Do you need a new website, or a better one?

"Do I need a new website" is the question owners bring me when sales are flat and the site is a few years old. It's the wrong first question, because a quote for a new site answers it before anyone has looked at where your visitors actually leave. A website is a path from stranger to call, and people fall off that path at one point, not everywhere. Find the point first. Most of the time it's one element and a small fix. Some of the time the whole design is wrong for your buyer, and then a rebuild is the honest answer. This is how to tell the difference.

Slow sales tell you there's a leak, not where it is

The owner's logic runs like this. Sales are slow, the site is old, so the site must be the problem, so replace the site. The first two steps are fine. The last one skips the only part that matters, which is where on the site people leave. A visitor who closes the tab at the top of your homepage and a visitor who bails at the contact form are two different problems with two different fixes, and a full rebuild replaces the leaking section along with every section that was working.

I look at a website the way I look at any funnel. Pull the numbers for each step, the same way you'd check whether any of your marketing is working, find the biggest drop-off, and fix that one first, because it throttles everything after it. If nine out of ten visitors leave before they reach a form, the form doesn't matter yet. Then I watch recordings of real visitors at that step to see what they did, and I walk the whole path myself, on a desktop and on a phone. That's the diagnosis. It doesn't take long, and it tells you where the leak is before you've spent a dollar.

The leak isn't always where the drop shows up, either. In one teardown I studied, a checkout page was losing buyers and no change to the page itself would fix it. The cause was a step earlier: the page before it was promising something the checkout didn't match. Fix the message upstream and the checkout recovered, no redesign involved. A rebuild quote can't see that. A diagnosis can.

Where the leak usually is, and the small fix that beats a rebuild

More often than not, the leak is at the top. The headline, the line under it, the first button. Every visitor sees that space, and on a phone it's most of what they see, so it carries more weight than anything lower on the page. If a stranger can't tell what you do, who it's for and what to do next in the first glance, they're gone, and nothing further down gets a chance to work.

Which is why the fix is usually smaller than the quote. A headline or button change comes ahead of a full redesign far more often than owners expect, because it's cheap, it's fast, and it changes the one thing every visitor reads. The best description of it I've heard is updating the sign on a store. Nobody rebuilds the store to fix the sign. And if the top passes and people still leave, the leak is lower, and the same method applies: find the step, change one thing, measure. It's the same argument I made about traffic: you don't have a traffic problem, you have a conversion problem, and the conversion problem lives at one step.

The money in a small fix looks modest until you run it. Say your site turns two visitors in a hundred into leads. A better headline takes that to 2.4 in a hundred. That's a fifth more leads from the same traffic, the same ad spend and the same site, for a change that's one line of text. And it does something a rebuild can't promise: it moves the number that decides what you can afford to pay for a visitor in the first place.

Your page's conversion rate sets what you can pay for a click

This is the part no redesign quote mentions. Work out the most a lead can cost you before it stops being worth having. Multiply that by how often your page turns a visitor into a lead. The result is the most you can pay for one click. Say a lead is worth paying up to $100 for. At two leads per hundred visitors, you can pay $2 a click and break even. At 2.4, you can pay $2.40. Same ads, same budget, and now you can afford clicks that a competitor who never fixed their page can't.

So the page's conversion rate isn't a design detail. It's the ceiling on every ad dollar you'll ever spend, and on how far your search traffic goes once it lands. A rebuild that makes the site prettier and leaves that rate where it was has done nothing for the only number that matters, and it cost you real money to do it. That's the test I hold any rebuild to: will it move the rate, or only the look? Asked that way, the question points straight at the cases where a rebuild does move the rate, because the current site is what's holding it down.

When you really do need a new website

There are three. First, the site was built for the wrong buyer. The colors, the photos and the tone all speak to someone your customer isn't, so the stranger's first feeling is "this isn't for people like me." You can't headline your way out of that. When the design is mismatched for the audience, stop testing copy on top of it and fix the design first. The sites that win aren't the most beautiful. They're the ones that feel native to the customer's world.

Second, it fails on a phone. The walk-through I do is desktop and mobile, and if the phone version hides the button or breaks the form, every visitor arriving from an ad or a search on their phone is leaving before your headline gets a chance. No small fix lands on a page nobody can use.

Third, you can't make the small fix. This one is my own position rather than something I can hand you a study for. If changing a headline means paying a developer and waiting a week, the site is blocking the one change that works most often, and you'll stop making changes. A site you can't edit is a site you can't improve. In those three cases a rebuild is the fix. In every other case, a rebuild is an expensive way to avoid finding the leak.

What a fair builder shows you before you pay

My rule is that I diagnose before I quote a rebuild. I'd rather tell an owner they need a free headline change than sell them a site they didn't need, because they come back when they do need one. So before you sign with any builder, me included, ask for three things. Where visitors drop off on the current site, with the numbers. Which single element they'd change first, and why that one. And the top of the proposed homepage, before any design work, so you can see the promise, the proof and the next step and judge whether a stranger would understand it.

A builder who goes straight from "sales are slow" to a price is selling you a website. A builder who shows you the leak is selling you the fix, and sometimes, in the three cases above, the fix is a website. You'll know which, and you'll know before the money moves.

Finding the leak is the first thing I do on every site I touch, and the rebuild question gets answered by what I find. If it's one element, you'll hear that. If it's the whole site, my website design work starts with positioning before pixels, which is the fix for the wrong-buyer problem.

Common questions

How do I know if my website needs a full redesign?

Three signs: the design speaks to a buyer who isn't your customer, it fails on a phone, or you can't change a headline without a developer. If none of those apply, find where visitors drop off first. It's usually one element, and a small fix beats a rebuild.

Can changing one thing really fix a slow website?

Often, yes. Every visitor sees the headline, the line under it and the first button, so a change there moves more than anything lower on the page. Taking two leads per hundred visitors to 2.4 is a fifth more leads from the same traffic, and it raises what you can afford to pay per click.

What should a web designer show me before I pay for a rebuild?

Where visitors leave your current site, with numbers. Which single element they'd change first and why. And the top of the proposed homepage before design work, so you can judge whether a stranger would understand the promise, the proof and the next step. A quote without a diagnosis is a guess.