Research commercialisation is the process by which new or improved technologies, products, processes and services generated through research are brought to market. That is the UKRI definition, and nearly every university page on the subject is a paraphrase of it.
The definition is correct. It is also written, almost everywhere you find it, by the institution that owns the intellectual property, for the academics who work there. Those pages stop at the moment a route is chosen and a contract is signed. What follows that signature is years of capital, patent maintenance and operator work that decides whether the science ever reaches anyone. This article covers both halves.
What is research commercialisation?
Durham University uses a deliberately wide definition: any mechanism by which intellectual property, know-how or expertise is developed and delivered into markets or paid-for applications, including spin-out companies, licensing, the sale of IP, and the commercialisation of know-how itself. That breadth matters. Commercialisation is not only patents. Software, data, trade marks and undocumented expertise all count, which is why the UK government's knowledge asset guidance applies the same logic to any public sector body sitting on patents, copyright, software, trade marks, know-how or data.
Spelling first, because it splits the search results: commercialisation and commercialization are the same word. British and Australian institutions use the s, American ones use the z, and the process behind both is identical.
Commercialisation is not research translation
This is the confusion behind most searches for the term. Australia's NHMRC draws the line explicitly: commercialisation is one branch of research translation, specifically the branch where outcomes reach a market as new technologies, processes, products or services. Translation also happens through policy change, clinical guidelines and practice, none of which involve a customer.
The practical difference is who pays. Translation into policy is funded by the public purse. Commercialisation requires someone to buy something, which is a much harder test and the reason most research never takes this route.
The four routes out of the lab
The University of Bristol's route guidance lists the realistic options and states plainly that licensing is the most common one.
License to an existing company
The institution keeps ownership of the IP and grants a company defined rights to make, use or sell within agreed limits, in exchange for fees, royalties or equity. Edinburgh Innovations describes this as often the most effective route, because the university retains oversight while the licensee does the development. It suits technologies with discrete, identifiable applications in a market that already exists.
Spin out a new company
A new legal entity is formed, the IP is licensed or assigned into it, and it raises the capital to finish the job. Edinburgh's rule of thumb is that spin-outs fit technologies with a wide range of applications or a complex path to product. Bristol's decision criteria are sharper: prefer a spin-out when the invention is a platform technology with many possible products, when no existing industry makes anything similar, when the market is large enough to justify substantial development investment, and when at least one credible inventor will actually join the company.
Assign or sell the IP outright
You transfer all rights, usually for a fee or shares. Bristol is blunt about the cost of this: you lose control over what happens to the IP, you lose your own ability to keep developing it, and inventors have no real ongoing involvement. It is not generally used by public research institutions.
Contract and collaborative research
Universities UK separates contract from non-contract research: in the first, a commercial partner approaches the institution with a specific problem, funds the work under a commercial agreement, and often holds first rights to the result. Edinburgh treats collaborative research as a full third route and notes it frequently includes a first option for the company to commercialise the output. If you are considering this route, it is worth understanding what contract research and development is as a service before you negotiate the IP terms.
Non-STEM research is not excluded from any of this. UKRI's ESRC toolkit sets out spin-outs, licensing, social enterprise and consultancy as social science routes, and argues the yardstick there should be social value and a self-sustaining operation rather than financial return alone.
Choosing between routes is not a matter of ambition. Simon Fraser University's technology licensing office lists the factors that actually decide it: time to market, technology readiness, how much capital the route demands, who holds operational control, and how much of the inventor's time the venture will consume. Licensing is capital-efficient and fast because the licensee absorbs the development burden. A spin-out costs more time and money up front and pays in equity and control.
The stage gate, disclosure to deal
The University of York publishes one of the few detailed stage gate processes you can read from outside the institution. Compressed, it runs like this.
Research produces something novel. The academic files an invention disclosure form describing what it is, why it is new, how mature it is, who funded it, which third parties have contractual claims, and whether it has been disclosed publicly. The commercialisation team reviews it and responds within about two weeks. If it survives, an initial evaluation follows: prior art search, market sizing, confidential conversations with companies to test industry pull, and a first view on the likely route. Only then does the institution commit to a licence, a spin-out or a joint venture, and usually only after further technical work to reach commercial proof of concept, which licensees and investors call de-risking.
Four things stop a disclosure dead at review, and they are worth memorising because they are avoidable:
- the institution does not own the IP
- usage rights have already been given to a third party
- the invention has been publicly disclosed
- it is still at the idea stage
The University of Reading's guidance makes the same point from the other direction: protect and manage the IP appropriately before you start talking to customers and industry experts, and involve the commercialisation team at idea stage, not once a paper is in press.
How long it takes, and what the institution keeps
York's published approximations are the honest ones: the shortest time to develop a licensing opportunity is around six months, and around eighteen months for a spin-out. The reason it stretches is structural. University technology is usually at too early a stage to attract immediate investment from a licensee or an investor, so timescales run longer than they would in a commercial setting.
Durham publishes the funnel numbers that most institutions do not. Its research commercialisation team handles roughly fifty to sixty research disclosures a year and spins out three to four businesses a year, with more than sixty companies created since the 2000s. Those two figures sit next to each other on the same page, and the implication is the part nobody writes down: somewhere between one in twelve and one in twenty disclosures becomes a company. The rest are licensed, parked, or quietly closed.
On terms, Durham normally takes founding equity of around 15 to 25 percent depending on how IP-intensive the deal is, with lower equity where the contribution is academic know-how rather than patented IP. There are typically no IP royalties on top of that equity, and no recovery of patent costs accrued before the company was formed. Durham states its approach is substantially aligned with the USIT Guide, and that it has substantively adopted the best practices of the independent Spinout Review published in November 2023, whose recommendations the UK government accepted in full. Those specifics vary by institution and are the single most negotiated part of any deal, which is why the university spinout equity split deserves its own conversation before you sign.
What the university pages leave out
Every source above ends at the signature. What exists the morning after is a technology at low readiness, a patent family that now costs money in every jurisdiction every year, no operating team, and a gap between grant-funded proof of concept and first revenue that no research council covers. This is where most commercialised research actually dies, and it is why it is worth knowing what research to commercialization actually costs before you pick a route.
Someone has to be on the other side of that handoff. That counterparty is what a deep-tech venture builder is: a party that finances the technology, files and pays for the patents, and staffs the company with operators rather than waiting for the inventor to become a chief executive.
EX EPIC works from that side. The group reports a capital track record of over EUR 160 million financing, patenting and deploying deep tech across four continents, with more than 200 waste-to-energy units deployed across 11 countries, 24 patent families filed and around 100 more validated in the pipeline, and more than 250 operators trained and placed into portfolio companies through its academy, according to its own published figures. The portfolio and the reasoning behind the model are set out in more detail by founder Julien Uhlig.
The point is not the numbers. It is that each of them corresponds to something a licence agreement does not supply: capital, a filing strategy across jurisdictions, and people who have deployed hardware before. Public money can cover part of the gap, which is why EU funding for deep tech startups is usually stacked underneath private capital rather than replacing it.
Is your research a commercialisation candidate?
Assembled from the criteria above rather than invented, five questions decide it:
- Is there a defined unmet need, and can you name someone who would pay to have it met?
- Is the result novel and still undisclosed, or has publication already closed the patent route?
- Is it a platform with many possible products, or a discrete improvement to an existing one? The first favours a spin-out, the second a licence.
- Will a credible inventor commit real time to it? Every spin-out needs a champion; licensing does not.
- Is the market large enough to justify the development spend that stands between the current result and a product?
If the answers point away from commercialisation, that is a legitimate outcome rather than a failure. Universities UK states it directly: not all research is done with the intention of being commercialised, and commercialised research is in no way more or less valuable than research that is not. For work that does pass the test, the sequencing of disclosure, protection, funding and route selection is the whole game, and it is covered separately in how to commercialize university research.
FAQ
Is research commercialisation the same as selling my research?
No. Selling means assignment, one of the four routes, and it transfers all rights and all future control away from you permanently. Most public research institutions avoid it and license instead. Durham also states plainly that raising revenue through licensing fees, royalties or equity sales is a by-product of commercialisation rather than its primary objective.
Can I commercialise research I have already published?
Often not by the patent route. Public disclosure before filing is one of the reasons York lists for an invention disclosure not progressing, alongside the institution not owning the IP. Routes that do not depend on novelty can still work: consultancy, software licensing, trade marks and know-how survive publication.
Do I have to leave academia to commercialise my research?
No. On the licensing route academic founders typically stay in academia while the technology still reaches the market through the licensee. The spin-out route is the one that needs someone to run a company, whether that is the inventor or a hired chief executive.
Can social science, arts and humanities research be commercialised?
Yes. UKRI's ESRC toolkit sets out spin-outs, licensing, social enterprise and consultancy as routes for social science, and argues these should be judged on social value and whether the operation sustains itself, not purely on financial return.
Who owns the IP, me or my institution?
It depends on the institutional policy, and it is not uniform. Most universities own IP arising from research they fund. Some run creator-owned policies, such as Simon Fraser University's Policy 30-03, which leaves ownership with the creator except where a separate agreement says otherwise. Read your institution's policy before you disclose anything to anyone.