Backlink Audit: You're Auditing the Wrong Half
Most audits spend their time hunting links Google already discounts. The output that changes anything is a map of which of your pages have earned links and which have earned none.
A backlink audit is a systematic review of every site that links to yours, what those links point at, and what they are actually doing for you.
That is the definition. The practice that grew up around it went somewhere else.
Open almost any audit template and the workflow is identical. Export your referring domains, run them through a third-party "toxicity" score, sort descending, build a disavow file, upload it, close the ticket. The deliverable is a cleaner list and the feeling is hygiene.
For the overwhelming majority of sites, Google says that work changes nothing. Its disavow help page is blunt about it: "In most cases, Google can assess which links to trust without additional guidance, so most sites will not need to use this tool."
So the cleanup half of the standard audit is mostly theatre. The half nobody runs is the half that pays.
A backlink audit is an asset inventory, not a hygiene check. Its real output is a map: where your authority is concentrated, where it is completely absent, which referring domains are genuine editorial relationships rather than one-off directory noise, and which links you have quietly lost. Audit for coverage, not for toxicity.
What is a backlink audit?#
A backlink audit is the process of pulling every known inbound link to your domain and answering three questions about it: who links to you, what they link to, and whether that link is doing any work.
Most people only answer the first. They scan the domain list, recognise a few, feel uneasy about the rest, and stop.
The second question is where the value is. A referring domain is not a property of your site, it is a property of a specific page on your site. Twelve links to your blog and zero to the page you sell from is a completely different situation from six and six, and the domain-level number hides it.
The third question is the honest one. A link in a paginated directory nobody reads and a link in a paragraph your buyers actually read are both one row in the export, and they are not the same asset. If you have never written down what a backlink is worth, the export will flatten that difference for you.
An audit that ends at the domain list has answered the least useful question.
Do toxic backlinks actually hurt your site?#
Almost never, and "toxic" is not a Google concept. It is a vendor metric. No search engine publishes a toxicity score, so every number you have seen is a third-party model guessing at one.
What Google does publish is a spam policy. It defines link spam as "the practice of creating links to or from a site primarily for the purpose of manipulating search rankings" and lists what counts: bought links, excessive exchanges, automated link creation, advertorials that pass ranking credit, low-quality directory and bookmark links, links stuffed into widgets or footers across many sites.
Read that list again and notice who it is about. Nearly every item describes something you did, or an agency did for you. It is a policy about link building, not link receiving.
Google is explicit on the other side too: "Google works very hard to make sure that actions on third-party sites do not negatively affect a website."
If you did not build the links, they are usually not your emergency.
The corollary matters for how you build. Google says the paid part of the web is fine as long as it is labelled: "It's not a violation of our policies to have such links as long as they are qualified with a rel="nofollow" or rel="sponsored" attribute value to the <a> tag." Those attributes are a disclosure, not a punishment, and Google's guidance on qualifying outbound links says links carrying them "will generally not be followed" rather than being ignored outright. If that distinction is fuzzy, the dofollow versus nofollow breakdown is the longer version.
When is disavowing actually the right call?#
When two things are true at once, which is rarer than the tooling implies.
Google states the test directly. You should disavow only if you have "a considerable number of spammy, artificial, or low-quality links pointing to your site," and those links "have caused a manual action, or likely will cause a manual action, on your site."
That is an AND, not an OR. A messy-looking profile on its own does not qualify. Neither does a ranking drop you cannot explain.
The second condition is checkable in thirty seconds. Open the manual actions report in Search Console. Google issues a manual action "when a human reviewer at Google has determined that pages on the site are not compliant with Google's spam policies," and the consequence is not subtle: "some or all of that site will not be shown in Google search results." If that report is clean, the disavow file you were about to upload is a no-op at best.
At worst it is not a no-op. Google calls the tool "an advanced feature" that "should only be used with caution," and warns that used incorrectly it "can potentially harm your site's performance in Google Search results." You are hand-deleting evidence from a system that was weighing it for you.
There is one genuine case: you or a previous agency bought links and the bill has come due. Even then Google's recommended process starts with removal, not disavowal. Pull your links from Search Console, find the ones that breach the link spam policy, and get them taken down at the source. Disavow is the fallback for the ones you cannot reach.
Disavow is an emergency brake with a specific trigger, not a maintenance task.
| Symptom | What it usually means | What to do |
|---|---|---|
| Spammy domains in your export, clean manual actions report | Normal background noise on any indexed site | Nothing. Record it and move on |
| Manual action for "unnatural links to your site" | Google believes you built or bought them | Remove at source first, disavow what you cannot remove, then file reconsideration |
| Ranking drop, no manual action, no link changes | Almost certainly not a link problem | Check the core update timeline and the query set before touching links |
| Referring domains falling month on month | Link decay: pages retired, sites redesigned, links dropped | Reclaim the recoverable ones, replace the rest |
| Links concentrated on one old blog post | You have authority, on the wrong page | Internal links from that post to the pages that matter |
| Money pages with zero referring domains | The coverage gap | Outreach, and a reason for anyone to link |
What should a backlink audit actually measure?#
Coverage. Which of your pages have earned links, which have earned none, and whether that distribution matches what the pages are worth to you.
That reframe moves the unit of analysis from the referring domain to your own page, and your own pages are the thing you control.
It matters because link-free pages are the default state of the web, not an edge case. Ahrefs' December 2023 index study found 96.55% of pages get zero traffic from Google, with another 1.94% scraping between one and ten visits a month. In the same dataset, of roughly 20 million pages with no referring domains at all, only 2,997 pulled more than 1,000 monthly search visits.
So the useful audit is a join, not a sort. Put the link export next to your page list and look for rows where commercial value is high and referring domains is zero.
The row you want to find is a page that matters to the business and has never been linked to by anyone.
Four things fall out of that view. Gaps: money pages with no referring domains, which are your outreach targets and usually the pages nobody thinks to pitch because they are not "content." Concentration: one post holding most of your links, which is real authority sitting in the wrong room. Decay: links you had last quarter and do not have now. Relationships: domains that have linked to you more than once, the warmest prospects you will ever have and the ones almost nobody tracks.
| What you look at | The question it answers | What you do about it |
|---|---|---|
| Referring domains per page | Where is our authority concentrated? | Internal links from the winners to the earners |
| Money pages with zero links | What are we not pitching? | Build the outreach list from this row |
| Lost links, last 90 days | What did we quietly stop owning? | Reclaim, or replace with a better target |
| Repeat linking domains | Who already likes us? | Ask them first, every time |
How do you run a coverage audit?#
In an afternoon, with whatever link data you already have. The tool matters much less than the join.
- Export your links. Search Console's Links report is free and is Google's own view, which makes it the right baseline even if you also pay for a crawler-based index.
- Collapse to referring domains per target URL. Not per site. The page-level split is the whole point.
- Bring in your own page list with a column for what each page is worth: revenue pages, high-intent comparison pages, the pillar content you want ranking.
- Sort by value descending, referring domains ascending. The top of that sort is your target list, and it is usually uncomfortable reading.
- Flag repeat domains and lost links separately. Two warm, cheap workstreams.
- Write down the reason. For each gapped page, one line on why anyone would link to it. If you cannot write that line, the gap is a content problem first.
That last step gets skipped, and it decides whether anything happens next. A page with nothing linkable about it does not have a link gap, it has a reason-to-exist gap, and no amount of outreach fixes that. Prospecting only works when the thing you point at is worth pointing at.
What do you do with the gaps?#
You turn the audit into a target list, and then you do the unglamorous part: contacting people.
This is where most audits die. The document is correct, the gaps are real, and nothing happens, because the next step is fifty personalised emails to fifty editors and nobody has the afternoon.
The bridge is a repeatable outreach motion: take the gapped pages, find sites that already link to comparable pages, find the person who owns that page, and pitch a specific, relevant addition. That is the whole loop described in link building outreach, and the tactics that feed it are the same ones as ever: broken link building for pages that replaced something dead, guest posting where a publication is genuinely a fit, and the wider set covered in how to get more backlinks.
GenLink exists for that half of the job. It is a backlink agent: give it the page you want links to and it finds relevant sites, finds the right contact, writes the pitch against what that site has actually published, sends from a warmed mailbox, follows up and triages the replies. It does not run your audit. It handles the fifty emails your audit says you need to send. There is a free tier, so you can point it at one gapped page and see what comes back.
An audit that does not end in a list of people to email is a document, not a decision.
Frequently asked questions#
How often should you run a backlink audit?
Quarterly for the coverage view, plus a check after any migration or large content cull, where lost links show up within weeks. Monthly is over-servicing unless you are actively running outreach, in which case you are already in the same data.
How do you check your backlinks for free?
Search Console's Links report. It gives you top linking sites, top linked pages and anchor text, and it is Google's own view rather than a third-party crawl. It undersamples next to paid indexes, but it is free, it is authoritative, and it has the page-level split a coverage audit needs.
Do you need a paid tool to run a backlink audit?
No, but paid indexes find links Search Console misses and make lost-link tracking much easier. Doing this once, start free. If link building is an ongoing channel, the historical view is worth paying for, because decay is invisible without it.
Can competitors hurt you with bad backlinks?
This is called negative SEO, and it is rarer than the tools selling protection against it imply. Google states it works hard to ensure actions on third-party sites do not negatively affect a website, and its own disavow test still requires a manual action. If your manual actions report is clean, you are watching noise.
What is a good number of referring domains?
There is no absolute number, only a relative one. Compare the specific page you want ranking against the pages currently ranking for that query, not your domain as a whole. A page with three referring domains can outrank one with three hundred if the intent match is better.
How long does it take to see results after a backlink audit?
The audit itself changes nothing, which is the point. What changes results is the outreach it triggers, and new links take weeks to be crawled and reflected. Treat it as a quarter-long plan, not a fix.
