Blog · SEO without the hype
Most sites do not need a backlink audit. That is not the answer the people selling them give, so it is worth starting there.
A manual action for unnatural links. Search Console tells you explicitly. This is the one case where a disavow file is not optional, because reconsideration requires demonstrating you cleaned up.
You inherited the domain or the agency. You bought a company, took over a site, or replaced an agency that would not describe its methods. You are auditing history you did not write. This is the most common legitimate reason.
Due diligence before an acquisition. If someone is buying a business whose traffic is part of the valuation, the link profile is part of the asset. Bought links are a liability that transfers with the domain.
An unexplained drop that correlates with link activity. Rankings fell, the timeline matches a period when links were being built, and nothing on-site changed. Even here, check the on-site explanations first. They are more likely.
Outside these four, an audit usually confirms what you already suspected and produces a disavow file that changes nothing.
Referring domain inventory. Every domain pointing at you, with its own authority signals, topical relevance, traffic, and indexation status. A link from an unindexed site is not a link.
Anchor text distribution. The strongest single signal in link analysis. Natural profiles are dominated by brand terms, bare URLs, and generic phrases. When 40% of your anchors are exact-match commercial keywords, nobody needs a tool to tell them what happened.
Velocity. When links appeared. A profile that gained 800 domains in one month three years ago and nothing since has a shape that is visible from orbit.
Topical coherence. Whether the linking sites have anything to do with your subject. A B2B software company with links from gambling directories and general-purpose "business" blogs in five languages is a specific pattern.
Placement quality. Sitewide footer links, author bios on syndicated posts, and comment-section links behave differently from an editorial mention inside relevant content. Position within the page matters more than most reports acknowledge.
Live status. A meaningful share of any historical profile is dead. Links to 404s, removed pages, and expired domains. Cleaning the report of these before analysis changes the picture.
This is where most link audits go wrong, in both directions.
The overselling version treats disavow as maintenance and hands you a file with 4,000 domains in it. The risk is real: disavowing links that were passing legitimate value is a self-inflicted ranking drop, and it is slow to diagnose because the effect appears weeks later with no obvious cause.
The under-reacting version says disavow never matters. That is true right up until you have a manual action, at which point it is the only path back.
Our working rule: disavow when there is a manual action, or when you can point to a specific batch of links you know were purchased and can identify precisely. Do not disavow because a tool assigned something a "toxicity score." Those scores are vendor heuristics, not signals from Google, and they routinely flag legitimate directories, aggregators, and press coverage.
If you cannot explain why a specific domain is in the file, it should not be in the file.
Roughly €600 to €2,500 for a standalone backlink audit in Europe, depending on profile size and whether remediation is included. Sites with under a few hundred referring domains sit at the bottom. Anything with a long acquisition history, multiple markets, or an inherited profile of tens of thousands of domains sits at the top.
The variable that moves the price is manual review volume. Tooling produces the inventory cheaply. Deciding what each cluster of domains actually is takes a person, and that is what you are paying for. An audit priced as if no human reviewed anything probably had no human review anything.
Outreach-based removal, where someone contacts webmasters to take links down, is usually quoted separately and is mostly theatre. Response rates are low, and Google has said disavow achieves the same outcome. Pay for it only if a reconsideration request needs evidence of good-faith effort.
That last point gets skipped and it is the one with commercial value. Knowing that 3% of your profile is junk changes little. Knowing that every competitor ranking above you has links from four trade publications you have never approached changes your quarter.
An audit that only removes is an audit that leaves you where you started, minus some noise. If links were a genuine factor in your ranking gap, the finding is that you need more of the right ones, and that is a different piece of work with a different budget.
For European B2B specifically, the sources that matter are narrow and mostly unglamorous: trade associations, industry publications, partner and integration directories, customer case studies published on the customer's own site, and conference or speaker pages. None of it scales, all of it is durable, and none of it appears in a toxicity report.
---
A review of every site linking to yours, assessing quality, relevance, anchor distribution, and acquisition pattern, to determine whether any of it is a liability.
Probably not, unless you have a manual action, inherited a domain or agency, are running acquisition due diligence, or have an unexplained drop that correlates with link building.
Roughly €600 to €2,500 in Europe, driven mainly by how much manual review the profile size requires.
An industry term rather than a Google one. Vendor tools assign toxicity scores using their own heuristics. Useful for triage, unreliable as a basis for disavowing.
Only with a manual action, or when you can identify a specific batch of purchased links. Disavowing on tool scores alone risks removing links that were helping.
Negative SEO is far rarer than the people selling protection against it suggest. Google discounts most spam automatically. Genuine cases exist and are unusual.
One to two weeks for most profiles. Longer for inherited domains with a long history, where the manual classification is the bulk of the work.
We'll check indexing, speed, content and the backlink profile, name the three biggest problems and the order of work. Real findings, no filler, no strings attached — the report is yours to keep either way.