USAspending.gov contains every federal contract award since 2008. About $650B per year flows through it. The site is awkward and slow, which is why most contractors skim the front page and walk away. The contractors who actually mine USAspending have a structural advantage over the ones who don’t.
This is the playbook.
The data model
Every USAspending row is one of four record types:
- A — Definitive contract (most common)
- B — Indefinite-delivery / IDIQ definitization
- C — Indefinite-delivery / GWAC / BPA award
- D — Definite-delivery / fixed-price award
For BD purposes, type A and D are where you mine for “what got awarded.” Type B and C are vehicle definitizations — they’re relevant for “who’s on which IDIQ” but don’t represent dollars that have actually flowed.
Five queries every BD lead should run weekly
-
My competitors’ last 30 days. Filter: recipient name = [competitor]. Date: posted in last 30 days. Result: every contract they won last month. You learn what they’re pursuing and what they’re winning.
-
Recompete window for my agency. Filter: agency = [your top buyer]. Date: posted 24-36 months ago. Result: contracts that will likely recompete in the next 12 months. Sort by ceiling — the biggest are highest-priority.
-
My NAICS, last quarter, by recipient. Filter: NAICS = [yours], date: last quarter. Group by recipient. Result: who’s winning in your space. The top 10 are your benchmark.
-
Set-aside saturation in your NAICS. Filter: NAICS = [yours], set-aside = [your eligibility]. Date: last year. Group by recipient. Result: how concentrated set-aside dollars are. If 80% goes to one firm, that’s a defended position. If it’s spread across 20 firms, the market is more contestable.
-
Geographic concentration. Filter: NAICS = [yours], by state of performance. Date: last year. Result: where the dollars actually land. Federal contracts are physically located somewhere, and that’s where customer-facing work happens.
What USAspending won’t tell you
Three blind spots:
- Recompete intent. USAspending shows the contract was awarded; it doesn’t show whether the agency intends to recompete vs. extend. That signal lives in budget documents and forecasts.
- Sub-contract flow. USAspending captures prime awards. The sub-contract flow under each prime is reported through FFATA but coverage is patchy. For task-order detail, FPDS-EZ (now folded into SAM.gov) is the better source.
- Pre-solicitation activity. RFIs, sources sought notices, draft solicitations — none show in USAspending. Those live in SAM.gov and the agency’s eBuy / FedConnect surfaces.
How VectorBrief uses USAspending
We mirror USAspending into our local Postgres + pgvector index. Every opportunity in our feed is enriched with the recipient’s last-12-month USAspending history, the agency’s last-3-year award pattern, and the incumbent (if any) inferred from prior awards on the same vehicle. That’s the data layer that drives our 5-factor scoring. /pricing.