FAQ

FedComp Index is an open source competitive intelligence platform for federal contractors. It pulls five years of award data from USASpending.gov, scores every contractor in a state from 0 to 100, and publishes the results as public dossier pages.

Each contractor gets a FedComp Index score based on two drivers: award volume (90% weight, log-scaled) and award recency (10% weight, bucketed). The score determines a posture class - Class 1 for scores 60 and above, Class 2 for 40-59, Class 3 for below 40.

Nevada is the first state with 349 scored contractors. The data refreshes daily. Rankings, individual dossier pages, and proximity maps are all generated automatically from public records.

Posture class is a classification used by FedComp Index to group federal contractors by competitive positioning. It's based on the FedComp Index score, which measures actual award performance, not eligibility or certification status.

Three classes:

• Class 1 - FedComp Index score 60 or above. Contractors with roughly $100M+ in federal awards over 5 years.
• Class 2 - Score 40-59. Roughly $5M-$100M range. Active mid-tier contractors.
• Class 3 - Below 40. Emerging contractors or those with lower award volume.

The thresholds are fixed. A contractor moves between classes only when their score changes enough to cross a boundary. Posture class is not a certification, not a set-aside category, and not based on company size.

Two index drivers. No normalization, no percentile ranking. The score is absolute - a contractor scoring 60 represents roughly the same award volume regardless of state.

Award volume (90% weight): total federal award dollars over the trailing 5-year window, log10-scaled and mapped to 0-100. Reference points: $10K maps to 0, $1M to 33, $10M to 50, $100M to 67, $1B to 83.

Award recency (10% weight): how recently the contractor won their last award. Past 12 months scores 100, 1-2 years scores 60, 2-3 years scores 30, 3-5 years scores 10.

The final score is: (volume driver x 0.90) + (recency driver x 0.10), clamped to 0-100.

Full methodology: fedcompindex.org/wiki/methodology

The proximity map on each contractor's dossier page shows which other contractors compete in similar spaces. It answers: who are you actually up against?

It works by comparing overlapping NAICS codes and awarding agencies between contractors. Two contractors who share the same obscure NAICS code are considered more proximate than two who share a common one, because a rare code overlap is a stronger signal of direct competition. Contract scale also factors in.

The map is directional. Your proximity map shows who you're likely competing against based on the awards you've won. Their map might look different.

Index drivers are the individual components that make up the FedComp Index score. Each driver measures one dimension of competitive positioning and contributes a weighted percentage to the final score.

Award Volume (90%) measures total dollars won over 5 years. A contractor with $100M will always score higher than one with $10M. Award Recency (10%) rewards contractors still actively winning. These are displayed individually on each dossier page so you can see exactly what contributes to a score.

The Tabularium is the data and resources section of FedComp Index. It collects all datasets, Python packages, and source links in one place.

Datasets are available on HuggingFace and Kaggle in CSV and Parquet format. The scoring engine and bundled data are on PyPI as fedcomp-index-scoring and fedcomp-index-data. The full source code is open on GitHub.

The name comes from the Tabularium in the Roman Forum - the state archive where official records were stored.

SAM.gov is a registration database. It tells you a contractor exists, their certifications, and their NAICS codes. USASpending.gov is a transaction database. It tells you what awards a contractor has won.

Neither provides scoring, ranking, or any way to compare contractors against each other. FedComp Index takes both sources, joins them, and produces the FedComp Index score, posture class, proximity map, and dossier pages. The raw data is the same public data anyone can access. The value is in the aggregation, scoring, and presentation.

Multiple platforms, multiple formats:

• HuggingFace - Nevada contractors
• Kaggle - Nevada contractors
• PyPI: pip install fedcomp-index-data

Key columns in the data: fedcomp_index (score 0-100), posture_class (Class 1/2/3), awards_5yr_total_usd, award_count, last_award_date, primary_naics.

All MIT licensed. Updated monthly.

Every scored contractor gets a public dossier page. It's a single-page competitive profile built from public data.

Each dossier includes: FedComp Index score and posture class, index drivers breakdown, spectator count (how many times the page has been viewed), contract history with dates and agencies, proximity map showing competitive overlap, certifications, and rank within the state.

Nevada currently has 349 dossier pages. Each one updates daily when new award data comes in. The URL is permanent.

Daily. The pipeline pulls incremental award data from USASpending Monday through Saturday. Full re-pull on Sundays. Scores recalculate and the site rebuilds automatically after each pull.

Datasets on HuggingFace and Kaggle update monthly. PyPI packages update monthly. The GitHub Wiki regenerates daily.

There's an antifragile gate in the pipeline - if a data pull produces more than 5% fewer contractors than the previous run, it rejects the new data and restores from backup. This protects against API failures or data corruption.