Prioritizing keywords with scores, KGR and CPC
These measures can organize research, not predict rankings or profit.
On this page
Direct answer
These measures can organize research, not predict rankings or profit. Volume × CPC ÷ KD mixes different measurements and is undefined at zero KD. KGR depends on unstable result counts and has no universal pass threshold.
When to use
You have many candidates and need a ranking system without mistaking a high score for validation.
Step by step
Preserve the raw metrics
Record volume, CPC, KD, methodology, market, and period. Keep missing, unknown, and zero values distinct; replacing zero KD with one creates arbitrary high scores.
Restrict formulas to research triage
Mark zero-denominator ratios not applicable and identify CPC as an advertiser-side signal. KGR is the estimated allintitle result count divided by monthly volume for the same market and period. A hypothetical 20 / 100 = 0.2 describes a sample ratio, not a pass threshold. Zero or unknown volume is not applicable; retain dates and count variation.
Apply nonnumeric decision gates
Check intent, accessible results, permissions, delivery costs, and differentiation. Reject an undeliverable high-score term; direct customer evidence may justify testing a lower-score term.
Worked example
Hypothetical exercise, not a real performance claim.
Suppose one term has volume 1,000, CPC 2, and KD 20; its ratio is 100. Another has volume 200, CPC 1, and KD 0; its ratio is not applicable, not an unlimited opportunity.
Action checklist
Reusable templates
Scoring review
Query / market / date:
Raw volume, CPC, KD, and tool:
Formula and zero / missing handling:
Research priority, not an earnings forecast:
Intent, delivery, permission, and cost checks:
Decision and human rationale:
KGR observation
Query:
Observed allintitle count, date, variation:
Volume tool / region / period:
Ratio or reason it is not applicable:
Actual result review:
Claims this cannot support:
Pitfalls
CPC is neither publisher RPM nor value per visitor. Result counts, formula thresholds, and alleged ranking-factor percentages are not official ranking rules.