KD ignores your own site, so it lies about what you can rank for

KD ignores your own site, so it lies about what you can rank for
Difficulty score measures the competition, not the gap between them and you. Across 1,000 terms we won and lost:
— Won terms, median KD ▇▇▇ 38
— Lost terms, median KD ▇▇ 29 (lower, yet lost)
— Won terms where our topical depth was higher ▇▇▇▇ 71%
— KD's correlation with our actual win rate ▇ 0.18
A low KD on a topic your domain has no authority in is still a wall.
The mistake: picking targets by absolute KD with no relation to your own footprint.
The fix: weight KD against your existing rankings in that cluster; a KD 40 term inside a topic you own beats a KD 20 term cold.
Benchmark of the week: KD alone explained under 4% of variance in our win rate.