Diluting a 41% exact-match ratio: one recovery timeline

Diluting a 41% exact-match ratio: one recovery timeline

A documented case from a B2B SaaS site (shared in a 2023 link-audit teardown) is worth examining because the numbers are unusually clean. The domain ranked 3-5 for its head term, then dropped to page 4 across 60+ keywords over a six-week window in 2022.

The audit found 41% of ~380 referring domains used the exact-match commercial anchor. The intervention had two arms:
— Disavow of 90 lowest-quality links carrying that anchor (roughly 24% of the EMD pool)
— A deliberate brand-anchor campaign that added 70 new links, pushing branded share from 18% to 44% over four months

Recovery to prior positions took ~5 months. Average position on the tracked set moved from 31 back to 6.

On one hand, this looks like a textbook over-optimization correction. On the other, the site also published 14 new pages and earned coverage during the same window, so link velocity and content freshness are confounded with the anchor change. We cannot isolate the anchor variable.

Limitation: single domain, self-reported, no control. What it suggests — not proves — is that the direction of the ratio shift (down on EMD, up on branded) coincided with recovery, consistent with Penguin-era observations.

Open question: was disavow or dilution the active ingredient, or were both downstream of the content work?