tappable? Press it. If your phone does not immediately offer to call, that number is decoration. Sometimes it is baked into an image or a logo, so there is nothing to tap. Sometimes a real link is there, but a chat bubble, a cookie banner, or a sticky header sits invisibly on top and swallows the tap. The button looks right and does nothing. You will never hear this from a customer, because a person who cannot reach you does not write in to say so.

In the first seconds, is it obvious what you do and where you do it? Cover the bottom half of the screen with your hand and read only what is left. If all that is visible is a tagline and a photograph, and the plain sentence naming your trade and your service area lives further down, the errand has not completed. A stranger should be able to answer right shop, right area without scrolling.

Is there a route to you that does not require hunting? Not a contact page four taps deep behind a menu. Not a form asking a homeowner with a leak for their company name and how they heard about you. One obvious route, visible where they land, matching how people actually reach your kind of shop — usually a call or a text.

Does it load on a bad signal? That is what turning wifi off was for. Pages are meaningfully slower on a phone than on the desktop where the owner reviewed them, and while the exact figures move around from study to study, the direction never does: people abandon slow pages, sooner than whoever built the page expects. Your site was approved on office wifi. It gets used in a driveway with one bar and a toddler yelling.

Write down whatever fails. Usually the list comes back short and dull. The number is an image. The trade is not stated where the thumb lands. The contact route is buried. A page drags because of a photograph nobody compressed. None of that is a design problem, and none of it needs a rebuild. Some of it is a fifteen-minute fix by whoever still has the login.

what a rebuild quietly resets

Here is the part that does not come up in the meeting where everyone agrees to start over.

Your current site, however dated it looks to you, has been accumulating something you cannot see and did not deliberately build. Search engines hold a settled record of its addresses, and other corners of the internet point at those addresses: a chamber of commerce listing, a supplier's dealer page, a local write-up from a long time back, a forum answer somebody left for a stranger. That happened slowly, with nobody managing it, and it is worth considerably more than the header.

A rebuild risks all of it, and the mechanism is boring rather than dramatic: page addresses change. The old address for your service page stops existing. The links that pointed at it now point at nothing. The search engine that had a firm opinion about your site has to go form a new one. This is a solved problem — the fix is a careful map from every old address to its nearest new one, made before launch — but only if somebody actually does it, and it is the step that gets skipped when all the attention is on how the new thing looks. Traffic drops, and the cause is almost never the design. It is the plumbing underneath.

Usability researchers have long framed this trade-off as radical redesign versus incremental change. A radical redesign gets you out of a corner you genuinely cannot fix a piece at a time. It also discards everything you had learned about what was working, replacing it with fresh guesses that all go live the same afternoon. Incremental change is slower, less satisfying, and usually safer: you keep whatever earned its place and only touch what did not.

when starting over really is the right call

The honest version of this argument has a limit, and pretending otherwise would just be a different sales pitch. There are real cases where a rebuild is correct.

The phone version does not exist. Not "looks dated on a phone" — genuinely renders as a shrunken desktop page you have to pinch to read. No punch list fixes that; the layout underneath has to change.

You cannot edit it. The maker is unreachable, the login died with a former employee, or every change disappears into a void. A site nobody can change is certain to drift further out of true every season. That is an ownership problem wearing a design costume.

The platform underneath is abandoned. If the thing it runs on stopped getting security updates long ago, you are not maintaining a website. You are maintaining a liability.

The business is not that business anymore. You dropped services, added better ones, moved towns, or now serve a different customer. When the site describes a company that no longer exists, no patch gets it back to true.

The punch list came back long. If nearly everything fails — number, clarity, route, speed — you are not looking at a site with problems. You are looking at problems with a site attached.

Notice what is absent: you are sick of the header. That is real, and it is not nothing. But it is a reason to change a photograph, not a reason to detonate the address structure of a business that has been quietly earning its search position since before you owned a smartphone.

For whatever it is worth from a company that would happily be hired for the rebuild: Catalyst Market keeps a gallery of ready-made, trade-specific templates you can preview in the browser without signing up for anything, and when a rebuild is warranted, a real person takes the brief and produces the finished result. That is the honest description, and it is also the wrong purchase for most people reading this — which is why the diagnostic came first and asks nothing of you.

the point

The reason owners rebuild is rarely the reason customers leave. You are the only person alive with a relationship to your header. Everyone else has a broken water heater and thirty seconds of patience.

Before replacing the whole thing, spend ten minutes being a stranger with a problem. Open your site on your phone, on a bad signal, from a search result, and try to complete the errand. If you can, the header is a matter of taste. If you cannot, what stopped you is usually small, specific, and fixable by Friday — and never the part you have been staring at.