The engagement ends with a hundred-page PDF
The engagement ends with a hundred-page PDF and a list of issues sorted by severity. Someone has to build all of it, that someone was never scoped, and the list is stale within a quarter.
Services
Technical SEO
The pages that carry your pipeline get found, indexed and kept there. We find what is blocking them, fix it in the build ourselves, and re-crawl every month, so a fix stays fixed and a regression shows up in the next report instead of in next year's traffic.
Book a Discovery Call
Found and Fixed
Why Most Technical Audits Change Nothing
The ProblemStandard Agency
A tool flags 4,000 issues that mostly do not matter
A tool flags four thousand issues and most of them do not matter. Without someone who can tell a real indexing problem from a warning, the budget goes on the wrong four thousandth.
The next CMS change reintroduces half the fixes
The fixes ship, the audit closes, and the next CMS change quietly reintroduces half of them. Nobody notices until traffic falls two quarters later, and by then it is a board question.
The Solution
The Legency Approach
The Legency ApproachA finding becomes a change on the site
We build the site, so a finding turns into a change we ship rather than a ticket for a team who were never briefed. Nothing gets recommended that we are not willing to build ourselves.
The pages that carry your pipeline come first
Findings are ordered by the traffic and revenue behind the affected pages. A canonical error on a page nobody visits stays at the bottom of the list where it belongs.
A regression shows up in next month’s report
The site gets crawled on the same cadence as everything else we measure, so a regression shows up in the next report rather than in next year’s audit, while it is still cheap to reverse.
The Work
Four Things That Decide Whether a Page Is Found
Crawl and Index
A page an engine never reaches cannot rank, convert or be cited. Canonicals, redirects, sitemaps, robots rules, and the junk URLs that eat a crawl budget before the real pages get seen.
Speed and Core Web Vitals
A slow page loses the visit before the form loads and ranks below faster ones. Field data first, lab data second: image weight, font loading, third-party scripts and layout shift, measured on the pages that carry traffic.
Structure and Schema
The markup that tells an engine what a page is and what to quote from it. One H1, a heading order that reflects the page, main content an engine can separate from the navigation, and schema per page type that describes what is there.
Machine Readability
The layer that decides whether an AI assistant can use the page at all: clean markup, a resolvable llms.txt, no content locked behind script, and internal links that resolve.
What It Moves
Google Clicks Up 90% in Four Months on One Site
K2 Relocate ranked for nothing commercial at the start of the engagement. The work was unglamorous: structure, internal linking, the markup that tells an engine what a page is, and the crawl waste that was burning attention on URLs nobody should ever see.
Four months later it is the fastest-growing site in a six-site estate, and its average position across the site reached 14.7, the best it has recorded. The board saw that line on a monthly report.
Nothing about that is a redesign. The pages largely say what they said in April.
Google Search Console, April to July 2026. Clicks to one site in the K2 Group estate. Axis starts at zero.
Crawl Waste
Most Estates Spend Their Crawl on Nothing
An engine gives your site a finite amount of attention. On a large site most of it goes on parameter URLs, filtered views and pages that should never have been reachable. Every one of those is attention not spent on a page you want quoted.
83%
475pages made machine-readableThe estate an engine can now separate from its own navigation
Ranked by Cost
Findings Are Ordered by What They Cost You
A crawler will hand you four thousand issues sorted by its own idea of severity. That ordering knows nothing about your traffic, so it will put a canonical warning on a page nobody visits above an indexing problem on the page that carries your pipeline.
We re-rank every finding against the sessions and the revenue behind the affected pages, then draw a line. Below the line goes in the appendix and stays there.
The report reads as a work order: what to change, in what order, and what each change protects.
Why the Crawl Runs Monthly
Fixes Do Not Stay Fixed
A template edit
One change to a shared component reaches every page using it. A heading level moved, a canonical dropped, a link turned into a button, and a hundred pages change behaviour at once.
A CMS migration
Slugs change, redirects get half-written, and the old URLs stay in the index for months pointing at nothing. This is the most common way a site loses traffic it had already earned.
A page published outside the process
A campaign landing page goes up in a hurry with no meta description, no schema and a noindex left on from staging. Nobody notices, because nobody is crawling.
A third-party script
A tag manager container grows by a script a month until the page takes four seconds to become interactive on a mid-range phone. No single change caused it, which is why no single person owns it.
Each of these is cheap to catch the month it happens and expensive to find two quarters later, which is the entire argument for a monthly crawl rather than an annual audit.
Case Study
475 pages made machine-readable, from zero.
We cut crawler time on junk URLs by 83% and rebuilt how the estate is structured, so the pages that matter are the pages engines spend their time on. ChannelSight is now the most-cited domain in its own AI answer set.
View Case Study
Start With a Crawl
Find Out What Is Costing You Traffic
A call about your site, your traffic and what is currently getting in the way of it. You will leave knowing what we would do first and why, whether or not you go ahead.
- What the crawl, index, speed and structure work covers
- How findings get ordered by the traffic and revenue behind each page
- Where we can fix things in the build, and where we cannot
Common QuestionsFrequently Asked
Frequently Asked
Questions
Everything that decides whether a search engine can reach your pages, understand them and keep them in its index. It covers crawling, indexing, site speed, structure, markup and internal linking. It does not cover what the pages say, which is content, or who links to them, which is authority.
A tool lists everything it can detect, weighted by its own severity scale rather than by your traffic. An audit is someone deciding which of those findings matter for your site, in what order, and what the fix actually is. The list is the input; the work order is the deliverable.
We fix them. We build the sites, so a finding turns into a change on the site rather than a recommendation handed to a development team who were never briefed. Where a fix sits in a system we do not control, we say so and scope it separately.
A full pass once a year, and a monthly crawl in between. Most damage comes from ordinary changes: a template edit, a CMS migration, a marketing page published outside the usual process. Monthly catches those while they are still cheap to reverse.
It decides whether the page can be used at all. An engine that cannot crawl your page, or that cannot separate your content from your navigation, has nothing to quote. Technical work alone will not earn a citation, and its absence reliably costs one.
For audit and diagnosis, yes. For implementation, it depends what the site runs on and who owns the deploy. We say clearly where we can ship the fix ourselves and where you would be handed a list to give someone else.