How Government Contract Data Reveals Industry Dynamics
The Biggest Customer Nobody Talks About
Total US federal spending runs into the trillions of dollars a year, and hundreds of billions of that goes out the door as contracts for goods and services. Software, satellites, consulting hours, janitorial work, cloud capacity. The federal government is routinely described as the largest single buyer on the planet, and unlike any private customer of that size, it publishes what it buys. Every contract action gets documented, categorized, and dropped into public databases that anyone can search for free.
Hardly anyone in the investing or strategy world actually uses them. Analysts pay real money for web-traffic panels and credit-card data while ignoring a procurement record that shows, down to the dollar, who is winning work from the biggest customer in the world. If you care about industry dynamics, competitive positioning, or a specific company's revenue pipeline, this data deserves a spot in your process. Here's how I actually use it.
Where the Data Lives
Start with USAspending.gov, which tracks all federal spending, including contracts, grants, and loans. It's searchable by contractor name, NAICS industry code, awarding agency, dollar amount, geography, and time period. The Federal Procurement Data System (FPDS) sits underneath it with more granular contract-level detail when you want to go deeper. SAM.gov rounds out the trio by listing upcoming solicitations, so it tells you where money is about to go instead of where it already went.
Between those three sites you can pull every contract a company has won, see who else wins the same kinds of work, and chart spending by industry category across a decade. USAspending also offers bulk downloads and an API, so once you know which codes you care about, you can automate the pull instead of clicking through search screens.
State and local procurement data exists too, though it's fragmented across hundreds of portals. Commercial aggregators like GovWin from Deltek stitch it together, and many state procurement sites publish award data directly.
Reading Industry Structure from Spending Patterns
Spending trends by NAICS code work as a rough map of where policy priorities are heading. Pick a code like 541512, computer systems design services, and chart federal obligations over ten years. When the dollars flowing into a category climb year after year, someone inside the government has decided that category matters, usually based on analysis the public never sees.
Cybersecurity is the clean example. Federal security budgets were climbing years before cyber became a boardroom topic in the private sector, so the procurement record served as an early read on the entire category. Defense spending shifts tell you about manufacturing capacity and technology priorities. Healthcare contracts through CMS and the VA show where the government sees the biggest gaps in medical technology and services.
The reason this works is timing. Procurement decisions reflect threat assessments and policy analysis that markets haven't priced yet, and by the time a trend shows up on earnings calls, the government has often been funding it for years.
Concentration is the other structural read. In most industries you estimate market share from surveys and proxies, but with government contracts you can compute it exactly, because you have the actual dollars each company received. When a handful of large primes win a growing share of awards in a category, that industry is consolidating. When the government leans into small business set-asides and pushes vendor counts up, it is deliberately injecting competition into a concentrated market. Both moves show up in procurement data long before they make it into an industry report.
Company-Level Competitive Analysis
Say you're analyzing a mid-cap defense contractor. Pull its history on FPDS and you get dollar amounts, awarding agencies, contract types, and performance periods for everything it has won. Run the same query on its competitors and you have genuine market share within specific government segments, built from actual awards instead of estimates. Say two rivals have split an agency's IT budget for years and one suddenly starts taking a visibly larger share of new awards. You just learned something the loser's investor relations team won't volunteer.
Contract types carry information on their own. Fixed-price work rewards efficient operators and punishes sloppy ones, while cost-plus work shifts risk back to the government and caps the upside. A portfolio drifting from cost-plus toward fixed-price says something about future margin volatility that the income statement will only confirm several quarters later. And a category moving from single-award to multiple-award vehicles means the government wants more competition, which usually means incumbents are about to lose pricing power.
Sole-source awards deserve special attention. When a company keeps winning non-competitive contracts, it almost always holds capabilities, clearances, or certifications its rivals lack. That's a genuine moat, and you'll never find it in a standard financial screen.
Forecasting Revenue from Backlog and Pipeline
Government contracts come with defined performance periods, option years, and ceiling values, which makes revenue at contract-heavy companies unusually forecastable if you're willing to do the work.
Start with the existing contract base. Say a contractor holds a five-year award worth $500 million including options. That's roughly $100 million a year of revenue visibility before the company wins anything new. Build that up across the whole portfolio and you have a bottoms-up backlog model. Then layer on the pipeline. Companies that win in a category tend to keep winning there, so if a firm historically converts about a third of what it bids and SAM.gov shows a heavy slate of upcoming solicitations in its core categories, you can sketch the bookings scenarios yourself.
This approach works best where federal work dominates the business, at names like Leidos, Booz Allen Hamilton, SAIC, and CACI. It also gives you a read on the government segments of Palantir, Microsoft, and Amazon, businesses those companies disclose only coarsely. Cross-check your bottoms-up numbers against the backlog and book-to-bill figures management reports in the 10-K on EDGAR. Big gaps usually mean you've misread option years or ceiling values.
On that last point, a ceiling and actual spending are different numbers. An IDIQ vehicle with an enormous headline value might only ever see a fraction of it obligated, so build your model on obligated dollars rather than announced totals.
Early Signals from RFPs, RFIs, and Sources Sought Notices
Before the government awards anything, it telegraphs intent. A Request for Proposal describes exactly what an agency needs, how it will score bids, and which capabilities matter most. The requirements section is a snapshot of the state of the art as the buyer sees it, and the evaluation criteria tell you whether price, technical capability, past performance, or small business participation will decide the award.
The scoring method itself tells you about the economics. A solicitation run as lowest price technically acceptable means the agency treats the work as a commodity, and margins will behave accordingly. A best value tradeoff means incumbents with strong past performance have the inside track.
RFP language also drifts over time, and the drift is informative. When cybersecurity requirements start appearing in solicitations for categories that never used to mention security, the threat picture has changed and budgets will follow.
Earlier still, agencies publish Sources Sought notices and Requests for Information before a formal solicitation exists. These often land months, sometimes a year, ahead of the actual procurement, which makes them one of the earliest public signals of where spending is headed. People who cover government markets seriously track them the way equity analysts track management guidance.
Subcontractors and Teaming Arrangements
Prime awards tell only part of the story, because the subcontracting layer underneath is where the real supply chain lives. Subaward data is patchier than prime data, but you can find it. Larger contracts require primes to report first-tier subcontractor spending, some subawards show up on USAspending.gov, and companies disclose major teaming relationships in SEC filings and investor decks.
Teaming patterns are worth mapping. Two companies that repeatedly bid together have complementary capabilities and a working relationship. A longtime subcontractor that starts bidding as a prime on similar work is telling you the competitive structure of that niche is shifting, and the incumbent now faces a rival that knows its playbook from the inside.
Private Companies Leave Public Footprints
This data is disproportionately valuable for researching private companies. Private firms owe the public essentially no financial disclosure, but if they sell to the government, their contract history is public record anyway. Sum a private contractor's awards over a period and you have a floor estimate of revenue. For firms that are mostly government business, that estimate lands close to the true number. For mixed commercial and government firms, it still hands you one solid piece of an otherwise dark puzzle.
Private equity teams use this on acquisition targets. Competitors use it to size rivals they compete with for talent and task orders. And when a government-heavy private company eventually files to go public, you can check the prospectus against a procurement record you already know.
A Simple Workflow, and the Honest Caveats
If you want to fold this into your research process, the sequence is short.
- Identify the NAICS codes relevant to your industry or target company.
- Search USAspending.gov or FPDS for spending trends in those codes over the past five to ten years.
- For a specific company, pull all of its awards and sort by dollar value and awarding agency.
- Check SAM.gov for upcoming solicitations, Sources Sought notices, and RFIs in the same categories.
- Watch recompetes. Formal performance ratings live in CPARS, which isn't public, so the best outside proxy for quality is whether incumbents keep winning their own work back.
Go in knowing the limitations. Classified contracts never appear. State and local data is scattered. The public databases lag actual awards by weeks or months, and headline ceiling values routinely overstate what actually gets spent. Even so, this is detailed, continuously updated information about the largest buyer on earth, it costs nothing to access, and most of the people you're competing with for insight have never opened it.