Most technical SEO is checklist work. Crawl the site. Export the errors. Hand the developer 400 fixes ranked by SEMrush severity. Six weeks later, nothing has moved on rankings because nothing on the checklist was actually what was holding the site back.
When two sites in the same category have similar content, similar authority, and similar intent coverage, technical SEO is the distinguishing factor. But only when it is done as diagnosis, not as a tick box exercise. I treat technical SEO the way a doctor treats a patient. Investigate first. Diagnose second. Prioritise by impact. Then prescribe.
Here is the difference, in plain English.
A good doctor looks at your symptoms, your history, and your context. They run the right tests, not all of them. They prioritise the conditions that are actually threatening your health, and they ignore the lab values that look unusual but don't matter for you specifically.
A good technical SEO works the same way. Two sites with identical SEMrush audits can have completely different priority lists. The agencies that hand over a 400 item action plan are doing the equivalent of prescribing every medication on the shelf. It looks thorough. It is actually negligence.
Almost every discovery call I take with a brand or in-house team starts in the same place. A previous audit. A long fix list. Six months of developer time. And rankings that did not move. Sound familiar? Here are the four things I hear over and over.
"Our last agency gave us a 400 item action plan. Dev spent six months working through it. Nothing changed on rankings. I can't tell what we got for the money."
What I hear from in-house SEO leads
"We've been told site speed is a ranking factor for years. Our competitors have terrible PageSpeed scores and they outrank us. Something isn't adding up."
What I hear from marketing directors
"Our SEMrush health score is 96 percent. The site is technically clean. So why does Google still seem to ignore half our pages?"
What I hear from in-house teams
"Every audit we get back looks identical. Same headings, same screenshots, same recommendations. None of them have a single sentence about our business."
What I hear from CMOs and founders
Same five patterns show up in almost every "previous audit" I review. None of them are about missing data. All of them are about how the data was read.
A technical SEO audit that begins with "let me crawl the site" has already lost. The right starting point is the business, the SERP, the competitors, and the revenue model. Crawl data only tells you what is on the page. It doesn't tell you what is actually limiting growth.
An auto generated "high severity" flag in SEMrush or Screaming Frog is a generic guess. A duplicate H1 on a low traffic page is high severity in the tool and irrelevant for the business. A canonical mismatch on the highest revenue category page is medium severity in the tool and a five figure problem in real life. The tool can't tell the difference. The SEO has to.
Health score is a vanity metric. Sites with health scores in the 90s consistently lose to competitors in the 70s because rankings don't care about the score. They care about whether the right pages are crawlable, indexable, and serving the right signals to Google. The rest is decoration.
"Fix 230 missing alt tags" tells a developer nothing about why it matters. "Add alt tags on these 12 high revenue product pages because Image Search drives 8 percent of organic conversions in this category" is a recommendation a developer can actually action with conviction. Same fix. Different value.
Most audits are delivered, billed, and forgotten. The fixes get half implemented. No one checks whether they worked. Technical SEO is a feedback loop, not a one off report. If you don't measure what changed after the work, you don't know what worked.
Six stages. Each one earns its place. None of them are skipped because a checklist would have made it faster.
Before touching the site, I map how the business actually makes money, which pages convert, who the real organic competitors are, and what the SERP looks like for the most commercial terms. Technical SEO disconnected from this context is just noise.
Search Console first (what Google is actually seeing). Server logs second (what Google is actually crawling). Crawl data third (what's on the site). Tool reports last (because they're a summary, not a diagnosis). Most audits get this order backwards.
This is the part the checklist agencies skip. Out of 400 issues a tool will surface, usually 5 to 10 of them are the actual constraint on growth. Finding those 10 is the entire job. The other 390 can wait, or be ignored entirely.
I rank fixes by the revenue they protect or unlock, not by what the tool flagged as critical. If a fix won't move rankings, conversions, or crawl efficiency on pages that matter commercially, it goes to the bottom of the list or off it entirely.
Every recommendation includes the commercial reasoning, the technical implementation, the priority, and the way to verify it worked. No vague "improve site speed." Specific implementation, on specific templates, with specific success criteria.
Four to six weeks after implementation, I re run the diagnostic on the fixed elements. Did the constraint clear? Did rankings respond? Did indexation behave? If something didn't work, that's data. The next prescription gets sharper.
Most technical SEOs hand over a recommendations document and walk away. That works until you realise the developer is booked out for six weeks and the fix is sitting in a backlog. I can ship a meaningful share of the work myself. Edit theme files, implement schema, adjust robots directives, write the canonical logic, build redirect maps, ship the technical content updates. The bottleneck moves.
The point is not that I replace your dev team. The point is that when something simple needs shipping, it ships. When something complex needs explaining, I can speak the language. Technical SEO recommendations only matter if they reach production. Mine usually do.
I'll do a free diagnostic walkthrough of where your real technical constraints sit before you commit to anything.
A UK ecommerce brand with a SEMrush site health score of 94 percent. Three previous agencies had run audits. Every audit said the same things. None of them moved the rankings on the highest revenue category pages.
The brand assumed the technical work was done. The score said healthy. The fix lists had been ticked off. And yet the highest commercial value templates were still losing to competitors with messier sites and lower health scores. The constraint wasn't in the audit. It was in the template logic that the audit had never been pointed at.
Here is how the diagnostic worked:
The brand was outranked on category terms by competitors with weaker brand authority. The audits had never asked why. I started there. What were the winning pages doing structurally that this site wasn't?
Crawl data tells you what is on the site. Search Console and server logs tell you what Google is doing with the site. Two very different questions. The audit reports had only ever answered the first.
Google was burning crawl budget on tens of thousands of low value filtered URLs and barely visiting the high revenue category pages. The SEMrush audits had flagged this as a low severity issue across the site. In reality it was the single biggest growth blocker.
Parameter handling rules to deprioritise low value combinations. Robots directives on facet URLs that didn't need indexing. Internal linking adjustments to flow authority to the templates that mattered. Implementation took the dev team under two weeks.
Crawl stats showed Google spending its budget on the pages that actually drove revenue. Indexation on category pages tightened. Rankings on high commercial intent terms moved within the quarter.
A 94 percent health score had hidden the real problem for over a year. The previous audits weren't wrong about what they reported. They were wrong about what they pointed at. Technical SEO is only useful when it asks the right question of the data. That is the whole job.
Every technical SEO engagement is scoped to your site, platform, and business model, but the building blocks usually include:
If you're a UK brand tired of technical SEO audits that look thorough but don't move the rankings, I'd like to hear what you're working on. The first call is free, and you'll leave with at least three things you can act on whether we work together or not.
BOOK A FREE CONSULTATION 30 MINUTE CALL · NO SALES PITCH · NO OBLIGATION