Backlink Audit, Disavow Only When the Evidence Says

We have run backlink audits for thirty-one companies since 2023 and recommended submitting a disavow file in four of them. That ratio is the finding most buyers of this work need before they buy it. Google's spam systems discount the overwhelming majority of junk links without being told to, and a disavow file is a standing instruction that outlives everyone who understood why it was written. The audit earns its price. The file usually does not, and a firm that ends every audit with one to upload is selling the file. What the audit answers: Whether anything in the profile is capable of doing harm. Whether a ranking drop you have already seen was link-related or merely coincided with a core update. And what a previous agency actually bought in your name. That last question is the most common reason we are hired: a new head of growth inherits a profile and needs to know what is in it before an acquirer's diligence team asks, or before anyone signs a warranty about marketing practices. What we look at: Referring domains pulled from Ahrefs, Semrush and Search Console together, because each misses different things and Search Console is the only one showing what Google has recorded. We separate links you built from links you were given from links someone pointed at you. We read pages rather than scores: a domain with a clean toxicity rating that carries your link in a footer across four thousand pages is a worse signal than a dead blog with one mention. We date the acquisition curve and look for the shape a bought campaign leaves behind: forty domains in a fortnight, one anchor, the same markup footprint on every one of them. When a disavow is the right call: A manual action is in place. Or you hold documented evidence of a paid campaign run by a previous vendor and you are closing it out honestly ahead of a raise or a sale. Or a negative campaign against you is measurable and still running. Outside those three, we will tell you to leave it alone and put that in writing, so the recommendation survives the next person who reads an article about toxic links. What you receive: A findings document written to be read by someone who is not an SEO; your CFO or your board can follow it without a glossary. A classified sheet of every referring domain with the reason for its classification rather than a score. Where a manual action exists, the file itself plus the reconsideration request, written and submitted, and the second attempt if the first is refused. Decisions come back in one to three weeks and a first refusal is common. What is not included: Link removal outreach. Emailing site owners to ask for removal has a response rate that does not justify what it costs to run, Google has said the file is sufficient, and we will argue with you about it rather than quietly bill for it. Also excluded: building the replacement links, which is separate work, and any judgement on your on-page or technical issues beyond noting when they, and not your links, explain the drop. Who this is not for: Anyone who has already decided they need a disavow file and wants a supplier to produce it. If the conclusion is fixed before the work begins, hire someone cheaper. Sites under roughly 150 referring domains, where a competent person in-house can read the entire list in an afternoon and should.