When the domain IS the anchor: an EMD double-counting case
Exact-match domains create a measurement problem rarely discussed. A 2023 audit of an EMD — the keyword literally in the domain — found that ~55% of its 'branded' anchors were also, by definition, exact-match keyword anchors, because brand and keyword were the same string.
The owners believed they had a safe, 55%-branded profile. Reclassifying brand-as-keyword anchors moved their effective exact-match ratio from a reported ~18% to ~52%.
The site had quietly slid from page 1 to page 3 over a year. After a deliberate campaign to add genuinely distinct anchors (URL, generic, partial-match unrelated to the head term), the effective exact-match share fell to ~30%, and positions recovered to mid page 1 within five months.
The lesson is classification, not tactics: exact-match domains can mask over-optimization because tooling counts brand and keyword as separate buckets when they're one string.
On one hand, the recovery tracked the reclassified ratio neatly. On the other, EMDs carry their own historical algorithmic baggage, confounding any anchor-specific reading.
Limitation: the 'effective ratio' is an analyst's reconstruction, not a measured quantity Google exposes.
Open question: does Google treat the EMD string in an anchor as branded, exact-match, or both simultaneously?
Anchor Theory
@AnchorTheory
When the domain IS the anchor: an EMD double-counting case
Этот пост опубликован в Telegram-канале Anchor Theory. Подписаться можно по ссылке: @AnchorTheory.