Troubleshooting
Check number of backlinks: a troubleshooting table
A link you can see in a browser but not in a report is a common complaint. Almost always the cause is one of eight dull things, listed here with a test for each.
Symptom, suspect, test
| Symptom | Suspect | Test |
|---|---|---|
| Link visible in browser, absent from the report | Crawl date: the index refreshes about monthly | Check how old the link is. Under a month, wait a cycle |
| Total differs by 2x between tools | Raw backlinks versus referring domains | Compare referring domains only |
| Links to old URLs vanish after a redesign | Missing 301 on the old address | Request the old URL and read the status code |
| Linking page returns 200 but nobody counts it | noindex or blocked by robots.txt on their side | Open their robots.txt and the page source |
| Links gone after an HTTPS or domain move | Redirect chain, or only the homepage redirected | Trace three old URLs hop by hop |
| Counts drop with no change on your site | The linking site pruned thin pages | Open five lost referrers and see if they still exist |
Links appear for www but not the bare domain | Two hostnames treated as two sites | Query both forms and add them up |
| A new link shows in one tool days before another | Different crawl schedules | Re-check after the next monthly refresh |
The cheapest test first
Run the free overview on the domain and read the referring-domain figure, then do the same elsewhere. Same ballpark means your tools agree and your worry is about one link. Far apart means a structural cause. The table above starts there.
Many readers meet this problem after a migration. A site that changed platform often keeps its content but loses the redirect map. Painful. Old addresses then return 404, and the links pointing at them stop passing anything. Fixing it is dull work: list the old URLs from your analytics, map each one, and redirect with a 301.
Crawl rules matter on the other side too. A referrer that adds noindex or blocks a path in robots.txt can drop out of everyone's reports without being deleted.
Also count the www form and the bare domain together before comparing totals. Tools split them differently.
One more habit helps: write the expected count before you look. If you predicted 120 referring domains and see 95, the gap is a number to explain, and explaining it is half the job.
A shop moves from /products/blue-widget to /shop/blue-widget and redirects only the homepage. Fifty links aimed at the old product URLs now land on a 404. Reports lose them over the next refresh cycles. After a page-by-page 301 map is live, the links count again once the crawlers revisit. Numbers here are illustrative.
Reconciliation checklist
- Same date window, same domain form, same metric in both tools
- Old URLs return a 301 to a live page, not a chain of three hops
- robots.txt on your site does not block the pages that earn links
- Search Console lists the links you expect for your own site
- You waited one full refresh before calling a link lost
When it really is lost
Sometimes the link is gone. Pages get edited, sites go offline. Then the only remedy is outreach, which belongs to link building and not to this page. Google Search Console shows your own links, and Majestic or Ahrefs can supply a third opinion.
For a calmer second opinion on your own profile, go through the monthly routine. Costs of each call are in the reference. Terms such as referring domain and redirect chain are defined in the glossary.