BERTlinker
Semantic internal linking analysis using Sentence Transformer embeddings. The hosted version was retired in September 2026.
The hosted version of this tool was shut down in September 2026 and is no longer available. This page is kept as a record of what it did. If you need this kind of analysis on your site, get in touch.
#about
BERTlinker found internal linking opportunities by meaning rather than keyword overlap.
It embedded every page title with a Sentence Transformer model, scored each pair of pages by cosine similarity, checked the live pages to see which recommended links already existed, and delivered the gaps as a per-page action plan in Excel.
It began as an open-source Streamlit app (Streamlit App of the Month, 2022) and ran as a hosted product at bertlinker.com from May 2026 until September 2026.
The hosted version is now closed.
I still run the same analysis as part of internal linking work for clients.
Award winnerUses crawl data
- Platform
- Browser-based
- Output
- Excel workbook with a per-page action plan, section-to-section link summary and all scored page pairs
- Category
- Internal Linking
Use cases
- Building topic clusters with contextual internal links
- Finding pages that gain no links from related content
- Improving site architecture based on content relationships
- Scaling internal linking for large e-commerce sites
#features
- Sentence Transformer embeddings of page titles
- Cosine similarity scoring between every pair of pages
- Live check of which recommended links already exist
- Per-page action plan grouped by site section
- Excel workbook with all results and a site map view
/contact
Want me to run this for you? I run this tool as a managed service, or build something custom around your data. You get the insights without touching the code.