Inventory

Every customer, and where their weight sits

All 33 customers in the snapshot. These figures measure configuration footprint, not business importance — there is no revenue, claim volume, or patient count in this data. What they do measure well is how much work each customer represents.

The inversion

ECP has the most SQL calculations of any customer and looks worst-blocked. But those uses are four distinct bodies, and nearly all are one modifier-55 rule. Femwellis mid-tier by SQL uses and has 123 distinct bodies, 119 of them needing legacy ETL data — by a wide margin the hardest customer, and counting uses hides it completely.

CustomerFee logicsCalcsTypesSQL bodiesBlockedOverlapsCannot fire
ECP1914,838234148
OneGI1573,3742261027
SGP1413,4201944011
MNEye1401,85521580517
OrthoAlliance1082,2982682817
RIA1082,425212173
Commonwealth791,9072120037
EyeSouth6760819179013
ARSA47689161009
Femwell436681512311909
EugeneGI4211140004
PAA424327710
StrideCare41776192103
HeatonEye40826213106
USDH39908160002
PrecisionHealth317351721030
TFOC305951060030
JewishBoard29379155301
MiddletownMedical28475181006
SeaRad26246124010
ISpine25554181000
TheOregonClinic242491700021
USHP23502203100
MannEye1830680010
SouthSoundRadiology182981812022
CCP155383300
NWDerm14198140000
HillsboroEye12308171001
Monarch12213110001
SWOrofacial10174143010
SunLife71640000
PPRM3321101
Wildflower2210000

SQL bodies = distinct queries to port, the real effort driver.Blocked = of those, how many need legacy ETL or payer-crosswalk data the replacement cannot reach. Overlaps = contract-map conflicts, each an onboarding blocker. Cannot fire = fee logics with no contract map, or one that selects nothing.

Clean on both axes

7 customers have no SQL calculations and no contract-map conflicts:EugeneGI, USDH, TheOregonClinic, NWDerm, Monarch, SunLife, Wildflower.

Concentration, everywhere

The estate is heavily skewed at every level. One adjustment type is nearly half of all calculations. Two SQL bodies are a third of all SQL. 172 of 268 SQL bodies are used exactly once. Plans built on averages will be wrong in both directions.