What this internal link opportunity finder does
The tool compares the pages in your own export, removes source-to-target pairs that already exist when link data is available, and ranks the remaining pairs for editorial review. It is designed for site owners, publishers, ecommerce teams, agencies, and developers who already have a crawl, CMS export, or content inventory but do not have an affordable way to turn it into a practical internal-link backlog.
How to use it
- Export one row per indexable page from a crawler or CMS. Include URL, title, and body text when possible.
- Export existing internal links as source URL and destination URL pairs, or include an outlinks column in the page inventory.
- Upload or paste both files, confirm the automatically detected columns, and choose the relevance and target-need settings.
- Review the highest-priority source-to-target pairs, verify the proposed context, and export the approved queue for implementation.
Worked example
A local-business site has a detailed page titled How to Build Useful Location Pages, but the crawl shows zero internal inlinks. A separate article about citation cleanup contains the sentence “Citation work is one part of local SEO, alongside reviews, useful location pages, and Google Business Profile optimization.”
The tool can propose the citation-cleanup article as a source, the location-page guide as the target, and useful location pages as an anchor candidate found in the source. The target receives additional priority points because it has no detected inlinks. The editor still decides whether the sentence and reader journey justify the link.
What each result means
| Result | How to use it |
|---|---|
| Source page | The page where an editor should look for a natural contextual link placement. |
| Target page | The destination that may need another relevant internal link. |
| Priority points | A transparent sum of topic relevance, phrase evidence, target need, and source readiness. Open the breakdown before acting. |
| Matched terms | Terms contributing most strongly to the lexical similarity. These help spot false positives. |
| Anchor found in source | A phrase already present in source copy. It is evidence, not an instruction to force exact-match anchor text. |
| Editorial fallback | A concise target label to consider only when the source copy is revised naturally. |
| Warnings | Missing link data, weak copy, or unknown inlink counts that reduce confidence. |
Common mistakes to avoid
Analyzing titles without page copy
Titles can produce broad topical matches, but body text gives much better evidence for where a link may fit. Export rendered or main-content text whenever the crawler or CMS supports it.
Skipping the existing-links export
Without source-to-destination data, the tool cannot reliably remove links that already exist. The results remain useful for review, but they should not be treated as a clean implementation queue.
Adding every high-scoring link
A relevant pair may still be awkward in the exact paragraph, repetitive for users, or strategically unnecessary. Review the surrounding sentence and navigation path.
Treating anchor suggestions as exact-match requirements
Use concise, descriptive wording that makes sense in context. The candidate phrase is a starting point and may need editing.
Practical use cases
- Content refreshes: find older articles that can support a newly updated guide or product page.
- Orphan-page cleanup: identify contextually related sources for pages with zero detected inlinks.
- Site migrations: rebuild useful contextual connections after URL consolidation or content moves.
- Ecommerce: connect guides, categories, and product-support content without relying on a platform-specific plugin.
- Agency delivery: export a reviewable source, target, anchor, rationale, and warning queue for a client or content team.
Limitations
- The relevance model is deterministic lexical analysis. It does not understand meaning like a human editor and can miss synonyms or produce topical false positives.
- English stop words are removed by default. Unicode tokenization supports other scripts, but results may be noisier for languages with different word boundaries or morphology.
- The tool does not measure rankings, conversions, PageRank, traffic, or user behavior. It does not guarantee crawling, indexing, or ranking changes.
- JavaScript-rendered text must already be present in the supplied export. The tool does not crawl or render websites.
- Inlink counts from a crawler can differ from a CMS or search engine view. Use them as prioritization evidence, not an absolute truth.
Frequently asked questions
Does this tool crawl my website?
No. It analyzes a page inventory and optional internal-links export that you provide. This avoids account access, paid crawler APIs, and server-side crawling.
Are my crawl exports uploaded or stored?
No. The analysis runs in your browser. The page does not transmit or retain the uploaded or pasted data.
What columns are required?
Each page needs a URL plus either a title or body-content field. Inlink counts, status, indexability, canonicals, target keywords, business priority, page type, and existing outlinks are optional but improve prioritization.
Does a high priority score guarantee that a link should be added?
No. The points organize editorial review using topic overlap, phrase evidence, target link need, and source readiness. A person should still confirm that the link helps readers in its exact context.
Can I use exports from Screaming Frog, Sitebulb, or another crawler?
Yes. The column mapper accepts common crawl-export headings, and you can manually map other column names. CSV, TSV, semicolon, and pipe-delimited data are supported.
How the tool works and its limits
Last reviewed: July 14, 2026 · Tool logic version: v1.0.0
Reference sources
- Google Search Central link best practices for crawlable links, contextual internal linking, and descriptive anchor text.
- Schema.org WebApplication for software structured-data vocabulary.
- MDN FileReader and MDN Blob references for local file reading and downloads.
Tool-specific limits
- Analysis is based only on user-supplied exports and deterministic lexical evidence; it does not crawl, render, rank-check, or use private search data.
- Points and anchor candidates prioritize editorial review and do not guarantee crawling, indexing, traffic, or ranking changes.
- The 2,000-page ceiling should be retained until production browser testing covers representative content exports and lower-powered devices.