Free page link analysis
Free Outbound Link Counter
See where a page links before you approve a publisher or revise an article.
Count internal and external links, inspect destination hosts and anchor wording, and review nofollow, sponsored and ugc attributes. Start with a public URL, pasted HTML or a saved HTML file.
From a count to a decision
Find the pattern.
Read the context.
Repeated destinations, vague anchors and concentrated hosts tell you where to look next. The report supports your review; it does not grade a website or predict ranking value.
Explore the review workflowThe working counter
Inspect the exact page.
Choose your source, run the check, then explore the links behind the totals.
Source and coverage
External destination concentration
Excluded references and source declarations
Inspect link evidence
Save the evidence
Exports include source information, counting rules and collected-row limits. View filters never change the snapshot's headline totals. Print includes the currently displayed occurrence page.
Compare a previously downloaded report
Comparison stays in browser memory. Use the same source scope and counting rules. Capped snapshots cannot prove a link was removed.
Only the shown summary and your notes are transferred. Review it and add contact details before submitting.
Public URL checks retrieve HTML through this site’s server. Supplied HTML is analyzed in your browser. Check the report’s source and coverage notes before comparing results.
A useful check in three steps
A report you can act on.
Use it for publisher research, citation checks, content reviews and template QA.
Choose a representative page
Enter the article or landing-page URL you want to inspect. For an unpublished draft, paste HTML or select an HTML file and supply a page address for relative links.
Review scope and evidence
Confirm the source, site definition and inspected region. Start with occurrence and unique counts, then read the destination hosts, link names and relationship attributes.
Save the useful findings
Filter the view, export the collected report or compare a saved JSON snapshot. Preview an evidence brief when you want help with the next editorial or publisher decision.
A publisher’s homepage may contain menus, category links, social buttons and footer references that differ from an editorial article. Check a representative article for a proposed article placement and compare similar page types.
Understand the evidence
What the outbound-link report shows.
The totals describe the accepted links in the inspected snapshot. Read the counting scope and coverage with the numbers.
| Report item | What it means | How to use it |
|---|---|---|
| Total links | Accepted HTTP/HTTPS link occurrences in the inspected HTML and selected scope. | Compare with the page type, navigation and editorial purpose. |
| Internal links | Occurrences classified as inside the chosen site definition. | Review useful routes into the site; check the scope before comparing pages. |
| Unique internal links | Distinct internal destination URLs under the displayed URL rules. | Separate repeated navigation from the variety of pages referenced. |
| External links | Occurrences classified as outside the chosen site definition. | Read citations, commercial references and their surrounding content. |
| Unique external links | Distinct external destination URLs. | See whether repeated references inflate the occurrence count. |
| External hosts & share | External destinations grouped by normalized hostname, with their share of external occurrences. | Find concentrated destinations and investigate why they recur. |
| Anchor & link-name evidence | Text and supported naming evidence, such as image-alt text and supplied labels. | Check whether readers can understand the destination; inspect the original markup where needed. |
| Relationship attributes | The anchor’s reported rel value and supported token flags. | Review how the source marks ordinary, paid and user-created links. |
| Source & coverage | Input mode, available source metadata, analysis time and collection limits. | Distinguish a collected snapshot from browser-rendered or exhaustive coverage. |
Overall relationship counts cover accepted links in the selected analysis scope. External-domain detail describes external links. Attribute filters and view controls can narrow a table without changing the underlying collected snapshot.
Occurrence versus destination
One URL can appear three times.
If the same external URL appears in three places, it contributes three external occurrences and one unique external destination under the tool’s URL rules.
A sentence, a source list and a reusable component may each have a valid reason to reference it. Count the repetition, then check whether it helps the reader.
Hostname versus ownership
Different hosts are not different companies.
With the default same-host comparison, www.example.com and example.com are equivalent. Other subdomains can appear separately, including docs.example.com.
Supported site-scope settings help you classify related hosts. A technical group is not evidence of business ownership, independent publishers or a link network.
Make comparisons reproducible
Know what entered the count.
Use the same source mode, site definition and content region when comparing two observations.
Resolve real destinations
The workspace preserves raw href evidence beside resolved URLs. Relative references use the effective document base; review the base shown in the report when the HTML contains a <base href> element.
Fragment-only, empty, email, telephone, JavaScript and data references do not enter the main HTTP/HTTPS link totals. Their evidence belongs in the report’s excluded-reference information.
Choose a useful region
Whole-page HTML may include navigation, headers, footers and other page components. Use the supported region controls or a CSS selector to inspect a narrower area when the markup allows it.
Location labels reflect the inspected markup. Read a main-content result beside the original article; a region label alone does not prove that every link is an editorial citation.
Keep the URL rules visible
Occurrence and unique views answer different questions. Query parameters, fragments and normalized hostnames can affect destination grouping, so check the displayed URL rules before interpreting duplicates.
The selected scope travels with the report. A changed grouping rule can change internal/external totals without any change to the page’s actual links.
Read the markup accurately
Followed, nofollow, sponsored and UGC.
Followed means the anchor has no nofollow token. Sponsored and UGC are recorded separately and can overlap with other flags.
| Markup | Observation in the report |
|---|---|
| No qualification token | An ordinary anchor without the supported relationship qualifications. |
rel="nofollow" | The source includes the nofollow token. |
rel="sponsored" | The source declares a sponsored relationship. |
rel="ugc" | The source declares user-generated content. |
rel="sponsored nofollow" | One occurrence contains both tokens and contributes to both counts. |
rel="noopener noreferrer" | Browser-related tokens; these do not themselves add a nofollow token. |
Do not add every attribute total and expect the sum to equal the link count. This Followed definition is broader than the everyday shorthand “dofollow,” which often excludes all three qualification tokens. Neither label proves that a search engine crawled, indexed or assigned ranking credit to the destination.
For a specific source-to-target check, including supported page-level directive evidence, use the Dofollow Nofollow Checker. A page-level nofollow instruction needs separate consideration even when an individual anchor has no nofollow token.
Review the page in context
Turn the report into an editorial decision.
Use the counter to organize the evidence, then read the actual references.
Start with the page’s purpose
A research guide may need primary sources. A comparison may reference manufacturers, documentation and reviews. A directory may intentionally contain many destinations. A high count alone does not show poor quality; a low count does not establish a good publisher.
Look at the most referenced hosts
Review host groups in occurrence-count order, check their share, then filter the link table by hostname to read the occurrences. Repeated citations to one relevant source can be sensible. Unrelated commercial references or anchors deserve closer review. The percentage tells you where to look; context explains the pattern.
Read the anchor in its sentence
Check generic wording, repeated sales phrases, missing labels and promises that do not match the linked page. Use the Anchor Text Optimizer & Analyzer for supplied wording and context; it does not certify an ideal anchor ratio.
Check the source behind the claim
Open important references and confirm that they support the statement beside them. The Link Relevance Checker supports wording and context comparison; lexical similarity does not prove editorial endorsement or ranking value.
Review the wider publisher evidence
Consider recent content, the intended audience, editorial standards, useful visibility evidence and placement terms. The Publisher Site Risk & Spam Signal Checker can prioritize further inspection; its flags are not a penalty diagnosis.
A historical, completed check
213 occurrences. 71 unique destinations.
A supplied research snapshot of the “What Are Backlinks?” article, dated , shows why both measures matter. These are historical results, not a current check.
| Measure | Returned result |
|---|---|
| Total accepted occurrences | 213 |
| Internal occurrences | 206 |
| Unique internal destinations | 67 |
| External occurrences | 7 |
| Unique external destinations | 4 |
| External hostname groups | 3 |
External occurrence distribution
The seven external occurrences led to four unique external URLs. Rounded shares total 100.1%, and the two Ahrefs hostnames are separate groups under the historical same-host rules.
Whole-page counts include navigation and other components. They do not mean the article body contained 213 links. Results change when the page or returned HTML changes.
Imagine two articles with twelve external occurrences. One cites original research, useful data and documentation; the other repeats an unrelated commercial destination through sales phrases. This is a fictional illustration: the total is the same, but the references, anchors and context call for different decisions.
Useful links over target ratios
Is there an ideal outbound-link count for SEO?
There is no universal ideal number or internal-to-external ratio for every page. Use citations that support the content and internal links that help readers continue. Removing useful sources to lower a number can weaken an article; adding irrelevant links to change a ratio can make navigation harder.
Qualify paid placements and advertisements appropriately, commonly with rel="sponsored"; nofollow can also mark that relationship. User-generated links can use ugc. A count or Followed label does not change the relationship that needs to be disclosed.
Our editorial standards explain the source and content expectations to consider alongside the report.
Choose the right workflow
Useful ways to use the counter.
Before choosing a guest-post publisher
Inspect articles that resemble your proposed placement. Review relevant citations, concentrated hosts, commercial anchors and existing relationship attributes. Confirm current terms for your exact brief.
Review publisher selection criteriaBefore publishing or revising an article
Inspect draft HTML, then check the public page after publication. Look for repeated destinations, generic anchors and unexpected links from a template or plugin.
Extract links from other supplied contentAfter a theme or navigation update
Select Full collected rows and save a JSON report, repeat the check with equivalent scope and compare the local snapshots. A count increase may come from menus, a footer or a widget instead of extra article citations.
Keep the source, date and page type. Comparisons cover saved observations; the counter does not create automatic monitoring.
When citations may be broken
A destination can appear in HTML while its target is unavailable or redirected. The counter records structural link evidence; it does not request every linked destination.
Run a separate bounded status reviewFor a status review, a blocked request, timeout or incomplete check is not proof of permanent deletion. Read the tool’s coverage with its result.
Keep useful evidence
Filter, compare and export with a clear scope.
Select the export that matches your task. Keep the source URL, analysis date and counting settings with any report you share.
Explore the current view
Use the workspace’s filters and sorting to inspect relevant hosts, anchors, attributes and regions. Occurrence and unique views reveal repetition and destination variety separately.
A filtered view is a subset of the collected report. Its controls do not make the source snapshot more complete.
Choose a view or full export
CSV choices distinguish the current view from the collected link data. JSON retains the structured report for later comparison. Text and print views support a readable brief.
Full means the dataset collected for this report, subject to source and analysis limits. It does not promise every link visible in an interactive browser.
Compare saved observations
Load a compatible saved JSON report to review changes against the current report. Inspect added or removed destinations and changed link evidence with the two source dates and scopes in view.
Incomplete snapshots or different settings can explain a difference. Keep those caveats beside your findings.
Coverage before conclusions
What a one-page link count can tell you.
A successful report describes its inspected source. A failed retrieval is not a zero-link result.
Returned HTML, not a browser session
Public URL checks read server-returned HTML. Links created only after JavaScript, login, consent or another interaction may be absent. Supplied HTML represents the file or text you provide, including any omissions.
Bounded source and report coverage
Public checks process one URL at a time and retrieve up to 2 MB of HTML. Row and collection coverage appears in the report. Read those notes before treating a long result as exhaustive.
Public source restrictions
Private, local, credentialed and non-standard-port source URLs are blocked. A source may also refuse an automated request, require authentication or return an unsupported response.
Counts support review
The counter does not discover inbound backlinks, crawl an entire site, confirm a link farm or diagnose a search penalty. It does not establish destination availability, indexing or ranking credit.
A public URL is submitted to the site’s server for retrieval. Supplied HTML is processed in your browser, and downloads are created from the report. Review the site’s privacy information for wider data handling; do not assume a server check has no logs or retention.
Frequently asked questions
Know exactly what you’re checking.
What is an outbound link?
With the default same-host setting, an outbound or external link points to a different normalized hostname. The checker treats www and non-www versions of the same hostname as equivalent. Other supported site-scope settings can change which related hosts count as internal, so read the report’s selected rule.
Does this count backlinks pointing to my website?
No. It counts links found in the source page or HTML you inspect. A backlink checker works in the opposite direction by finding pages that link to your target.
Are internal and external links both counted?
Yes. The report includes accepted internal and external occurrences, with distinct destination counts and detailed link evidence. The selected site scope determines the classification.
Does the counter include menus and footers?
Whole-page analysis includes qualifying links in navigation, footers and other components. Use the supported region controls or a CSS selector for a narrower scope. Region labels depend on the inspected markup and should be checked against the original page.
Why is the unique count smaller than the link total?
Repeated destinations contribute multiple occurrences but fewer unique URLs. Read both figures with the displayed URL-normalization rules to understand repetition and destination variety.
Why are two subdomains listed separately?
Destination-host grouping uses normalized hostnames, not company identity. A main site and documentation subdomain can appear as separate groups even when one organization operates both. Supported site-scope settings can classify related hosts as internal without proving ownership.
What does Followed mean here?
It means no nofollow token was detected on that anchor. Sponsored and UGC flags are separate and can overlap. The label describes anchor markup; it does not establish crawling, indexing or ranking credit.
Can sponsored and nofollow counts overlap?
Yes. One anchor can contain both tokens and contribute to both counts. Relationship totals are not mutually exclusive categories.
Do noopener and noreferrer make a link nofollow?
No. They describe browser behavior and referrer handling. They do not themselves supply the nofollow token.
Does the report confirm page-level following directives?
Read this counter’s relationship flags as anchor-level observations. Use the dedicated Dofollow Nofollow Checker when you also need its supported source-page directive evidence.
Can it confirm that every destination works?
No. Destination status checking is a separate task. A link may be present in the HTML while its target is unavailable, redirected or unsuitable. Use the Broken Link Finder for a bounded status review and read incomplete checks separately.
Does it run JavaScript or inspect logged-in content?
No. A URL check reads HTML returned to the server without operating a full browser session or inspecting private content. Browser-created links may be missing. If you supply saved or rendered HTML, the report analyzes that supplied snapshot.
Can I paste HTML or select a local HTML file?
Yes. Choose the supplied-HTML input and paste the markup or select a supported HTML file. Provide the page address for resolving relative links and defining the source host. This mode processes the supplied content in your browser and does not fetch linked destinations.
Why might a page fail to load?
A source may block automated requests, require authentication, return an unsupported response or fall outside the permitted public-source rules. Recheck the URL and the message. A failed fetch does not establish that the page has no links.
Will the exported CSV match my filtered table?
Choose the filtered-view export to save matching collected rows across all table pages, or the full export to save all collected rows. CSV, JSON and text follow that choice. Headline metrics remain attached to the full analyzed snapshot. Use full JSON for later comparison. Every export remains subject to the source and analysis limits shown in the report.
Can I compare a page before and after an update?
Yes. Select Full collected rows and save a compatible JSON report, run the later check and load the earlier report for local comparison. Compare source modes, dates and scopes before treating a difference as a page change. This does not create automatic history, scheduled monitoring or alerts.
Can a report identify a link farm or a search penalty?
It can reveal patterns worth reviewing. Counts, concentration and markup alone do not prove a link farm, search penalty or harmful backlink profile.
Is the tool free?
The public checker is offered as a free page-inspection tool. Publisher placements, managed reviews and other campaign services have their own terms and costs.
Is analysis local to my browser?
Supplied HTML is analyzed in your browser. A public URL check sends the source URL to the site’s server for retrieval, then the workspace analyzes the returned HTML. Downloads are created from the report. Consult the site’s privacy information for broader data handling rather than assuming no logs or retention.
Does continuing with a report send an enquiry?
No. The report lets you preview an evidence brief and continue to Contact. Review the message and choose whether to submit it there. The handoff does not order a placement or guarantee acceptance.
Inspect the links that matter
Better evidence for your next decision.
Review a representative page, read its references in context, and keep the report with your campaign brief.