Why Most Milestone Structures in License Agreements Reward the Wrong Behavior

Milestone structures are supposed to keep licensees honest. Pay a fee when you hit clinical Phase I. Pay another when you file your first NDA. Miss the dates, lose the license. Clean, simple, accountable.

Photo by SHOX ART on Pexels.

Except they almost never work that way in practice.

What happens instead: licensees negotiate every milestone date to be as late as legally defensible, build in termination rights that let them walk before any real money is due, and treat the milestone schedule as a liability to be managed rather than a roadmap to be followed. The university gets the illusion of accountability. The licensee gets optionality. Neither party ends up where they expected.

The problem runs deeper than bad negotiation habits.

Milestones are usually written to track regulatory events, not commercial progress.

Phase I completion. Phase II initiation. First commercial sale. These are real events, but they are lagging indicators. By the time a licensee misses a Phase II milestone, you already know the program has been deprioritized for 18 months. The milestone didn't predict that. It just confirmed it.

Worse, regulatory milestones create perverse incentives in earlier-stage deals. A licensee working on a software platform or agricultural biotech product has no Phase I to hit. So the milestone schedule gets adapted awkwardly: "first paying customer" or "prototype delivery" written in by lawyers who don't fully understand the technology. These proxy milestones often capture the wrong moment in the development cycle, either too early to mean anything or too late to course-correct.

Consider what a well-designed milestone structure actually needs to accomplish. It should surface early signals of commercial commitment. It should give the licensor meaningful leverage before the program is dead, not after. And it should track the licensee's actual behavior: hiring, spending, partnering, filing.

Here's a diagram of how most milestone structures fail versus where intervention actually needs to happen:

graph TD
    A[License Signed] --> B(Development Begins)
    B --> C{Internal Reprioritization}
    C -->|Program deprioritized| D[Milestone Missed]
    C -->|Program active| E(Next Milestone Hit)
    D --> F[Licensor Notified]
    F --> G[Cure Period]
    G --> H((License Terminated))
    E --> B

The gap between C and D is where everything falls apart. Internal reprioritization happens quietly. A milestone missed is the public announcement of a decision that was made months earlier, behind closed doors.

So what should licensors do?

First, build reporting obligations that are genuinely informative. Quarterly development reports are standard. Quarterly reports that require the licensee to document budget spend on the licensed technology, headcount assigned, and partnership discussions underway are much rarer. They're also far more useful. A licensee who goes two quarters without spending money on a program is telling you something.

Second, tie at least some milestones to input behaviors rather than output events. "Submission of a development plan within 90 days" is an input milestone. So is "completion of formulation studies" or "engagement of a contract manufacturing organization." These happen before regulatory touchpoints and give the licensor signal much earlier. A licensee who won't commit to a development plan within the first quarter of the license has already told you where this is headed.

Third, price the cure period carefully. Most agreements give licensees 60 to 90 days to cure a milestone failure. For a development milestone, that's often enough time to generate paperwork without generating progress. Consider whether the cure obligation should require a specific deliverable (a written catch-up plan with budget commitments, a signed CRO agreement) rather than simple notice of intent.

There's also a structural question worth asking before you draft the schedule at all: what does the licensee actually have to do to succeed commercially with this technology? That sequence of activities, not the regulatory pathway, should form the spine of the milestone schedule. Regulatory events can be waypoints. The real milestones should be tethered to development work.

One more thing worth saying plainly. Milestone structures are not a substitute for licensor engagement. The most effective technology transfer offices treat milestone schedules as conversation starters, not enforcement documents. Regular check-ins with the licensee's development team surface the reprioritization signals long before a missed deadline makes them official. That relationship doesn't happen automatically. Someone has to build it.

Writing better milestones matters. So does showing up to find out if anyone is actually following them.