1 in 3 corporate-startup partnerships misses its goal. Start with the shortlist.

TL;DR

About one in three corporate-startup partnerships in Europe misses its goal, according to Sopra Steria's Open Innovation Report 2025. INSEAD researchers found that understanding each other's needs matters most for success, and a 2024 study shows that corporates already struggle at the very start, with finding suitable startups. Part of the problem is where the search begins. Most databases are organized around startups and sector labels, while an innovation team usually starts from a technology it needs. A shortlist that starts from that capability also finds the patents, spin-offs and young companies a startup search would miss, and three checks before the first call keep the wrong ones off it.

Most large companies now work with startups

In Sopra Steria's Open Innovation Report 2025, written with INSEAD professors and based on a survey of more than 1,600 corporates and startups in 12 European countries, 72% of large organizations said they had collaborated with startups. For many of them, startups have become a normal place to buy technology from, next to their usual suppliers.

Corporate money has moved the same way. Global Corporate Venturing counted 3,068 corporations investing in startups in 2025, a record, and around one in five startup funding rounds now includes a corporate investor.

Two in three partnerships reach their goal

The same report found that 65% of organizations reached the objectives they set for their startup collaborations in 2025, up from 58% in 2023. That still leaves roughly one in three partnerships falling short, and the spread between companies is wide: those with a dedicated open innovation team reported a 73% success rate, against 49% for those without one.

Why corporate-startup partnerships fail

Most often, the two sides want different things and find out too late. In the 2023 edition of the report, INSEAD researchers named understanding each other's needs as the most important driver of success. The corporate is looking for a solution to a business problem that fits the way it buys, while the startup wants market access and a customer it can point to. Startups themselves call cultural differences the biggest barrier.

A 2024 study in the Journal of Open Innovation, based on a survey of corporates, adds a cause that sits even earlier. Difficulty identifying suitable startups was one of the main barriers the companies reported. When the wrong company is on the shortlist, a well-run pilot only shows that more clearly.

Start from the technology, then find who has it

When a pilot fails, the review usually looks at how it was run, and less often at how the startup got onto the list. The usual routes, a keyword search in a database, a meeting at a conference or a tip from a colleague, all start from the startups that are easiest to see.

Innovation teams know that the startup is a means to an end: what they are after is a technology that solves a problem in their business. The tools they search with are mostly organized the other way around, by company and by sector. The same capability may sit in a patent held by a university or another company, in a spin-off that has not raised any money yet, or in a young company from another sector that has never seen your industry as a market. A search built around startups and sector labels leaves those options off the list. One that starts from the capability brings them in, and the first conversation with whoever holds it begins with the problem you want solved. We wrote about where these companies leave their first traces, from patent offices to university spin-off lists, in 5 places to find early-stage startups before they raise.

Three checks before the first call

Once the capability is clear, three questions settle most of the shortlist before anyone books a meeting.

1. Does the technology solve your problem, today? Describe the problem in a full sentence, then look at what the technology does now and for whom. Two customers in your industry make for a very different first call than none.
2. Can whoever holds it deliver? For a company, that means it is still active, at a stage where it can take on a corporate customer, and run by a team that has done something like this before. For a patent, it means the rights are available to license and the holder is open to working with you.
3. Do the claims hold up? Customer logos, patents, funding rounds and the team's background all leave public traces, and checking them now costs far less than finding out halfway through a pilot.

Building the shortlist with Kuanta

In Kuanta you search our database by the technology itself, described the way you would explain it to a colleague. It holds validated startups, each checked as active and under ten years old, and each with its sector, stage, traction signals, competitors, latest news and verified contact details for the founder. When you also want to see patents, you can add them to the same search, and each patent that matches comes with an explanation of why it fits and a suggested way to work together. That takes you most of the way through the first two checks. You can see how it works on our scouting page.

For the final few, a deep evaluation takes care of the third check. It tests each startup's claims against filings and public records and scores it on the framework that investors in its industry use, so you go into the pilot knowing why you picked this one over the others. More on that on our evaluation page.

Sources

Sopra Steria, Open Innovation Report 2025, executive summary · Sopra Steria, press release on the Open Innovation Report 2025 · INSEAD Knowledge, The obstacles to partnering with start-ups · Banka et al., Understanding corporate concerns: barriers and challenges in corporate-start-up collaboration, Journal of Open Innovation, 2024 · Global Corporate Venturing, Corporate investment in startups at record high, 2026