Swapping anchors on existing links: a 40-link experiment

Swapping anchors on existing links: a 40-link experiment

A 2023 outreach team ran something rarely documented: they re-contacted publishers and changed the anchor text on 40 existing live links pointing to one URL, from keyword-rich to branded/natural, holding everything else constant — same pages, same placement.

The tracked URL had been stuck at positions 11-14 for its primary term for eight months. The anchor profile for that URL went from 62% keyword-match to 33% over the edit window.

Reported outcome at 12 weeks: the term moved to position 7, and two long-tail variants the page wasn't targeting entered the top 10.

Because the links themselves were unchanged — same authority, same context — this is closer to a controlled anchor experiment than most case studies. The main confound is time: Google re-crawled and re-scored independently of the edit, and eight months of aging may have mattered regardless.

Confidence: this is one of the cleaner natural experiments I've seen for anchor text in isolation, but n=1 URL and no control URL means it remains anecdote, not evidence.

On one hand, isolating the anchor edit is methodologically strong. On the other, the long-tail gains hint the page was simply maturing.

Open question: would a control URL, aged identically with no anchor edit, have moved too?