Picture a startup founder, eighteen months into building a product, sitting across from a patent attorney who slides a quote across the table: $15,000 to file a single software patent. The founder winces, writes the check anyway, and files something related to a recommendation algorithm they built. Three years later, that patent sits in a database, untouched, generating zero revenue and blocking zero competitors. The founder couldn’t even explain what it covers anymore. This scenario plays out thousands of times a year, and yet companies keep filing. The question worth asking is: why?
Tech companies routinely invest in assets that look irrational on the surface but serve deeply strategic purposes. Software patents are one of the most expensive and least understood versions of this pattern.
The Patent as a Signaling Device
Here’s the part that patent attorneys don’t lead with in their pitches: the majority of software patents are filed not to be enforced, but to be counted.
Venture capitalists, acquirers, and enterprise sales prospects all ask the same question during due diligence: how many patents do you hold? It’s a proxy metric for innovation, the way follower counts became a proxy for influence. It doesn’t measure the right thing, but everyone uses it anyway because it’s easy to measure.
A company with forty patents looks more defensible than one with four, regardless of whether those forty patents would survive a legal challenge. IBM famously files thousands of patents per year, many of which exist purely to pad portfolios and generate licensing revenue from companies who settle rather than fight. But for startups, the calculus is even more cynical: patents are bought to impress people who don’t read them.
This is signaling economics. The patent isn’t the product. The number is.
The Defensive Stockpile Strategy
There’s a second reason companies file patents they never use, and it’s actually more rational than the signaling game. It’s called defensive accumulation, and it operates like nuclear deterrence.
The logic goes like this: if a larger competitor decides to sue you for patent infringement (and in software, almost everyone is technically infringing on something), your best defense is a counter-threat. You file suit claiming they infringe on your patents. Both sides suddenly face expensive litigation, and the incentive to settle out of court rises dramatically.
This is why Google spent $12.5 billion acquiring Motorola in 2012. The products were largely irrelevant. They were buying 17,000 patents as a legal shield. The acquisition was a patent stockpile dressed up as a hardware strategy.
For smaller companies, the same logic applies at a smaller scale. You don’t need to win a patent lawsuit. You need to make winning expensive enough for the other side that they back off. A portfolio of ten questionable patents can accomplish this almost as well as two airtight ones, because legal fees are the real deterrent, not legal merit.
Early-stage startups increasingly treat strategic assets as cheat codes, and patent portfolios fit squarely into that playbook when they’re used correctly.
Why Most Software Patents Wouldn’t Survive a Fight
Here’s the uncomfortable truth that sits beneath all of this: a significant portion of filed software patents are, to use the technical legal term, junk.
Software patents are notoriously difficult to scope correctly. The claims need to be broad enough to be valuable but narrow enough to be novel. Patent examiners are overworked and under-resourced, reviewing applications in technical areas where the prior art landscape is enormous and constantly shifting. The result is that patents regularly get approved that would be invalidated the moment a well-funded legal team started pulling at the threads.
Companies know this. Their own patent attorneys know this. But invalid patents still scare small competitors who can’t afford the discovery process to prove invalidity. The threat is the weapon, not the patent itself.
This is structurally similar to something we see in other parts of the tech industry, where the surface-level product and the actual strategic mechanism behind it are entirely different things. Tech giants hide their most powerful retention tools so well that users call them conveniences. Patents work the same way. The stated purpose (protecting innovation) and the actual purpose (creating friction and fear) barely overlap.
The Cost Nobody Talks About
All of this creates a real drag on the ecosystem that doesn’t show up in any company’s quarterly report.
Small developers and startups routinely avoid building certain features or entering certain markets not because they couldn’t compete technically, but because the patent landscape is too murky to navigate without a legal budget they don’t have. This is called the patent thicket problem, and it’s particularly severe in software where overlapping claims are everywhere.
The irony is that the companies most hurt by patent thickets are often the ones most likely to drive genuine innovation. The incumbents with massive patent portfolios aren’t using those portfolios to build better products. They’re using them to make the game expensive enough that challengers think twice.
When you add up the filing fees, the attorney costs, the maintenance fees paid annually to keep patents active, and the internal legal resources required to manage a portfolio, a mid-sized tech company might spend millions per year on patents that do nothing except exist. That’s capital that isn’t going into engineering, product, or customers.
What Would Actually Protect Innovation
The honest answer is that trade secrets and speed-to-market protect software innovation far more reliably than patents do.
By the time a software patent is filed, examined, and granted (often eighteen months to three years after filing), the technology landscape has frequently moved on. The feature that felt like a competitive moat at filing might be table stakes by grant date. Meanwhile, a company that invested those same legal dollars into shipping faster and building a better product would have a more durable advantage.
The most profitable startups often win by doing the thing that looks strategically boring, and in the patent context, that means spending less on legal theater and more on building things worth copying in the first place.
The founders who understand this game don’t opt out of patents entirely. They file selectively, target the claims most likely to survive challenge, and treat their portfolio as a negotiating tool rather than an enforcement mechanism. They know the patent is probably never going to court. They’re filing it to check a box on someone else’s due diligence form and to make a potential aggressor do a little math before picking a fight.
It’s not glamorous. It’s not what gets taught in innovation policy discussions. But it’s what actually happens, and understanding it changes how you think about the entire system.
