Technical SEO Audit with Log and Rendering Forensics

You will get eleven findings, ranked, each with the evidence that proves it and the change written the way your engineers need to receive it. Not a crawler export with 400 issue types, three of which matter. We have inherited enough of those documents to know that the reason they never get implemented is not laziness on the dev team's side; it is that nobody triaged, so every item arrives with equal weight and the ticket dies in the backlog. How the crawl is actually run: Seeded from four sources, not from the homepage. Server logs, Search Console performance and inspection data, your XML sitemaps, and internal links. The gap between those lists is where the findings live: URLs Googlebot requests that you do not link to, URLs you link to that it has never fetched, sitemap entries that return 404 to a crawler and 200 to a browser. We crawl twice, once rendered and once not, and diff the two, because on a JavaScript site the difference between the raw DOM and the rendered DOM is where canonical tags, hreflang and internal links quietly disappear. Logs, when we can get them: Thirty days of raw access logs from nginx, Cloudflare Logpush or your CDN of choice, verified by reverse DNS so we are reading Googlebot and not something wearing its user agent. Requests by directory against clicks by directory tells you where crawl is going and what it is being wasted on. Without logs the audit still runs, and we will mark the sections where we are inferring rather than observing, so you know which conclusions are softer. Rendering, indexing and the parts people skip: Index status pulled per URL through the URL Inspection API rather than a site: query, which is an estimate and has always been one. Soft 404s that return 200. Parameter combinations multiplying quietly behind a filter UI. Core Web Vitals read from CrUX field data, INP first, because lab numbers from a laptop on office broadband have misled more teams than they have helped. Redirect chains at the edge that the application layer cannot see. What is not in scope: We do not fix the code. We hand over the finding, the reproduction, the proposed change and a way to verify it, then work with your engineers to get it shipped, but the commits are theirs. No content strategy, no keyword mapping, no link work, no CMS migration project management. We do not sell a retainer on the back of this and will tell you plainly when the answer is that your site is technically fine and the problem lives somewhere else. Not a fit if: Your site is under a few hundred URLs, where this is expensive relative to what it can find and an hour of consulting serves you better. You cannot give a staging environment or read access to logs and Search Console, since without them we are guessing in public and charging for it. You want the crawl file and nothing else, which is a fair thing to want and about a tenth of the price from someone else. Every engagement ships with the raw crawl file alongside the written document, so a second opinion costs you nothing and our reasoning can be checked line by line. Findings are re-verified after your team ships, against the same crawl configuration, so the closing report is measured rather than assumed.