Every value here feeds the scoring engine directly — no deploy
needed. Edits are validated on save, versioned (the config
fingerprint is stamped on every evaluation) and logged, so a shift in
verdicts is always traceable to a settings change. Fields that differ
from the file defaults are highlighted; "Reset all" returns everything
to defaults.
01
Coverage factors
Observed spend ÷ factor = estimated total spend. The single biggest
lever in the tool. Mark a vertical calibrated only after
the Phase 1 pass against real client spend.
02
Ad-to-sales ratios
Estimate B = annual revenue × ratio × working-media share (the
share is under Thresholds). Ratios are advertising
expense to sales, never marketing-to-revenue; current values are
placeholders (R9) pending the SEC-derived table.
03
Proxy benchmarks
Used only when a channel has no direct measurement (no AdClarity
export): creative counts × $/creative, visits × paid share × CPC.
04
Capacity modifiers
Additive weights; net clamped below. Group names are logged for the
double-counting review (R10) — they do not change scoring.
05
Compatibility flags
Weights are stored positive on BOTH sides — reds are subtracted by
the formula (sign convention, R6). The neutral baseline is computed
from the checked flags and subtracted from every verdict (R5).
Neutral baseline (computed): — · net = greens − reds − baseline
06
Thresholds & bands
Verdict cut-offs, tier bands, clamps, checks. Agency
margin: the pillars estimate working media; when enabled the
headline spend and tier are grossed up by the rate so they compare
with client billings, which include margin.
07
Top verticals — static fallback
Used only when the live NetSuite query returns nothing. Tick
confirmed to activate the list.