Growth Strategy · Technology Transfer
The Assets Were There. The Framework Wasn't.
A company can spend decades building genuinely proprietary technical capability and never move a fraction of its value outside the division that built it. The constraint is almost never the technology. It's the absence of a repeatable way to transfer it.
Why Most Companies Never Build This
Technology transfer — moving a capability developed inside one organization into use somewhere else — is a genuine, well-established field, not a niche practice. The Bayh-Dole Act of 1980 is the clearest evidence of how seriously this is taken at a policy level: it gave U.S. universities the right to own and license patents on federally funded research, specifically to create the incentive structure for moving lab discoveries into commercial use. Over the twenty years from 1996 to 2015, academic patent licensing enabled by that single piece of legislation is estimated to have contributed $591 billion to U.S. GDP and supported 4.3 million jobs.
The global IP licensing market itself has grown from roughly $27 billion in 1990 to approximately $340 billion in 2024, a growth rate that has outpaced global GDP for three consecutive decades. The demand side of this is not in question. What's inconsistent is the supply side — most organizations that could transfer internal capability outward simply never build the operating mechanism to do it.
IBM's licensing program didn't succeed because the patents were valuable. Plenty of companies hold valuable patents and license almost none of them. It succeeded because IBM built a repeatable operating process around moving them out.
A Real, Well-Documented Case
IBM is the clearest public example of what a standing capability-transfer program actually looks like in practice.
Under Marshall Phelps, who built IBM's intellectual property and licensing organization starting in the early 1990s, the company turned patents that had been sitting unused inside its own labs into a recurring revenue line — reported at well over a billion dollars a year through the 1990s and 2000s, at a time when most companies sitting on comparably large patent portfolios were licensing almost nothing.
The gap wasn't IBM's technology; competitors had access to comparable internal capability. The gap was that IBM built a repeatable operating function for it, and almost no one else did.
A Real, Well-Documented Case
IBM has run an active patent licensing program since 1985 — not occasional deals, a standing operating capability. For most years since 1996, that program generated over $1 billion annually in direct licensing revenue, peaking around $1.7 billion in 2000. The company holds roughly 100,000 patents, and for nearly three decades was the top annual recipient of U.S. patents of any company.
The program has also genuinely declined — licensing income fell to $626 million in 2020, and to $397 million and $366 million in 2022 and 2023 respectively, as legal and market conditions shifted. That decline is worth including precisely because it makes the case more credible, not less: even a program built at this scale, with this much real revenue behind it, required continuous operating discipline to sustain — it wasn't a one-time unlock that ran on autopilot for forty years.
What the Mechanism Actually Requires
The common failure isn't a lack of valuable internal capability. It's that no one inside the organization owns the question of whether a capability built for internal use could become an external one — and even when someone does ask that question, there's rarely a repeatable process for answering it, only a one-off deal negotiated from scratch each time.
IBM's program works as a standing capability because it separates three distinct functions that most one-off technology deals collapse into a single ad hoc negotiation: identifying which assets have external value, valuing and packaging them for a specific licensee, and enforcing and renewing the resulting agreements over time. Treating those as one undifferentiated task is a large part of why most internal capability transfer efforts stay one-off instead of becoming repeatable.
01
Identify
Determine which internal assets have external value — the question almost no organization assigns clear ownership to.
02
Value & Package
Package the asset for a specific licensee, with terms and framing suited to that buyer rather than a generic offer.
03
Enforce
Monitor compliance with the license terms and step in when a licensee violates them — the ongoing operating discipline a one-off deal never has to build.
04
Renew
Revisit and renegotiate terms as the relationship and the asset's value evolve, rather than letting the agreement lapse or run on autopilot.
Why This Takes Discipline, Not Just a Decision
Deciding to license an asset once is a single decision, made by whoever happens to notice the opportunity. Building the capability to do it repeatedly is an organizational commitment — someone has to own identifying candidate assets on an ongoing basis, not just react when one surfaces. Most organizations that could build this never do, not because the economics don't work, but because no one is ever explicitly given the job.
The Step IBM's Numbers Don't Show
IBM's revenue history, the Bayh-Dole framework, and the growth of the IP licensing market are all public, independently verifiable facts — none of it is proprietary.
What none of it explains is how an organization that has never licensed anything out before actually builds the muscle to do it repeatedly, rather than as a single negotiated deal. Strategic Innovation — Vaxa's proprietary methodology, built and referenced across engagements including Nokia, Intel, DuPont, and P&G — supplies that staged handoff: a first engagement Vaxa leads directly, a second where ownership shifts toward the internal team, a third the client team runs alone.
The sequence itself, not the decision to license something, is what doesn't show up in the public technology-transfer literature.
The Actual Question
When your last adjacency decision named its capability, was it tested against where that capability had actually been proven — or did the category label do the reassuring instead?
This relates to Vaxa's Growth Strategy.
Related Material
IBM Annual Reports, 2022–2023 — direct IP licensing income figures.
Bloomberg, "IBM's Patent Income Slips as Companies Resist 'Godfather' Deals," March 2021 — historical IBM licensing revenue, 1996–2020.
Thompson Patent Law, "How Licensing Intellectual Property Turns Patents Into Revenue" — global IP licensing market growth ($27B 1990 → $340B 2024) and Bayh-Dole Act economic impact data ($591B GDP contribution, 1996–2015).