Choose the question before choosing a checker
Checking backlinks can mean discovering reported references, inspecting one known link or evaluating an entire profile. Start with a target and a question. A tool that inspects a supplied page does not discover all pages linking to a domain. A marketplace order list describes purchases and delivery records, not an independent web-wide backlink index.
For your own property, use a report you are authorized to access. For a competitor, use an external index whose access and limits cover that research; your ownership of your own site does not grant access to someone else’s Search Console. If you already have a source URL, you can investigate that reference directly. The output should state which route you used.
Obtain records from an appropriate real source
For an authorized property, the Search Console Links report provides a sampled route from target or publisher to source pages. Use the dedicated walkthrough for navigation and exports. For an external-index example, Semrush documents its Backlinks report: choose the intended target scope, open its backlink rows, apply relevant filters and export the resulting table as CSV or Excel. Check your account’s current export allowance before starting.
Keep the original provider, target scope, filters, export date and source file. A domain report and a single-page report are not interchangeable. Provider first-seen and last-seen dates describe that provider’s observations. They are not a current inspection performed by you and not automatically the publisher’s creation or deletion dates. Do not rename one provider’s score as another’s metric.
Follow a small fictional checking assignment
The example shop asks about references to its planner page. An authorized export supplies two source URLs. In the invented follow-up, the first source contains the expected destination with sponsored and nofollow attributes; the second returns HTTP 403. A separate order has no publication URL. These are three evidence states, not three confirmed live links.
The six-row table is an original investigation log, not a customer report or a live scan. The report’s earlier collection date and the investigator’s later observation date must remain separate. A blocked request cannot establish that a link is missing. An order awaiting a source URL cannot be checked as a published reference just because payment exists.
| Step | Fictional evidence | Action or result | Limit to retain |
|---|---|---|---|
| Define target | https://shop.example/planner | Choose a page-level question | Do not replace it with an unlabeled domain total |
| Choose source | Authorized property report or licensed external index | Record access, target mode and filters | Neither route proves every web link |
| Keep collection context | Export collected 2026-10-20 | Retain the original file and provider dates | Collection is not a fresh source-page inspection |
| Inspect first source | https://publisher.example/resources → planner; rel="sponsored nofollow" | Observed matching target on 2026-10-21 | Attributes do not guarantee rankings |
| Inspect second source | https://review.example/post returns HTTP 403 | Access unresolved on 2026-10-21 | Not proof of deletion |
| Separate an unfulfilled record | Order has no publication URL | Await source evidence; not checked | Payment is not proof of a live reference |
Inspect the actual element and its destination
Open a chosen source and locate the exact reference. Check href, anchor text and the complete rel value, noting redirects and whether you examined initial HTML or a rendered DOM. Google’s relationship guidance distinguishes paid and user-generated contexts. The common label dofollow is not an HTML attribute you must find; also inspect relevant page-level instructions before making a broader claim.
Record whether the destination and agreed attributes match. Do not silently remove sponsored or nofollow from a record because another field calls it follow. Preserve conflicting evidence for review. For a failed request, note the time and access result and obtain legitimate alternative evidence if needed. Avoid turning one provider’s old row into a claim about what is currently visible.
Save a short, repeatable check
Save target, exact source, provider record, observation date, access result, observed attributes and the unresolved question. Use the filled CSV to discuss the six decisions and the blank version for your own work. Supplied URLs can be tracked in the known-link monitor within its documented scope. That does not add undiscovered publishers to the list.
If the question is indexing, treat it separately. Google’s URL Inspection tool operates on a URL in the currently opened property. Access to your target property does not let you inspect an unrelated publisher’s property. A clickable link, a successful local fetch and an indexed source are different observations. Do not report one as proof of the others.
Download the worked checking log · Download a blank checking log
Report what was checked, including what remains unknown
A useful handoff says that the supplied first source matched the destination with specified attributes, the second remained unresolved due to access, and the third order still lacked a source URL. Include dates and evidence locations. Do not summarize the three as “all backlinks verified” or calculate a whole-site total from this small selection.
This guide explains the procedure; it does not deliver a free arbitrary-site index, a price finder, a spam score, software recommendations or guaranteed link opportunities. Deeper quality, gap, profile and indexing tasks have their own evidence requirements. For repeated checks, preserve the same target scope and collection method, then investigate changes instead of rewriting prior observations.