Methodology

How the FedComp Index score is calculated

The FedComp Index score runs 0 to 100. It ranks federal contractors by competitive positioning using public data only. It doesn't measure ethics, quality, or probability of winning any specific award. It measures two things: how much you've won, and whether you're still winning.

Two drivers
Driver Weight Signal
Award Volume 90% Total dollars won (past 5 years), log10-scaled to 0-100
Award Recency 10% When was the last award? Bucketed, not a daily decay.
Award Volume (90%)

Total contract dollars won over the past five years, run through log10 and mapped directly to 0-100. The scale covers $10K to $10B:

$10K0
$1M33
$10M50
$100M67
$1B83
$10B100

The log scale is not arbitrary. Without it, one $500M contractor mathematically crushes the entire field. With it, each order of magnitude gets equal representation. A $100M contractor and a $10M contractor are genuinely different. The score says so.

These are absolute values, not relative to the state pool. A score of 67 on volume means $100M in five-year awards regardless of who else is in the ranking.

Award Recency (10%)

A bucketed read on the last award date. It doesn't decay every day - it steps down when you cross a threshold:

Won in past 12 months100
1-2 years ago60
2-3 years ago30
3-5 years ago10

This driver is a tiebreaker, not a signal. At 10% weight, it can move a score by at most 10 points. Its job is narrow: separate "still actively winning" from "went quiet." A contractor with $100M who hasn't touched a contract in three years is a different proposition than one with $100M who won last quarter. Recency makes that cut. It doesn't pretend to do more than that.

Dollar amount is irrelevant here. Frequency is irrelevant. The only question is when.

Readiness Classes
Class 1Score 60+ - roughly $100M+ in 5-year awards
Class 2Score 40-59 - roughly $5M-$100M
Class 3Score below 40 - below $5M or minimal activity

Fixed score thresholds, not percentile cutoffs. Class 1 means the same thing regardless of how many contractors are in the state. Adding 1,000 new small contractors doesn't demote an existing Class 1 contractor. The class is anchored to performance, not rank within the pool.

Proximity Map

Each dossier shows a Proximity Map - the six contractors most likely competing for the same work. It uses NAICS and PSC codes from won awards, not SAM registration claims. Anyone can register for a NAICS code. Winning an award in it is different.

Two weights shape the match:

Rarity. A shared code gets weighted by 1/n2, where n is how many contractors have won awards in that code. Two contractors sharing a PSC code that only three companies have ever won - that's a strong signal. Two contractors sharing a code with 200 winners - that's background noise. The squared term makes rare overlaps dominate.

Scale. The overlap score gets multiplied by (min/max)2 of the two contractors' total award volumes. A $50M contractor and a $40M contractor get nearly full credit. A $50M contractor and a $500K contractor get almost none - they're not showing up on the same bids regardless of what codes they share.

Both must be strong or the match collapses. Sharing a rare code at a wildly different scale doesn't count. Sharing a common code at the same scale doesn't count. You need both.

Data sources
  • SAM.gov - entity registration, certifications, NAICS codes
  • USASpending (FPDS-NG) - award history, dollar amounts, agency, PSC
  • SBA - certification verification

The date in the footer shows the last scoring run.

What it doesn't measure

Win rates on specific solicitations, contracting officer relationships, proposal quality, pricing strategy, backlog, or anything requiring inside knowledge. The score comes entirely from public records. It tells you what happened, not why, and not what comes next.