<b>Exact-match anchors above 18% of new links correlate with a measurable plateau, not a penalty</b>

<b>Exact-match anchors above 18% of new links correlate with a measurable plateau, not a penalty</b>

Across 220 link-velocity cohorts (n=220), the story isn't a manual penalty — it's diminishing returns.

New-link anchor mix vs 6-month ranking trend:

— Exact-match under 8%: median +6.2 positions
— 8-18%: median +4.8
— 18-30%: median +1.1
— Over 30%: median -0.4

The inflection is sharp around 18%. Beyond it, each additional exact-match link buys almost nothing. This is consistent with a relevance-discounting model, not a penalty switch.

Directional — observational, no controlled test.

<b>Cap exact-match at roughly 1 in 6 new anchors and reallocate the rest to partial-match and branded.</b>