<b>The risk asymmetry between commercial and informational exact-match anchors</b>

<b>The risk asymmetry between commercial and informational exact-match anchors</b>

Not all exact-match anchors carry equal penalty risk, and treating "exact-match %" as one number erases the most important distinction.

The hypothesis: anchors matching high-commercial-intent queries ("buy cheap X," "best X casino bonus") sit closer to the spam decision boundary than anchors matching informational phrases ("how X works"). The reason is base rates — manipulators overwhelmingly target money terms, so the classifier has learned that commercial exact-match in volume is a strong spam prior. Informational exact-match is comparatively rare in spam, so it draws less suspicion.

On one hand, this is consistent with the lived pattern that gambling, pharma, and loan niches face anchor scrutiny at ratios that would be unremarkable in a hobby vertical. On the other, it's hard to disentangle anchor risk from the elevated baseline scrutiny those entire verticals receive regardless of anchors.

A 2021 vertical-segmented analysis found top-ranking commercial pages running notably lower exact-match ratios than top informational pages — circumstantial support for an intent-conditioned threshold.

Limitation: query intent is itself fuzzy and machine-classified, so any "commercial vs informational" split inherits that noise.

Practical read: budget exact-match far more conservatively on commercial targets than informational ones.

Open question: is the threshold conditioned on the anchor's query intent, the page's intent, or the vertical's spam base rate — and can these even be separated?