Backlink Audit and Disavow Decision for Industrial Sites

Six of the last forty backlink audits we ran ended in a disavow file. The other thirty-four ended in a written recommendation to leave the file untouched, because the traffic decline that prompted the call turned out to be a template change, a canonical pointing at the wrong variant, or a competitor who started publishing. Disavow repairs a specific injury. Used as routine hygiene it removes links that were doing work, and nobody notices for eight months. What this engagement decides: Whether links are your problem. If they are not, you get the evidence for that and the name of what is, which is usually more useful than the audit you asked for. If they are, you get the file, the reconsideration request, and the paperwork behind both. How the inventory is built: Search Console's links export, Ahrefs and Majestic, reconciled against each other at registrable-domain level rather than URL level; the disagreements between the three are informative on their own. We then crawl the linking pages ourselves. Is the link still in the served HTML rather than injected by script, has the page been rewritten since placement, does the site still resolve at all. Thirty days of raw server logs, when you can get them released, tell us which of those pages Googlebot has visited this year. A link on a page last crawled in 2019 is not what is hurting you. What Italian industrial profiles contain: Directory submissions from the 2009 to 2014 period, when every provincial web shop sold manufacturers a segnalazione package. Consortium, chamber of commerce and industrial district listings, which are legitimate and which people keep trying to disavow. Expired domains recycled into Italian content farms, usually carrying your own catalogue copy spun past recognition. Scraped mirrors of your product pages hosted abroad, which is a copyright matter and not a link matter. And often a portfolio farm run by a previous agency, linking each of its customers to all the others. When we file: Two triggers, and no others. A manual action visible in the Search Console report, or a purchase history you can confirm, meaning an invoice for a link package or a placement spreadsheet somebody inherited. Then the file goes in at domain level, and the reconsideration request is written as a plain account of what was bought, when, from whom, and what has been done since, with the invoices attached. Reviews have taken two to six weeks in our experience and the first answer is often no. What we will not do: Build links. That is different work and a different budget. Run removal outreach, because paying a webmaster to take a link down funds the people who place links in order to charge for removing them. Promise a ranking recovery, which depends on your own site more than on anything we submit. Sell this as a subscription: a profile that needed auditing in March does not need auditing again in June. Who should not buy this: Anyone who has already had two reconsideration requests declined. That is a specialist job by then and we will give you a name rather than take the third attempt. Anyone whose brief is to clean up the profile with no symptom attached, because there is no clean, there is an inventory and a judgement. And anyone who cannot give us an hour with an engineer or a product manager. The person who knows why a standards body links to your fatigue-testing page settles more questions in that hour than a week of tooling does.