Smart Interlinking Audit Free Tool
Internal links (links from one of your pages to another) help visitors and Google find related content on your site — and pages with no internal links pointing to them (“orphans”) get buried. This tool takes a list of your own pages and, for each one, suggests the most closely related other pages to link to. It also guarantees no page is left an orphan — every page gets at least one suggested inbound link. It is a read-only report — nothing on your site is changed; you decide which suggestions to add.
What are “topic tags”? They are just a few short words you pick to describe what each page is about — for example running-shoes, reviews, nike for a running-shoe review. The tool links pages together when they share tags, so two pages tagged running-shoes get suggested to each other. You choose the tags; there is no fixed list.
How to use it: in the left box paste your page addresses, one per line. In the right box, on the same line, put that page’s topic tags separated by commas. Line 1 on the left pairs with line 1 on the right, line 2 with line 2, and so on.
About Smart Interlinking Audit
What This Tool Actually Does
Paste in a list of your own pages with their topic tags and you get back a table of suggested internal links: which of your pages should probably link to which other pages, and why. The "why" is the useful part: every suggestion comes with the actual shared tags that produced it, so you're never staring at a black-box recommendation with no explanation attached.
The matching method is Jaccard similarity, a plain word-overlap calculation explained in full further down this page. Every page in your list is also guaranteed at least one suggested inbound link. The tool actively bridges any page that would otherwise end up with zero, rather than just reporting the problem and leaving it there. And to be direct about the most important trust point up front: this is a report only. Nothing on your site changes when you run it. It suggests. You still decide, and you still do the actual linking by hand.
Why Internal Links Matter for SEO
Internal links do more real work than most site owners give them credit for, and it's worth being specific about what that work actually is rather than repeating "internal linking is important" as an unsupported claim.
Crawl discovery. Search engine crawlers find the overwhelming majority of pages on a site by following links, not by magically knowing a URL exists. A sitemap helps, but it's a supplementary signal, not a substitute: Google's own guidance treats sitemaps as a hint about which URLs exist, not a guarantee they'll all be crawled with equal priority. A page that's linked from several other relevant pages on your own site gets crawled more often and more confidently than one sitting alone with only a sitemap entry pointing at it. Every internal link is effectively a vote that says "this page is connected to something real," and crawlers weigh that signal.
Link equity distribution. Authority on a site isn't spread evenly. Your homepage and your most-linked pages accumulate the most internal link equity, and every outbound internal link from those pages passes some of that equity along to whatever it points at. A page with zero inbound internal links gets none of that equity at all, no matter how strong the rest of your site is. This is exactly why a single high-authority hub page linking out to a cluster of related content is a more effective structure than a flat pile of unconnected pages competing for the same crawl and ranking attention independently.
User navigation and engagement. A visitor who lands on one page and finds two or three genuinely relevant links to related content is more likely to stay on your site, read more, and eventually convert or return. Compare that to a visitor who lands on an island page with no obvious next step. This matters for behavioral signals search engines do track (time on site, bounce rate, pages per session), and it matters independently of SEO, as basic usability.
Reducing orphan pages. A page with zero internal inbound links is functionally invisible to both crawlers and human visitors unless someone already has the exact URL memorized or bookmarked. It might still show up in search results occasionally if it's indexed, but it's cut off from the rest of your site's link equity and from any reasonable path a visitor would take to find it browsing normally. This tool's core design goal, guaranteeing zero orphans in its output, exists specifically because this is the single most avoidable internal-linking failure, and it's one a simple audit like this one can catch and fix without needing a full crawl of a live site.
Not all internal links carry the same weight, and it helps to be specific about the different kinds a site typically has, since a tool like this one only addresses one of them directly. Primary navigation links (the top menu, footer menu) get crawled constantly and pass consistent equity site-wide, but they're broad by nature and can't cover every relevant pairing between individual pages. Contextual in-body links, meaning a link inside an actual paragraph pointing to a genuinely related page, tend to carry the strongest topical signal, because both the anchor text and the surrounding content tell a crawler exactly why the two pages are connected. Related-posts or "related tools" widgets, the kind this tool's own suggestions are built to feed, sit in between: broader reach than a single hand-placed in-body link, but more topically specific than a generic nav menu. A page that only shows up in your XML sitemap and nowhere else, by contrast, isn't really internally linked at all. A sitemap is a hint to crawlers, not a link, and it doesn't carry equity the way an actual clicked-through link does.
None of this means every possible internal link is worth adding. A link that exists purely because two pages happen to share a tag, with no real topical relevance to a reader, does more harm than good: it clutters the page and signals a weaker connection than it claims. The suggestions this tool produces are a starting point for a human decision, not a mandate to link everything to everything. That distinction gets its own dedicated section further down this page, because it's worth being explicit about rather than assumed.
What You Put In: Pages and Topic Tags
This tool works from two short lists that you provide, side by side: your pages on the left, and a few topic tags for each page on the right.
A topic tag is just a short word you pick to describe what a page is about. There's no fixed list to choose from and you don't need tags already set up anywhere in your CMS, you make them up on the spot. For a page reviewing running shoes you might use running-shoes, reviews, gear; for a marathon plan, training, marathon, plan. The tool then links pages together when their tags overlap, so two pages you both tagged running-shoes get suggested to each other, and the shared tags are shown as the reason.
You enter these as two parallel boxes on the page. In the left box you paste your page addresses, one per line. In the right box, on the same line as each page, you put that page's tags separated by commas. Line 1 on the left pairs with line 1 on the right, line 2 with line 2, and so on. That's the entire input, no file uploads, no crawling, no account. The exact step-by-step is in the "How to Use This Tool" section below.
What "Jaccard Similarity" Actually Means
Jaccard similarity sounds more technical than it is. It's a simple ratio: the number of tags two pages share, divided by the total number of unique tags across both pages combined.
A concrete example makes it clearer than a formula does. Say Page A is tagged seo, backlinks, links and Page B is tagged seo, links, generator. The shared tags are seo and links, two tags in common. The total unique tags across both pages, counted once each, are seo, backlinks, links, generator, four tags total. Two shared divided by four total gives a similarity score of 0.5. If Page A and Page B had identical tag sets, the score would be 1.0, a perfect match. If they shared no tags at all, the score would be 0.0.
This tool computes that score for every possible pair of pages in your list, then, for each page, picks the three other pages with the highest score as its suggested related links. That's the entire mechanism. It's honest to be plain about what this is and isn't: it's word-overlap math on whatever tags you supplied, not a semantic understanding of what a page is actually about. Two pages about genuinely different topics that happen to share a generic tag like "seo" or "tool" will score some nonzero similarity even if a human reader would never link them together. That's exactly why the tool's suggestions are a starting point to sanity-check, not a verdict to apply blindly, a theme this page comes back to more than once because it's the single most important way to use this tool responsibly.
The Zero-Orphan Guarantee, Explained
Picking each page's top 3 most similar matches, on its own, doesn't guarantee every page ends up with at least one inbound suggestion. It's entirely possible for a page to be nobody else's top 3: a genuinely unique page, or one whose tags don't overlap much with anything else in your list, could end up with zero pages suggesting a link back to it, even though it's suggesting links outward to others.
This tool corrects for that automatically. After the top-3 matching runs, it checks for any page that received zero inbound suggestions. For each one it finds, it identifies the single most similar page that doesn't already link to it, and adds that orphan page to that candidate's own outbound suggestion list. The result: every page in your list ends up with at least one suggested inbound link, guaranteed by design rather than left to chance.
Why this matters more than it might sound: an imperfect match is still better than zero. A forced link between two moderately related pages still gives an otherwise-orphaned page a path into your site's link graph. A real user can still discover it, and it still receives some share of internal link equity. Leaving a page with genuinely nothing pointing at it is a worse outcome than a slightly loose suggested connection, which is exactly the tradeoff this design makes on purpose. The "Reading Your Results" section below explains how to spot these bridged suggestions in the actual output table, so you can weigh them a little more critically than a suggestion backed by real shared tags.
How to Use This Tool
- In the left box ("Page URL"), paste your pages one per line. A full address like
https://example.com/best-running-shoes/is ideal, but any consistent label you'll recognize works too. - In the right box ("Topic tags"), on the same line as each page, type that page's tags separated by commas, for example
running-shoes, reviews, gear. Line 1 on the left pairs with line 1 on the right, line 2 with line 2, and so on. - Give each page a few specific, genuinely descriptive tags. Three to six per page is usually enough for the similarity scoring to produce meaningful results; too few tags per page gives the tool very little to compare.
- Click the "Suggest Internal Links" button.
- Read the summary line and the link-distribution box at the top of the results to confirm how many pages were analyzed and that no page was left an orphan.
- Review the results table, then treat every suggestion as a candidate worth a quick relevance check before you manually add any link, never as an instruction to apply automatically.
A few practical rules worth knowing before you paste a long list: each page can appear once (if the same page shows up twice, only the first occurrence is used), each page is capped at 20 tags and each tag at 60 characters, and the tool audits up to 3,000 pages. A page line with no tags beside it is simply skipped rather than causing an error. If the two boxes get out of alignment, a tag line with no page next to it, the tool tells you so you can line them back up rather than silently guessing.
Reading Your Results
The output has a few distinct parts. Here's what each one is telling you:
| What you see | What it means |
|---|---|
| Summary line ("Analyzed N page(s) from your list") | Confirms how many pages were actually read and scored; worth a quick check against your input if the number looks off |
| Link distribution box | The minimum, maximum, and average number of inbound suggestions across every page in your list, plus confirmation that the zero-orphan guarantee held |
| "Suggested pages to link to" column | Up to three page addresses per row, the actual suggestions for that page |
| "Why (shared tags)" column | The specific tags two pages have in common, shown next to each suggestion; this is your evidence for whether the suggestion actually makes sense |
| A suggestion marked "bridged (0-orphan fix)" | This particular pairing shares no tags in common. The connection exists to satisfy the zero-orphan guarantee, not because of genuine topical overlap. Treat these with a bit more scrutiny than a suggestion backed by real shared tags before deciding whether it's worth a real link |
A quick way to prioritize when you have a long results table and limited time: start with rows whose suggestions carry the most shared tags, since those are the strongest, most self-evident matches. Come back to the "bridged" rows last, and only act on them if a quick manual look confirms the two pages are actually relevant to each other despite the tag mismatch.
This Tool Suggests. You Still Decide.
This point is worth stating plainly, more than once, because not every tool in this category works the same way. Some internal-linking tools, plugins that auto-insert links directly into your published content based on their own relevance scoring, will change your live pages without asking you to review each individual suggestion first. This tool does none of that. It only reads the data you paste in, computes similarity scores, and displays a report. It never writes to a database, never edits a page, and never touches your site's actual content in any way. The code behind it is documented as strictly read-only for exactly this reason.
Practically, that means the output you see is the entire effect of running this tool. Nothing changes anywhere until you, personally, go add a link somewhere, in your CMS, in your HTML, wherever your actual internal links live. That's a deliberate design choice, not a missing feature. An automated tool guessing at topical relevance from tags alone will occasionally be wrong, sometimes obviously so, and a wrong suggestion applied automatically is a genuinely bad link sitting live on your site until someone notices and removes it. A wrong suggestion sitting in a report you haven't acted on yet is nothing at all, just a line in a table you're free to skip.
Use the report as a work queue, not a to-do list to blindly clear. Read each suggestion, check whether the pairing genuinely makes sense to a human visitor landing on that page, and only then decide whether it's worth the few minutes it takes to add the link by hand.
Where Tag-Based Similarity Falls Short
Being upfront about the limits of this approach matters more than overselling what it does. A few honest caveats:
- Shared tags aren't the same as true topical relevance. Two pages can share a generic tag like "seo" or "checker" and still not be genuinely useful to link together, while two pages with zero tag overlap but a real conceptual connection won't get suggested at all. The math only sees what's in the tag data, nothing more.
- It doesn't read actual page content. This tool never crawls a live page or analyzes its text. It works entirely from the tags you provide. Garbage or overly generic tags produce garbage or overly generic suggestions; the output is only as good as the tag data behind it.
- It doesn't see your live internal-link structure. This tool has no idea which links already exist on your site right now. A suggestion might duplicate a link you've already added elsewhere, or miss the fact that a page is already well-linked through content this tool never saw.
None of this makes the tool useless. Every audit tool trades some accuracy for speed and ease of use, and the tradeoff here is a reasonable one for what this tool is actually for: a fast, free, first pass at internal-linking opportunities, not a definitive technical audit. It makes it a starting point rather than a final answer. For a small or medium site, or for a quick sanity check on a handful of pages, that's genuinely enough. For a large site where getting internal linking exactly right carries real stakes, a big e-commerce catalog, or content in a sensitive, high-scrutiny topic area, pair this report with a real crawl-based audit (Screaming Frog and similar tools read your actual live link structure and page content, at the cost of more setup) and, ultimately, human judgment about what a real reader on that specific page would actually find useful to click next.
One practical way to partly offset the tag-quality limitation: write your tags the way you'd describe the page to a colleague in a few words, not the way you'd stuff a meta-keywords field. Specific, concrete tags, such as air-fryer, chicken-wings, crispy rather than just recipe, food, produce noticeably more meaningful similarity scores than broad, generic ones, because generic tags match almost everything and specific tags only match what's genuinely close.
Putting Suggestions Into Practice
Getting a report full of suggestions is the easy part. Turning a handful of them into links that actually help your site takes a little discipline, and it's worth laying out concretely rather than leaving as "use your judgment."
Sanity-check relevance before anything else. Open both pages side by side, mentally, and ask whether a real visitor reading one would genuinely want to click through to the other. A suggestion with three or four strong shared tags is usually an easy yes. A suggestion with one generic shared tag, or a "bridged" label with no shared tags at all, deserves a slower look. Sometimes it's still a good link, sometimes it isn't, and the only way to tell is to actually think about the two pages rather than trust the row in the table.
Prefer natural, in-content anchor text over a forced link dump. The strongest internal links read naturally inside a sentence a visitor was already reading, something like "if you're also dealing with X, our guide to Y covers that," rather than a bare list of "related tools" bolted onto the bottom of a page with no context. A suggestions report tells you which pages to connect; it says nothing about how to phrase the actual link, and that part is worth real writing effort, not a copy-paste of the page title as anchor text.
Prioritize the pages with the lowest inbound counts first. The link-distribution box at the top of your results shows the minimum inbound count across your whole list. Pages sitting at that minimum, especially any that only got there through a "bridged (0-orphan fix)" suggestion, are your genuine priority. A page that's already receiving several suggested inbound links is in reasonably good shape; a page barely scraping by on one forced bridge is the one actually at risk of staying effectively orphaned in practice if nobody acts on the suggestion.
Don't link everything to everything just because a suggestion exists. A results table full of suggestions is not an instruction to add every single one. Over-linking a page with a dozen tenuously related links dilutes the value of every link on it and makes the page harder to read. A handful of genuinely relevant, well-placed links beats a wall of marginal ones, both for readers and for how link equity actually distributes: spreading it across too many weak connections thins it out rather than reinforcing the strongest, most relevant paths through your site.
Re-run the tool after a round of changes. Once you've added a batch of new internal links based on a first pass through the report, running the audit again, updating your pasted tag list if any pages changed, gives you a fresh read on whether your inbound-link distribution actually improved, rather than assuming it did.
Check the real-world effect with Google Search Console, not just this tool. Search Console's Links report shows your actual live internal-link counts per page, and Coverage/crawl-stats data can show whether a previously orphaned page starts getting crawled more often after you add real links to it. This tool's job ends at the suggestion stage; confirming whether a change actually mattered on the live site is a separate step worth doing for pages you cared enough about to prioritize in the first place.
A small, realistic scenario ties this together: say a run of the tool flags one of your published guides as sitting at the minimum inbound count, one suggested link, and that suggestion is a "bridged" one with no shared tags. Rather than adding that weak forced link, a quick look at the page's actual topic points to two other guides on your site that clearly cover adjacent ground but weren't in your original tag list at all, because you tagged narrowly the first time. The fix in that case isn't the tool's literal suggestion, it's improving the tag data and running the audit again with better tags, then adding the genuinely relevant links the second pass turns up. The tool's real value here was pointing at the weak spot, not dictating the exact fix.
Related Tools
A couple of other free tools on this site pair naturally with an interlinking audit, depending on where you are in the process. Running all three in sequence, on your own pages, gives a fuller picture of your internal-linking health than any single one on its own.
- If you want a tool focused specifically on catching pages with zero inbound links from a real internal-link graph you paste in (rather than suggesting new links to add), the Orphan Page Finder is the more direct diagnostic. It tells you which pages are currently orphaned; this tool tells you what to link them to.
- If some of your orphan or low-inbound pages turn out to be old URLs that have since moved, the Redirect Chain Checker confirms whether those URLs are routing cleanly or stacking up unnecessary hops before you invest time linking to them.
- Before you start pasting your own list together, the Website Links Counter is a quick way to get a first read on how many total links a given page already carries, as a sanity check before you decide it needs more.
Frequently Asked Questions
Does this tool actually add links to my pages?
No. It's strictly a report-only tool: it reads data, computes similarity scores, and displays suggestions. It never writes to your site, your database, or any page in any way. See the "This Tool Suggests. You Still Decide." section above for the full reasoning behind that design choice.
What exactly do I put in the two boxes?
In the left box, your pages, one per line (a full URL is ideal, but any label you'll recognize works). In the right box, on the same line as each page, that page's topic tags separated by commas. Line 1 on the left pairs with line 1 on the right, and so on. Topic tags are just short words you pick to describe each page, for example running-shoes, reviews, gear; you don't need pre-existing tags anywhere in your CMS.
How many tags should I give each page?
Three to six specific, genuinely descriptive tags per page is usually enough for the similarity scoring to produce meaningful matches. Too few tags gives the tool very little to compare against; overly generic tags (a single word like "tool" or "seo" repeated across every page) will inflate similarity scores between pages that aren't actually that related.
What does "bridged (0-orphan fix)" mean in my results?
It marks a suggestion where the two pages share no tags at all. The connection was added specifically to satisfy the guarantee that every page gets at least one inbound suggestion, not because the tag data itself supports a real match. The "Reading Your Results" section above covers how to weigh these compared to suggestions backed by genuine shared tags.
Is Jaccard similarity the same thing as Google's ranking algorithm or "topical relevance"?
No. It's a simple, transparent word-overlap calculation on whatever tags you provide, not a semantic understanding of page content and not anything Google itself uses to rank pages. It's a useful, explainable starting signal for finding candidate internal links, not a stand-in for actual topical or semantic relevance. The "Where Tag-Based Similarity Falls Short" section above covers this limitation directly.
My site doesn't use tags anywhere, can I still use this tool?
Yes. You don't need pre-existing tags in your CMS. Write a short, honest list of topic keywords for each page yourself when you build your list (three to six words that genuinely describe what that page covers). The tool doesn't care where the tags came from, only whether they accurately describe the page.