How I Track $500B in US Federal Contracts in Real Time and What I've Learned So Far

The U.S. federal government publishes every contract it awards. Daily. For free.

The DoD releases a list every afternoon. SAM.gov tracks every civilian agency award in near real time. USASpending.gov holds a historical record going back to 2008. Every dollar, every company, every contract, fully public.

The problem is that the data is a firehose with no filter.

Why I built this

I'm an investor and quantitative trader. My edge hypothesis is simple: knowing things before other investors leads to better returns. Government contract awards are one of the few categories of genuinely public, genuinely actionable information that almost no one is processing in real time. Bloomberg Government does it for $8,000 a year. Outside of that, retail investors are largely on their own.

So I built GovEdge. The goal was not to create new data, the data already exists. The goal was to structure and organize it so you can orient yourself in minutes instead of hours. While everyone else is lost in PDFs and procurement databases, I wanted a single place where you could see what's happening, which companies are winning, and how significant those wins actually are.

The subsidiary problem nobody talks about

Something that surprised me early on is that most federal contracts are not awarded to the company you think.

When the DoD awards a $2B contract to "Lockheed Martin Missiles and Fire Control," that's a subsidiary. When a cybersecurity contract goes to "Leidos Innovations Corporation," also a subsidiary. The public parent company is several layers up, and the raw government data does not connect those dots.

This means that if you're reading raw procurement data, most of it looks like it's going to entities you've never heard of. Building the mapping layer, linking subsidiaries to their publicly traded parents, turned out to be one of the hardest parts of this project and also the most valuable. Without it, the data is noise. With it, every contract award becomes a data point in a company's actual financial story.

The most important thing I've learned

After ingesting thousands of contract awards, the single most useful metric is also the simplest one: what percentage of a company's annual revenue does this contract represent?

A contract worth 1% of annual revenue is interesting. A contract worth 10% is significant. Anything above that is the kind of event that should make investors pay attention.

And in practice, it does. These awards don't take months to get priced in. They reprice in days. The advantage GovEdge creates is identifying them in hours, often the same morning the award is published.

Most contract awards flowing through the system on any given day are routine:

• Modifications
• Extensions
• Small task orders

They matter in aggregate but no single one moves the needle. The meaningful events are rare, maybe one or two a week that rise to that level across the entire market. That rarity is exactly what makes them valuable. If high-revenue-impact contracts happened daily, they'd be priced in immediately. Because they're infrequent, catching them early is a real edge.

What this means in practice

The contracts that matter most are the ones that are large relative to the company receiving them. Not just large in absolute dollar terms, but large as a fraction of what that company earns in a year. That single filter cuts through most of the noise and surfaces the handful of events each week worth actually acting on.

GovEdge is built around this idea. Every contract in the feed is scored by its revenue impact, so the signals worth paying attention to rise to the top automatically.

---

The data has always been public. We just made it usable.