A SEO audit for businesses is not a 120-page PDF with red traffic lights. It is a diagnostic process that, if done well, ends with a short list of decisions that someone has to approve and execute. And that is where almost everything breaks down.
I have seen technically impeccable audits that ended up in a drawer because the consultant delivered 340 recommendations without prioritizing them and without talking to the people who had to implement them. I have also seen companies spend €6,000 on an audit for an ecommerce site with monthly revenue of €40,000 and discover that 70% of the problems were the result of two poorly built templates. Context, always.
Why 80% of commissioned audits are never executed
Based on recurring conversations in the industry and agency documentation, the problem is usually not the detection of errors, but translating them into actionable tasks. An audit that says “improve Core Web Vitals” is useless. One that says “the LCP of the category template is 4.8s due to a 1.2 MB hero image served uncompressed and without a modern format; switching to WebP and removing the slider reduces it by 2.1s” is useful.
The second reason for abandonment is political. In mid-sized companies, the SEO team does not have access to the developer. The audit reaches marketing, marketing forwards it to IT, IT ignores it because it has a closed quarterly roadmap. Without a prioritization plan based on effort/impact and without a named owner, the audit dies.
Real cases where the audit changed something (and why)
Case 1: the company with 14,000 URLs and 900 visits per month
An industrial-sector client with a product catalog had 14,000 URLs indexed, of which 11,200 were filter variants generated by the CMS. Google was crawling 40,000 URLs daily and only visited each product page once every three weeks. The audit was not about content: it was about crawl budget.
The solution was to block the filter parameters in robots.txt, add a canonical to the parent category, and create static pages for the eight filter combinations that did generate traffic. In three months, crawling of product pages went from one visit every 21 days to one every 3 days, and product impressions rose without touching a single description.
What I learned: in projects with large catalogs, auditing crawling and rendering first saves weeks. Reviewing titles and meta descriptions beforehand is a waste of time.
Case 2: the structured data error that nobody was looking at
A furniture store implemented Product structured data without a valid offers field. Search Console showed 2,300 “missing value” errors. The team had assumed it was a plugin problem. It was not: the price field was generated with a shortcode that left the value empty on products with variable pricing.
The diagnosis required cross-referencing the template’s source code with the URL inspection API. The fix was a conditional in the shortcode. The result was not spectacular, but it was real: pages with a visible price began showing the product rich result with availability, and the CTR of those URLs improved measurably in Search Console the following month.
This does not work for every business. If you sell services or products available on request, Product structured data does not apply, and forcing it generates inconsistent signals that Google may ignore.
Case 3: the migration that was done twice
An online training company changed domains without a prior audit. They lost 60% of their traffic in two weeks. The subsequent audit revealed three things: the 301 redirects were done page-to-page for only 200 URLs (the rest fell through to the homepage), internal links still pointed to the old domain in the menu, and the new sitemap had been uploaded but the old one was still accessible and had crawl priority.
The lesson from this case is that audits for businesses should be done before a migration, not after. A well-made migration checklist costs less than recovering lost rankings.
How I structure an audit that gets executed
I do not start with tools. I start with a 45-minute meeting with the decision-maker and the implementer. Specific questions: who touches the template? who approves changes in production? is there a maintenance window? what is measured as success: leads, sales, calls?
Then come four blocks, in this order:
1. Crawling and indexing (without this, everything else is noise)
I review robots.txt, sitemaps, status codes of the main templates, canonicals, and parameters. I use server logs when the client has them. A figure that almost nobody calculates: how many unique URLs Google crawls per day versus how many the site has. If the ratio is bad, there is an architecture problem or a problem of infinite URL generation.
2. Templates and rendering
I do not audit individual URLs. I audit templates. If the category template has a problem, it is multiplied by 200. This includes titles, meta descriptions, H1s, structured data and, above all, what Googlebot sees after executing JavaScript. An SPA with injected content may look correct in the browser and be empty for the crawler.
3. Content and intent
I cross-reference Search Console (queries with impressions and low CTR) with the pages that receive them. I look for cannibalizations where two URLs compete for the same query and one of them should redirect or be merged. This step usually generates more impact than any technical adjustment in companies with abandoned blogs.
4. Authority and links
I review the link profile without obsessing over third-party metrics. I am interested in detecting broken internal links, orphan pages, and whether the internal linking architecture distributes authority toward the pages that convert. In B2B projects, poorly planned internal linking is a more frequent problem than the backlink profile.
What an SEO audit CANNOT do
It cannot guarantee rankings. It cannot compensate for a bad product or a website with prices outside the market. It cannot fix a team that does not have time to implement. And it cannot replace analytics: if there are no properly configured conversion events, the audit will tell you there is traffic but not whether it serves any purpose.
Nor does it detect business problems. I have seen technically perfect audits on websites that had lost traffic because their sector had contracted or because a competitor had lowered prices. SEO does not explain everything, and an honest auditor says so before invoicing.
Deliverables that actually get used
My standard delivery has three pieces: a findings document grouped by impact (not by technical category), a spreadsheet with tasks prioritized by effort/impact and an assigned owner, and a 60-minute session with the development team to review the first ten tasks. If the client does not have a technical team, the audit includes a proposed implementation scope with a third party, not a wish list.
According to Google’s official documentation on crawling and indexing, changes take time to be reflected; that is why follow-up should be planned at 30, 60 and 90 days with specific metrics: valid indexed URLs, impressions by template, and average CTR by page group.
A SEO audit for businesses is, at bottom, an exercise in translation: converting technical signals into business decisions with names and surnames. If it does not end in a plan with owners and dates, it is not an audit; it is a report.