The short version: A website that loads quickly can still be quietly broken in all the ways that actually matter — dead links, no proper structure for search engines to read, and barely enough content to rank for anything. “It loads fine” tells you almost nothing about whether it’s working. Below: what’s usually broken under the bonnet, why most “you need a rebuild” advice is wrong, and how to tell an honest redesign from an expensive one.
Here’s a conversation I have more often than I’d like.
A business owner tells me their website is fine. It loads quickly, it looks tidy, nobody’s complained. So as far as they can tell, it’s doing its job.
Then we look underneath, and it’s a different story entirely.
The problem is that “it loads fine” is the website equivalent of “the car started this morning.” Good. Glad to hear it. That’s the floor, not the ceiling. Plenty of cars start every morning right up until the day they don’t, and plenty of websites load beautifully while leaking customers you’ll never know you lost — the person who tapped a button that went nowhere, the one who couldn’t find what they wanted, the dozens who never found you in search at all.
You don’t see those people. That’s exactly why this is dangerous.
What “broken underneath” actually looks like
Picture your website as a shopfront on a busy street. Loading speed is whether the lights are on and the window looks good. Important — but it’s the easy part.
Now imagine someone walks up to that lovely-looking shop. The front door’s jammed. Half the aisles dead-end at a wall. And when they do get in, the shelves are nearly empty. They’re not buying anything. They’re leaving, and they’re not coming back.
I reviewed a site not long ago that looked perfectly respectable from the footpath. Loaded fine. The owner genuinely thought it was performing. Under the hood: buttons scattered across the site that led to 404 error pages — dead ends, every one. No real heading structure, so search engines had no idea what any page was actually about. And the whole site — eight pages of it — contained maybe 600 words in total.
Six hundred words. Across eight pages. That’s not a website, that’s a brochure someone left in the rain.
None of that showed up in “does it load.” All of it was costing them customers.
The own-goal almost nobody catches: heading tags
If I had to name the single most common technical mistake on DIY and cheaply-built sites, it’s heading tags. And it’s worth understanding, because it’s a perfect example of something invisible to you that’s screaming at Google.
Think of heading tags — H1, H2, H3 — as job titles in an org chart, not shirt sizes. The H1 is the CEO: there should be one, and it tells search engines what this page is fundamentally about. H2s are the department heads. H3s report to them. Read top to bottom, that hierarchy is how a search engine understands the shape and meaning of your page before it reads a single full sentence.
Now here’s where it goes wrong.
Page builders like Elementor default new headings to H2. So someone building their own site — or someone they hired without the technical know-how — drops in heading after heading, never changes the setting, and ends up with a page that has no CEO at all. Just a room full of middle managers and nobody in charge. Or the opposite: twenty headings all fighting to be the most important, which is its own kind of chaos.
And my favourite, because I’ve heard it so many times: “The H1 looked too big, so I made it an H2.” Or an H3.
That’s fixing the wrong thing entirely. The tag is the job title — “H1” tells search engines this is the most important heading on the page. How big it looks is a separate styling setting that has nothing to do with the tag. So if your main heading looks too big, just shrink it in the styling and leave it as an H1 — a thirty-second job. Changing it to an H3 instead is like demoting someone from Manager to Junior because you wanted their name printed smaller on the staff list. You didn’t fix the text size — you just told everyone they’re no longer in charge.
Get this wrong across a whole site and you’ve made yourself genuinely hard to understand — not to humans, who’ll muddle through, but to the search engines deciding whether to show you to anyone at all.
Why “you just need to optimise it” is sometimes a trap
Let’s talk about the advice you’ll get once you start asking around, because some of it is a trap in the other direction.
Conversion rate optimisation — CRO — is a real and valuable discipline. The good people in it are very good, and worth every cent. They’ll genuinely make your site convert better.
But there’s a flavour of operator who’s discovered that “you’re not optimised” is a magic phrase that sells things. So you’ll hear a brand-new, perfectly sound website needs to be torn down and rebuilt because it “isn’t optimised” — and then watch a generic funnel template get wheeled in with your content awkwardly pasted into someone else’s layout.
That’s backwards. Optimisation is, by definition, what you do to an existing build. It’s the tune-up, not the trade-in. Want a more prominent call-to-action button? Add it. Want your Google reviews displayed, a clearer enquiry path, social proof above the fold? All of that bolts onto the site you already have. None of it requires demolishing the thing and starting over.
If someone’s first answer to “my button isn’t getting clicks” is “you need a whole new website,” keep your hand on your wallet.
So when do you actually rebuild?
Here’s the honest version, and it’s the answer I give clients even when a smaller fix would’ve earned us less.
It depends entirely on the extent of the damage.
If the bones are good — solid structure, modern build, decent speed — then almost everything is a fix, not a rebuild. We optimise, we sharpen, we move on. Cheaper for you, faster for everyone.
But sometimes the old ute simply won’t pass rego. When a site is slow and built on an outdated page builder and has no real technical structure underneath it, you reach a point where patching each problem individually costs more time than building clean. At that point a fresh build on a modern framework isn’t the upsell — it’s genuinely the faster, cheaper path, because we’d spend longer unpicking the old mess than we would doing it properly from scratch.
The skill isn’t in always recommending a rebuild, or never recommending one. It’s in knowing which situation you’re actually in. And that requires someone looking under the bonnet honestly.
What the difference actually looks like
This is a real before-and-after from a site we worked on — same content, same business, just sorted out properly underneath.
Before: an F grade, a 6.8-second Largest Contentful Paint, and 14.1 seconds to fully load — over four megabytes and 250 separate requests. After: an A grade, LCP down to 1.6 seconds, fully loaded in 2.6. Same business. The customer who was leaving at second four now sticks around.
And it’s not just a number on a speed-test tool. Sort the foundations out and the search engines reward it. Here’s organic traffic on a site after we fixed the structural issues holding it back:
The trend that matters: more of the right people finding the site, month over month, without paying for every click.
During my time at Kartia I’ve personally built 70-plus websites and worked on plenty more, so I’ve seen just about every version of “it loads fine, but…” there is. The patterns repeat. The fixes aren’t mysterious — they just take someone who knows where to look.
How to find out what’s actually going on with yours
You can’t fix what you can’t see, and you shouldn’t have to take anyone’s word for it — including mine.
That’s the entire point of a website review. We get under the bonnet of your existing site, find what’s quietly costing you customers — the dead links, the missing structure, the speed problems, the thin content — and tell you plainly what’s worth fixing and what’s worth leaving alone. No pressure to rebuild. If a tune-up is the right call, that’s what we’ll tell you.
And if you’d rather take our findings and sort it yourself? You’re welcome to. The review is yours either way.
Book a website review with Kartia and let’s find out whether your site is actually working — or just loading.
Frequently asked questions
Do I need a website redesign, or can I just optimise the one I have?
It depends on the state of what’s underneath. If your site is structurally sound and built on a modern platform, most issues — speed, calls-to-action, content, search visibility — can be fixed through optimisation without a rebuild. A full redesign is usually only the smart move when a site is slow, built on outdated technology, and lacks proper technical structure all at once, because at that point patching each problem costs more than building clean.
Does a slow website actually affect my Google ranking?
Yes. Page speed and the broader “page experience” are factors Google uses when deciding how to rank you, and slow load times also push visitors to leave before they convert. So a sluggish site loses twice — fewer people find it, and more of the ones who do give up early.
Do multiple H1 tags hurt my SEO?
The bigger problem isn’t the exact number of H1s — it’s having no clear heading hierarchy at all. Many DIY sites have no H1, or headings chosen for how big they look rather than what they mean, which leaves search engines unable to understand the page. Aim for a logical structure: a clear main heading that describes the page, with subheadings nested sensibly beneath it.
What is a website review?
It’s a structured audit of your existing site that identifies what’s holding it back — broken links, weak technical structure, speed issues, thin or missing content — and lays out what’s worth fixing versus what’s fine as-is. It tells you whether you need a redesign or just targeted improvements, before you spend money either way.
