Skip to content
TrustList
Blog

Corporations joined 39 of the 102 startup rounds we covered but led only two

Editorial

By TrustList Editorial

We sorted every investor in 102 October rounds: companies or their venture arms took part in 39 but led two. Suppliers, customers and platforms buy a window, demand and an option. Here is what founders give up.

About Corporations joined 39 of the 102 startup rounds we covered but led only two

Corporations joined 39 of the 102 startup rounds we covered but led only two

Between 2 and 14 October 2026 we published 103 funding-round reports on TrustList, one of them by mistake a second time, which leaves 102 distinct rounds. We went back through every one and sorted each named investor by what it is: an independent venture firm, a company or its venture arm, an asset owner such as a pension or sovereign fund, a public or development body, or an angel. The result was not what the usual picture of startup finance suggests.

A company that is not itself an investor took part in 39 of the 102 rounds. That is more than one round in three. Yet a corporate investor led only two of them: Dell Technologies Capital led the Series A of Ascerta, the Bellevue company formerly called Pay-i, and Best Practice Software, an Australian clinical software maker, led the round in Medcast. Independent venture firms led 67 rounds, and 19 reports named no lead at all.

So corporations are everywhere in these rounds and in charge of almost none of them. This piece looks at why large companies take small stakes in young ones, what they get, what the startup gives up, and how tax and accounting rules shape the size of those stakes.

How we counted. We read the 103 funding-news items published on TrustList from 2 to 14 October 2026 and listed every investor each item names as taking part in that round, leaving out investors mentioned only for earlier rounds. Each was classed as independent venture, corporate (a company that is not primarily an investor, or its venture arm), asset owner, public or development, angel, or other. Fleuret AI was published twice and is counted once. The items are themselves sourced to company or investor announcements; we did not add investors the announcements leave out.

Who the corporate investors were

The names that came up most often were the venture arms of companies selling the infrastructure that AI startups run on. NVIDIA took part in three rounds, Microsoft's fund M12 in three, and Alphabet's GV in three. Salesforce Ventures and Dell Technologies Capital appeared twice each. Samsung, Qualcomm Ventures and CapitalG, Alphabet's growth fund, appeared once or twice.

Behind that cluster sits a second group that tells a more interesting story: companies investing in startups that sell to them, or to their own customers.

The clearest case is mortgage software. The Series B of Vesta, which builds a loan origination system, drew money from Pennymac, New American Funding, Citi Ventures, nbkc bank, First American and FirstKey Mortgage. Several of those are the very lenders a loan origination system is sold to. Car companies did the same for their dealers: Flai, which sells AI agents that answer dealers' calls and messages, counted Toyota Ventures, The Friedkin Group and Findlay Automotive Group among its investors.

A bank backed the software it helped design: Promenaut, a London company building governed AI agents, launched with HSBC as both design partner and strategic investor. A hospital group backed hospital software, as MemorialCare's innovation fund took part in Vitalize, which automates hospital staffing. Even software vendors backed a tool for their own users: the observability vendors Datadog and Grafana Labs both invested in OllyGarden, which cleans up OpenTelemetry data before it reaches tools like theirs.

In Japan the pattern runs through banks and insurers. The contact-centre AI company Flyle raised from Aflac Ventures Japan, Japan Post Capital, Mitsubishi UFJ Capital, Mizuho Capital and SMBC Venture Capital as well as Salesforce Ventures. Two more Japanese rounds, AIONA and Henry, drew on bank-owned venture arms and regional banks.

The compute suppliers follow the same logic from the other side. NVIDIA's three rounds were GMI Cloud, a GPU cloud, Mecka AI, which supplies training data for robots, and Nous Research, which builds open models. Each is a heavy buyer of the chips NVIDIA sells. The server maker Supermicro invested in Verda, the Helsinki AI cloud formerly called DataCrunch, which runs on exactly that kind of hardware.

Read together, the 39 rounds sort into three motives. A supplier invests in a customer that will buy more of its product. A customer invests in a supplier it wants to survive and to shape. A platform invests in companies that make its ecosystem more useful. Pure financial investing by a corporation is the exception, not the rule.

What the parent company gets

The corporate venture arms we read describe their work mostly in terms of what they give founders, but the value to the parent is visible between the lines.

Intel Capital, which calls itself the first corporate venture capital firm, founded in 1991, says on its own site that it made "nearly 1,000 highly curated Global 2000 customer introductions" for portfolio companies in 2025 and placed 250 Intel engineers and consultants with them. Introductions to Intel's customers and the loan of Intel's engineers bind a startup's product to Intel's platform. In January 2025 Intel said it would spin the arm out as a standalone fund with Intel remaining an anchor investor, a sign that even the oldest corporate investor sees tension between strategic and financial goals.

Salesforce Ventures, founded in 2009, says enterprise technology "has been our sole focus" and that it has invested more than 6 billion dollars in more than 630 startups, guiding over 200 to an IPO or acquisition. It describes a "deep partnership with Salesforce" through which founders can use the parent's platforms "to generate visibility", so each investment also widens the market for Salesforce itself. Microsoft's M12 is just as direct: "We harness the power of Microsoft to help today's most innovative startups become tomorrow's technology leaders." BMW i Ventures, investing since 2011, closed a 300 million dollar third fund this year to back AI "across the automotive ecosystem", the supply chain BMW buys from.

Research gives the strategic case some weight. Chemmanur, Loutskina and Tian, in the Review of Financial Studies in 2014, found that startups backed by corporate venture capital "are more innovative, as measured by their patenting outcome, although they are younger, riskier, and less profitable" than those backed by independent firms. They point to two causes: the parent's industry knowledge, because of the "technological fit" between parent and startup, and a "greater tolerance for failure". A parent that understands the technology and can wait for it is well placed to learn from it.

What a corporation buys with a minority stake, then, is a mix of four things:

  1. A window on technology it would otherwise only see when a competitor bought it.
  2. Demand for its own products, whether chips, cloud capacity or a software platform.
  3. An option on the company, since an investor that knows a startup well is in the best place to buy it later. When Salesforce agreed on 5 September 2024 to buy Own Company, the data backup firm, its release described Own as "a Salesforce Ventures portfolio company"; the deal closed on 18 November 2024.
  4. A financial return, which the large arms do track, though few say it is their main aim.

Accelerators, the cheaper way in

A stake is not the only route. Many large companies run accelerators that take no equity at all, because what they want is the relationship.

Google for Startups Accelerator takes cohorts of 10 to 15 seed to Series A companies, preferably working in machine learning and AI, and offers "equity-free support" with mentoring from Google teams, cloud credits and early access to Google's AI products. Microsoft for Startups offers up to 150,000 dollars in credits and access to "the Microsoft network of customers". Neither page says what the company gets in return, and they do not need to: a startup that builds on a cloud during its first two years tends to stay on it.

Banks run the same model with a stake attached. Morgan Stanley's in-house accelerator, Morgan Stanley Inclusive & Sustainable Ventures, gives each startup in its cohort an equity investment, a five-month programme, mentoring and office space, and connects founders with the bank's bankers, financial advisers and technology staff. The bank says the programme has put more than 40 million dollars into more than 160 startups and organisations since it began. Its pages frame the aim as inclusion and impact rather than return, but the founders it backs also become clients, partners and sources of deals for an investment bank.

MassChallenge, an accelerator that partners with corporations, set out the corporate case in 2020 as seven benefits: keeping pace with technology, insight into emerging markets, gains from early investment, a voice in young companies, diversification, a vetted pool of investments and talent, and job creation. It is a promotional piece written to win corporate partners, but it names the motives plainly. The second and sixth, insight and a vetted pipeline, match what the 39 rounds above show.

What the startup gives up

Corporate money is rarely bad money, but it is never only money, and the two leads in our sample show why founders hesitate to let a corporation set the terms.

The best-known problem is disclosure. Dushnitsky and Shaver, writing in the Strategic Management Journal in 2009, studied 1,646 startup-stage ventures and found that where intellectual property protection is weak, a startup and a corporate investor are less likely to do a deal when the invention targets the corporation's own industry. The founder fears that explaining the product in due diligence is the first step to having it copied. Where patents and trade secrets protect the startup well, the overlap makes a deal more likely, because the corporation brings real expertise and the risk of imitation is lower.

Other costs are softer but real. A strategic investor's presence can make a rival hesitate to buy from the startup, or to buy the startup. Information rights that let the parent see detailed figures can carry the same risk as due diligence, repeated every quarter. And a corporation's priorities change with its own strategy; Intel's decision to spin out its arm is a reminder that a parent can step back.

That may explain the pattern in our count. In 37 of the 39 rounds with a corporate investor, an independent venture firm, an asset owner or a public fund led and set the price, and the corporation took a seat alongside. The founder gets the relationship without handing the terms to a party whose interests may later diverge.

Questions worth putting to a corporate investor before signing:

  • Does the parent get any right of first refusal or notice if the company is offered for sale, and for how long?
  • Which people at the parent will see board papers and financial reports, and are product teams walled off from them?
  • Is the investment made from the parent's balance sheet or from a fund with outside investors and a fixed life?
  • What happens to the commercial relationship if the venture arm is closed or spun out?

How tax and accounting shape the stake

Large companies have no tax relief aimed at venture investing in the UK today. The Corporate Venturing Scheme, introduced in 2000 for ten years to encourage it, expired on 31 March 2010, and HMRC's manual says it "has not been renewed". What remains is the general regime for company shareholdings.

The most important piece of it is the Substantial Shareholding Exemption. Under Schedule 7AC of the Taxation of Chargeable Gains Act 1992, a gain a company makes on selling shares is not a chargeable gain if it has held a substantial shareholding, broadly at least 10 per cent of the ordinary share capital, for a continuous twelve months in the six years before the sale, and the company invested in is a trading company or the holding company of a trading group. Since April 2017 the investing company itself no longer has to be trading. For a corporate investor that holds 10 per cent or more of a startup for a year, a profitable exit can be free of corporation tax. Below 10 per cent the gain is taxed like any other.

Accounting pulls in the other direction. Under the international standard IAS 28, a holding of 20 per cent or more of the voting power is presumed to give "significant influence", and the investor then accounts for the startup by the equity method: the investment starts at cost and its carrying value moves with the investor's share of the startup's profit or loss. Since most startups make losses, a large stake brings those losses onto the parent's income statement. A stake below 20 per cent is usually carried at fair value instead.

Those two thresholds, 10 per cent for the tax exemption and 20 per cent for equity accounting, frame the decision whenever a corporation takes more than a token position. None of the round reports we counted states the size of a corporate investor's stake, and companies rarely publish it, but a participant alongside an independent lead usually holds a few per cent at most. A corporation that wants a stake large enough to qualify for the exemption but small enough to keep the startup's losses off its income statement has a fairly narrow band to aim for, and a founder who knows that band can predict how far a strategic investor will want to go.

Corporate investors in our catalogue

Our venture capital firm category lists the independent firms that led most of the rounds above, and every round mentioned here has its own report on our funding news page, where the investors are named as the companies state them. For founders weighing an offer from a strategic investor, the useful comparison is not between corporate and independent money but between a round where a corporation joins an independent lead and one where it sets the terms alone. In the fortnight we counted, founders chose the first almost every time.

Related on TrustList:

Sources

Categories & features

TrustList Weekly

The week in software and IT, in one email

The news that matters to buyers, new rankings and our own research. Every Thursday, free, and easy to leave.

We will email you to confirm. Unsubscribe with one click in any issue. Privacy policy

Keep exploring

More on TrustList

Everything here links back to the same verified catalogue. Pick your next stop.