120624-WP_Header Image.png

Featured image by Mara Potter/PitchBook News

Weekend Analysis

Where are all the consumer AI startups—and why aren’t VCs funding them?

VC funding for consumer AI startups is less than half that of enterprise.

When I was at the startup conference Slush last month in Helsinki, artificial intelligence dominated not only the panels but also my conversations with founders and investors.

While I expected this, one thing struck me in the 10-plus conversations I had with VCs over the two-day period: Almost all of them were focused on enterprise.

The words “consumer” or “B2C” rarely featured, and when they did it was mainly to remark that comparatively few investments in that market appear to be taking place.

Previous tech eras such as the rise of the internet or mobile targeted the consumer first and business later. Big consumer companies like Amazon and Uber led the initial wave of both of these technologies. And yet, in the age of AI, the reverse seems to be true.

Venture funding for B2B AI startups stands at $16.4 billion this year, according to PitchBook data. But the amount going to their B2C counterparts is less than half that, at only $7.8 billion. For generative AI startups specifically, deal value for B2C is only 6.4% of B2B’s current total for 2024.

Yet by one estimate, the consumer AI market by 2032 will be twice the size of the enterprise market for AI.

With the balance tipped so far in favor of enterprise, the question is: Is there a lack of consumer AI companies, or is VC just not funding them?

A vicious circle

“It’s the chicken or the egg, really,” Antler partner Tobias Bengtsdahl said. “I do think that VCs are sleeping on consumer right now, and that might be influencing where founders are building.”

This article appeared as part of The Weekend Pitch newsletter. Subscribe to the newsletter here.

From the investor side, the current VC climate has influenced the flow of capital to AI, according to Bengtsdahl. Since the onset of the VC downturn in 2022, investors have shifted away from consumer startups in general. Firms like Andreessen Horowitz have reportedly overhauled their consumer teams.

A challenging macroeconomic climate caused by rising inflation, higher interest rates and markdowns in valuations that occurred heavily on the consumer side made consumer a risky bet.

The shift to a profitability focus also made the consumer market less attractive. Enterprise clients are more likely to commit to multi-year contracts and subscription-based models, creating predictable and recurring revenue streams.

Even in the riskier world of AI, this downturn mindset persists. But, according to Bengtsdahl, one hangover from the VC bubble days of 2021 is also influencing VC investments.

“You can’t get away from the herd mentality, even now,” he said. “Once one person invests in something, everyone else follows. The simple fact is that consumer is out of fashion. Founders [in that space] have been treated very harshly since 2022, and that hinders others.”

A different world

A key factor holding back B2C AI, according to Point72 Ventures managing partner Sri Chandrasekar, is that the consumer landscape has significantly changed over the past two decades. User acquisition costs are higher and—compared, for example, to the internet era—there are incumbents in place that pose a real threat.

“A startup is just as capable of approaching a large enterprise customer and pitching their solution as any existing software vendor,” Chandrasekar said. “Consumer data, on the other hand, is often owned by incumbents such as Google or Meta. In that sense, I believe these companies already have an advantage over startups, which must first acquire that data and the insights it provides before they can deliver a novel experience.”

Data is the lifeblood of AI development, and building models requires vast amounts of information and is very expensive. Big tech companies have already committed vast resources to AI and used their capital to hire the best talent to develop their models.

Meta, for example, has announced plans to use personal data from its users to train the generative AI developed by its subdivision Meta AI. The company is focusing on developing chat functions and targeted ads using the technology. Google has its Gemini chatbot and offers consumer-facing products like AI image editing.

But there are areas where newcomers may have an edge, according to Bengtsdahl. An AI-powered social media platform may be a tough sell with consumers given how entrenched incumbents are in the market, but other spaces where big tech has less credibility, such as mental health solutions, offer opportunities for startups.

The bigger the risk, the bigger the reward

As things stand, the enterprise market seems like a safer option for investors looking to capitalize on the growing interest in AI. But while riskier, the consumer sector’s rewards have the potential to be much higher for VCs.

According to Zion Market Research, the market size for consumer AI is predicted to grow to around $1.3 trillion by 2032. For enterprise, it is estimated to reach only around $560 billion by the same year, according to Precedence research.

It is too soon to tell from VC exits, but if past performance is any indicator of the future, the right bet in the consumer space is also likely to pay out higher.

For the period between 1995 and 2022, the value generated by the top five consumer exits was 2.3x higher than that of enterprise companies, according to PitchBook data.

But for exits to happen, investments need to close. According to Chandrasekar, differentiation is key to consumer AI’s future.

“Many consumer AI startups today offer capabilities that are likely to be developed by existing big tech companies,” he said. “To attract excited investors, I believe consumer startups need to demonstrate that they are uniquely positioned to offer a brand-new capability or experience that can reach a critical mass of consumers and cannot be easily replicated by incumbents.”

Featured image by Mara Potter/PitchBook News

Learn about our editorial standards.