Most companies remember to audit their website when something is already broken: traffic dropping, sales falling, doubts about whether the new site actually worked. By then you're paying twice: for the website you built and for the patches that come next.
A proper website audit is not an autopsy. It's a blood test before surgery. And the right time to ask for one is almost never the time most people ask for it.
What a website audit really is (beyond the Wikipedia definition)
If you Google the term you'll find dozens of definitions repeating the same line: "technical review of a website". That definition is correct and, at the same time, useless. It falls short.
A serious website audit starts before looking at the site. It starts by understanding the market the site competes in, because without that context every technical recommendation that follows is opinion dressed up as method. Once you have that context, the diagnosis unfolds across six layers that influence each other:
- Strategic layer (market, competition and ideal customer): where real demand sits, who your customer is, how they search, what your competitors are doing, what they have that you don't, and the other way round. This layer is what gives meaning to all the others. Without it, you optimise for an imaginary market.
- Technical layer: indexability, speed, server errors, redirects, mobile-friendliness, schema, security. What Google sees when it crawls.
- SEO layer: current rankings, keyword opportunities, cannibalisation, structure, internal linking, authority.
- Content layer: how your published content covers the real searches your audience is doing, quality, depth, intent. What you decide to say and how you say it.
- UX and conversion layer: the paths users take to contact, forms, friction, visual hierarchy, value proposition clarity. What happens once the visitor is in.
- AI search visibility layer (GEO/AEO): whether ChatGPT, Perplexity and Gemini cite you when someone asks about what you offer. The new layer, the one almost nobody is looking at.
Running an "SEO audit" without looking at UX is watching half the film. Running a "technical audit" without understanding the market is checking a car's engine without knowing where you're going. A real website audit is the sum. If an agency offers you only one layer, or skips the strategic context, they're selling you a piece, not the diagnosis.
What a website audit is actually for
The textbook answer is "to detect problems". True, but incomplete. A well-framed audit does four things, and only one of them is about finding errors.
1. It puts a number on the hidden cost of not touching anything.
Most websites have between five and twenty invisible money leaks: one page cannibalising another and so neither ranks, a JavaScript that blocks half the catalogue from being indexed, a contact form that stopped sending emails three months ago and nobody noticed, broken redirects inherited from an old migration. Each one is small. Added up, they're usually the reason you're spending on Ads to paper over what isn't converting organically.
2. It tells you what to invest in, in what order.
A good audit does not deliver a list of a hundred things to fix. It delivers a list of a hundred things prioritised by impact and cost. What needs fixing this week, what next month, what should be left forever because it isn't worth it. Without that prioritisation, any SEO or development budget is a lottery.
3. It helps you decide whether to redesign or not.
This is the awkward conversation almost nobody has. Sometimes the problem is not that the website looks ugly: it's that the URL structure is messy, there's a missing landing page per service, or the blog is abandoned. In those cases, redesigning the home page is like painting a damp wall. Wasted money. And sometimes, yes, you do need to redesign the whole thing. But that decision isn't taken on intuition or "this site is three years old already". It's taken with data.
4. It creates a baseline you can measure against.
After any change (redesign, migration, campaign, new service) you need to be able to compare against something. If you didn't audit before, you have no starting point. Any later result is opinion.
Before or after building the site? The most expensive trap in digital marketing
This is where the common sense you've been told is exactly what's costing you money.
The sequence most companies follow is: (1) decide to build a new site, (2) commission it, (3) launch it, (4) at some point hire SEO. By the time they reach step 4, they're already inheriting problems that could have been avoided at step 1: improvised URL structure, duplicate content, redirects done wrong from the old site, page hierarchy that doesn't match how the customer actually searches. Each of those problems has a fix, but the fix costs time and money. And it was all preventable.
The right sequence is the inverse. Three scenarios, three moments:
If your current site has been running for years and you're thinking about redesigning.
Audit before touching anything. You need to know what to protect (which pages bring traffic today, which keywords bring leads, which backlinks you can't afford to lose) and what to rethink from zero. Without that prior snapshot, a redesign can throw away years of SEO work. We've seen it happen so many times we lost count.
If you're launching a brand new website from scratch.
Audit at the end of the design process, before launch, not after. When the site is built but not yet public. At that point you can fix structural errors at zero cost (changing a URL, adding schema, adjusting internal linking) that after launch are no longer free: any change implies redirects, communication to Google and risk.
If you're migrating a site (technology change, domain change, hosting change).
Double audit: before to map what exists and plan redirects one by one, and after to verify nothing broke. Migrations are the moment when most organic traffic is lost through carelessness. Not an exaggeration: we're talking about 40-60% drops in sites that were working fine.
We applied this to our own house. When we migrated geckostudio.es from WordPress to Astro this year, we audited the 781 URLs of the site before touching anything. We mapped which were still alive, which needed redirecting, which no longer made sense. The day of the launch, rankings held. That kind of result isn't luck: it's having done the homework in the right order.
The most expensive mistake in digital marketing isn't skipping the audit. It's doing it late.
How we run a website audit at Gecko
There are as many ways to audit as there are agencies, and many end with a hundred-page PDF nobody reads. Our process is this one, in six phases, designed to deliver something actionable instead of decorative. The difference from most agencies is in how it starts: we don't open Screaming Frog on day one.
Phase 1 · Strategic study: market, customer and competition.
Before touching the site we audit the context the site lives in. Three fronts:
- Market analysis: real demand size, sector trends, seasonality, segments where there's room and segments that are saturated. If you work in a sector with strong seasonality (tourism, marine, holiday rentals), decisions emerge here that change everything else.
- Real buyer personas: who your customers are, how they search, what pains you solve, what objections they have. Not the template personas from a marketing manual: the ones built by crossing your brief with real Google Analytics 4 data and search behaviour.
- 360 competition analysis: which keywords bring them traffic, what content works for them, which backlinks they have and you don't, what they're doing well and badly. We cross-check it with Ahrefs and Search Console data so it's diagnosis, not impression. This is where the real gaps come up: what competitors aren't doing and you could occupy.
This phase is what gives meaning to the next five. Skipping it is what turns most audits into technical lists the client reads and stores in a drawer. With context, the list becomes a plan.
Phase 2 · Technical crawl and real data.
Now, yes, we open the site from the inside. We crawl the entire site with Screaming Frog, connect Google Search Console, Google Analytics 4 and, when applicable, Ahrefs and Bing Webmaster Tools. We pull the complete URL inventory, 4xx/5xx errors, redirects, structured data, real Core Web Vitals and current rankings. This phase is not opinion, it's data.
Phase 3 · Content and architecture analysis.
We cross what you have published with what your audience actually searches for (we know this from Phase 1). We detect cannibalisation (two of your pages fighting for the same keyword, both losing), content gaps (searches you should have a page for and don't) and architecture problems (important pages three clicks away from the home, empty pages receiving links). This is where the discoveries that surprise the client most usually appear.
Phase 4 · UX and conversion funnels.
We walk the site as a real user, on mobile and on desktop. We measure forms, map the paths to contact, identify friction points. If you have Hotjar, Clarity or similar, we cross-check with it. The question we answer is concrete: out of the traffic that already arrives, why isn't more of it converting?
Phase 5 · AI search visibility (GEO/AEO).
The new layer. We test real queries on ChatGPT, Perplexity, Gemini and Bing Copilot related to your services and sector, and we check whether you are cited, whether competitors are cited, or whether nobody appears. We review schema, FAQ, mentions, citability. Almost no agency is looking at this yet. And almost all search is moving towards it.
Phase 6 · Prioritised report and action plan.
The deliverable is not a hundred-page PDF. It's a report covering the things that matter, ordered by impact and cost, with a clear action plan: what to do this week, what next month, what later. We close with a walkthrough call because a report without context is wet paper.
Average time for the complete process: between two and four weeks, depending on site size and the depth of the strategic study. The first partial deliverable lands in week one.
Frequently asked questions about website audits
What exactly is a website audit?
It's a complete diagnosis of a site's state across six layers: strategic (market, customer, competition), technical, SEO, content, UX and conversion, and AI search visibility. It isn't a standalone technical review or a PDF of errors: it's the starting point for deciding where to invest in your site over the next several months.
What's the difference between a website audit and an SEO audit?
An SEO audit looks only at organic rankings. A website audit includes SEO but adds UX, conversion, technical, content, market context and visibility on AI engines like ChatGPT or Perplexity. The SEO audit is one piece inside the website audit, not a synonym.
How long does a professional website audit take?
Between two and four weeks depending on site size and the depth of the initial strategic study. Small sites (fewer than 50 URLs) and well-known sectors can close in two weeks. Large sites with catalogues, multilingual setups or highly competitive industries take three or four.
When is a good time to audit a site?
Before redesigning, before migrating, before investing in Ads or SEO, when organic traffic drops without a clear cause, when the site is more than two years old without a review, or when you inherit the site from another agency. If you're going to take a big decision about your site in the coming months, audit it first.
Do you need to audit a brand new site?
Yes, but at the end of the design process and before launch, not after. At that point structural errors (URLs, schema, linking) can be fixed at zero cost — after launch they require redirects and risk. If the site has just launched and has no traffic yet, it's better to wait three to six months for real data before auditing again.
Is a website audit useful for an online shop?
It is, and it usually adds more value than on a corporate site. In e-commerce, technical issues (product cannibalisation, indexable filters, broken pagination, incomplete product schema) have direct sales impact. A well-run audit on a shop with a hundred products can recover 15-30% of organic traffic that was being missed.
When asking for a website audit makes sense (and when it doesn't)
Not every project needs a full audit. An audit makes sense when it answers a concrete question and enables a decision.
It makes sense to ask for a website audit if:
- You're thinking about redesigning or migrating and haven't touched anything yet.
- Your organic traffic has dropped without a clear cause in recent months.
- Your site is more than two years old and has never had a serious review.
- You're about to invest in Ads, SEO or a major campaign and you need to know where the traffic will land.
- You're considering positioning yourself in AI-powered search and don't know if your site is ready.
- You've just inherited the site from another agency or developer and need to know what you've got on your hands.
It doesn't make sense yet if:
- Your site is new and doesn't have traffic or data yet. Better invest those months in publishing content and audit again in six months.
- You're about to pivot the entire business. Audit after the pivot, not before.
- Your only doubt is whether the site "looks pretty". That's not an audit, it's a visual review and it takes an hour.
If your case fits the first group, the most profitable thing you can do this quarter is to know with data what's happening on your site before spending another euro on it. We do it at Gecko, we did it on our own site, and it tends to save more money than it costs.
