Why Most Patent Maintenance Decision Processes Burn Through Budget Without Protecting Anything Real

Patent maintenance fees are a slow bleed that most tech transfer offices treat as a fixed cost of doing business. They aren't. Every annuity payment on an unworked patent is a choice, and most universities are making that choice on autopilot.

Photo by Tibor Szabo on Pexels.

The numbers deserve attention. A US patent can cost $15,000–$25,000 in total maintenance fees across its life, before you add prosecution costs or foreign filing fees. A mid-sized research university might hold 800–1,200 active patents at any given time. Do the math. A substantial fraction of that portfolio will never generate a dollar of licensing revenue, yet the fees keep flowing because no one built a real process for deciding otherwise.

Why does this happen? Three reasons, mostly.

First, the people making maintenance decisions are usually the same people who prosecuted the applications. Ask a patent attorney whether to abandon a patent they spent three years drafting and you'll get a predictable answer. The sunk-cost logic runs deep, and the discomfort with abandonment is genuine. Nobody wants to be the one who dropped a patent that turned out to matter.

Second, most universities review maintenance decisions on a purely calendar-driven schedule: a fee comes due, someone runs a quick check on licensing status, and the default answer is "pay it." The burden of proof sits on abandonment rather than continuation. Flip that burden and the calculus changes immediately.

Third, the criteria used for continuation decisions are almost always backward-looking. Is there an active license? Has there been recent interest from industry? These questions tell you about the past. They don't tell you whether the technology has a plausible commercialization path in the next three to five years, which is the only question that matters for a maintenance decision.

Here's what a more honest process looks like:

graph TD
    A[Fee Due Notice] --> B{Active License or LOI?}
    B -- Yes --> C[Pay: Obligation Exists]
    B -- No --> D{Credible Commercialization Path?}
    D -- Yes --> E{Market Exists Within 3 Years?}
    D -- No --> F[Abandon]
    E -- Yes --> G[Pay: Strategic Asset]
    E -- No --> F

The key move is the "credible commercialization path" gate. Not "could someone theoretically use this," not "the inventor thinks it's important." A credible path means: there is an identifiable customer segment, a realistic licensing candidate, or an active spinout that needs this claim scope. Without one of those three, you're paying maintenance fees on hope.

Foreign portfolio decisions make this worse. PCT filings routinely proceed into national phase on patents that have zero prospect of generating international revenue, largely because the decision gets made at 30 months without proper commercial review. National phase entry into Germany, Japan, and Australia costs real money. The technology transfer office rarely has the market data to challenge the filing recommendation at that moment, so the default is to proceed.

One structural fix worth considering: decouple the maintenance decision from the patent attorney relationship entirely for assets with no active commercial activity. Assign a licensing manager to run a brief commercialization screen on every patent facing a major fee decision (three-and-a-half year, seven-and-a-half year, and eleven-and-a-half year in the US). The screen doesn't need to be elaborate. It needs four data points: current licensing status, the last meaningful industry inquiry date, the technology readiness level, and the competitive patent landscape. Twenty minutes of work. Most decisions will be obvious.

The pushback from inventors is predictable and worth taking seriously, up to a point. Inventors often have the best market intelligence on their technology because they're still active in the field. Their opinion should be solicited. Their vote shouldn't be decisive. The conflict of interest is too direct.

Some universities have experimented with defensive publication as an alternative to abandonment: publishing the technology to create prior art, blocking competitors without paying maintenance fees. This is underused and often misunderstood. Defensive publication makes sense for process improvements and incremental advances where blocking competitors matters but licensing revenue is unlikely. It doesn't require patent counsel, it's nearly free, and it doesn't carry the psychological weight of "abandonment."

Portfolio pruning conversations stall because everyone focuses on what might be lost. Shift the frame: ask what you'd file today if you were starting fresh with the same budget. Most universities would hold a third of the patents they currently maintain. That gap between what you'd choose and what you're paying for is the real cost of a bad maintenance process.