Comment from Matt Asay, MongoDB
February saw the announcement of yet another corporate venture fund, prompting CNBC reporter Jordan Novet to quip, “If your company has no corporate-venture arm, what is it even doing?”
Novet’s comment came on the heels of MongoDB’s new fund [disclosure: I work for MongoDB], but the intentional irony strikes home because it comes on the heels of recent announcements from Databricks (December 2021) and Snowflake (November 2020), joining larger tech companies that launched their own venture funds years ago, including Salesforce (2017), Microsoft (2016), Google (2009), Red Hat (2000), and Intel (1991).
Why? What is the point of these different tech venture funds? Are they merely a sign of well-funded partnerships?
Partnerships masquerading as investments
Sometimes that is surely the case. Though Salesforce launched its first $50m Impact Fund with high-minded verbiage about benefiting society and social good, it also highlighted the goal to “strategically invest in companies built on the Salesforce Platform, delivering unique capabilities to Salesforce customers.” Could this be done through partnership? Of course. But investing arguably makes “some partners more equal than others,” to paraphrase George Orwell’s Animal Farm. Nor did it hurt that sometimes these investments have paid rich financial returns for Salesforce, netting it $2.17bn in 2020.
In other words, there are likely all sorts of motivations in play for each of the corporate venture firms.
Take Red Hat, for example. The company was famously parsimonious under former CEO Matthew Szulik. (I once flew to Raleigh for a job interview with Szulik, and his first question was how much my flight cost.) Red Hat’s venture fund might have seemed an extravagant way to spend the company’s (then) small profits. Speaking at the time, Szulik stressed that Red Hat Ventures was intended to be “much more than a source of financing” and instead saw a need for “true partnerships” whereby “we accelerate their ability to reach the global Linux market and they enhance our ability to offer the broadest range of Internet solutions to enterprise customers.”
Think of these partnerships as a way to buy some loyalty from partners the investor needs to round out its product suite, even as the investor hedges its bet by accelerating the startups’ access to large markets. In Databricks’ fund announcement, it was quite clear on this: The “Lakehouse Fund will focus on early- and growth-stage companies extending the lakehouse ecosystem or using the lakehouse architecture to create the next generation of data and AI-powered companies.”
Put this way, a few hundred million here and there is a small price to pay for influencing markets to lean your way. Or, as MongoDB’s Serge Tanjga put it, “We are not just building a product; we are building an ecosystem.”
In a recent Grant Thornton survey of 150 tech CFOs, respondents were asked how their companies planned to fund growth during the next two years. Acquisitions notched the top spot (58%), followed by divestitures (25%) and nontraditional mergers and acquisitions such as special-purpose acquisition companies, alliances, and joint ventures (15%). When asked how they expected to reach new customers, direct selling (58%) came in first, followed by customer acquisition through third parties (25%). Corporate venture funding, in other words, is not the preferred way that tech companies seek growth. As one very unscientific source (Hacker News) suggests, marketing is arguably the top thing companies can and should do to find new customers.
Yet, it is telling that companies as varied as Databricks, Salesforce, and Intel continue to pour copious quantities of cash into supercharged partnerships-qua-venture investments to attempt to move markets in their favour. Yes, companies will still do marketing, just as they will open source software or provide open APIs to increase developer interest, and (or) build out partner programs to align corporate interests. Venture funds are just one more way that companies hope to “build an ecosystem, not merely a product,” to paraphrase Tanjga.
Do venture funds work? Those that have done them longest, such as Intel, seem to believe so. Intel Capital has invested billions of dollars in hundreds, maybe thousands of companies since 1991 and can point to both significant venture returns and support for Intel processors among its investment portfolio.
Money may be the ultimate takeaway. Intel, like Salesforce and other successful corporate venture initiatives, insists on financial returns, not just ecosystem development, as central to its investment thesis. Market-making is a difficult, long-term endeavour, and venture investment is just one contributing factor. The companies that do it best do not seem to privilege strategic considerations over financial returns. If your company is considering a venture arm, perhaps you should take this approach, too.