Enterprise companies and generative AI: Just looking?

W
elcome to the TechCrunch Exchange, a weekly startups-and-markets newsletter. It’s inspired by the daily TechCrunch+ column where it gets its name. Want it in your inbox every Saturday? Sign up here.

This week, I am diving deeper into what generative AI means, or doesn’t mean, for enterprise buyers. I also have some notes on why your company may want to be like Figma, and how the investing side of the market is adjusting to down rounds being the new normal. — Anna

Not-yet-unlocked potential

When The Exchange looked into Battery Ventures’ state of cloud software spending report, we started by focusing on what the title promised: fresh data on cloud software spend. And it turned out to be more encouraging than we expected.

We then looked into another piece of good news for founders: that startups building tech to automate tasks and drive quick productivity gains might be able to duck the downturn. It was based on a report data point showing that automation had risen among enterprise budget priorities.

But in the back of my mind, I also kept thinking about some of the report’s comments about generative AI — and not just because superlative takes on the topic have been ubiquitous ever since.

If anything, you could call Battery’s view on generative AI conservative, but that would be unfair. After all, the VC firm was just relaying findings from its Q1 survey, which gathered responses from 100 C-suite execs (CXOs) managing around $30 billion in spend.

“Generative AI is one of the buzziest topics in the tech world, as technologies like ChatGPT hold tremendous, transformative potential. From the perspective of enterprise tech buyers, however, it may be just that — potential,” Battery said in a blog post accompanying the report.

This interpretation was based on survey results showing that only 32% of companies were looking at exploring generative AI applications. And even when doing so, it was “primarily for internal productivity, workflows, knowledge transfer and cost reductions — not the glittery, futuristic applications one might have predicted,” Battery said.

For clarity, this isn’t about dabbling with ChatGPT or Midjourney for fun; this is about trialing these approaches in an enterprise environment, with an actual budget attached.

Companies with deeper pockets seemed more open to exploring generative AI applications than their smaller peers: 48% of CXOs with a budget over $250 million were already experimenting in this field in Q1, compared to 25% of those with less budget. But there were still big questions left to answer.

“Enterprises are still exploring how to interact with foundation models and whether the right approach is bringing a model to their proprietary data (Model to Data) or their data to an external model (Data to Model),” Battery reported.

The foundation model layer is where you’d find OpenAI, but also Stability AI, Anthropic and others.

While some foundation models are open source, it seems increasingly likely that the biggest, best-funded players will win. But when it comes to supporting enterprise companies in leveraging this technology, there’s still a lot of white space and adjacent opportunities for startups to pursue.

Startups as enablers

That enterprise buyers don’t have their AI roadmap planned out could actually be an opportunity for startups.

According to Battery, “Many organizations are still working to operationalize their data, which is leading to accelerated adoption of technologies like data labeling, embedding models, vector databases and search, data quality, observability and governance.”

TechCrunch’s enterprise desk has been covering these types of startups and technologies for years, so this isn’t new. But hearing that these adjacent verticals might enjoy new tailwinds as a result of the rise of generative AI — and hearing it from investors — is a relevant piece of information.

And while the generative AI space itself is starting to look increasingly busy, there are still gaps to address. Per Battery’s survey, “48% of CXOs who are exploring generative Al today noted that governance was the top tool that is missing within the current toolchain.”

Fear and opportunity

Governance concerns should be taken seriously, but that doesn’t change the increasingly common prediction that generative AI is here to stay. In its State of the Cloud 2023 report, Bessemer Venture Partners included the following address to entrepreneurs:

[T]o all the SaaS founders, we know your legal team is already nervous.

These breakthrough models are as astonishing and powerful as they are nascent. Technology is accelerating at rates we have never seen before, and businesses and society-at-large will be playing catch-up for some time. There are real legal and ethical concerns around IP, biases, accuracy (i.e., hallucinations), and regulation.

Over the coming months, fear will build as we enter the trough of disillusionment. Companies will continue to ban employees from using AI at work, regulators will raise more questions, and AI-powered apps will make real mistakes.

It is important for companies to work closely with their legal teams and take appropriate measures to mitigate these risks when implementing AI-powered tools.

However, if you’re a CEO of a cloud company, opting for no AI strategy is equivalent to signing the business’s death warrant.

I wouldn’t go as far as foreboding doom or death, but it is also becoming clear that late adopters might get leapfrogged by competitors who can afford to experiment with generative AI, not tomorrow, not next week, but right now.


From Figma to FABRIC

Are there characteristics that every successful enterprise software company shares? Seasoned investor Kobie Fuller thinks so. The Upfront Ventures partner said he came up with an acronym to sum these up: FABRIC, which stands for being fast, addictive, bold, rewarding, integrated and built on community. And if you’re looking for a role model, “Figma embodies all these components of FABRIC,” Fuller told TechCrunch.


I’m the bad guy, duh

Down rounds require someone to play the part of the “bad guy” — setting a new, lower valuation that existing investors will also accept — and Tribeca Venture Partners is ready to step into the role, PitchBook News reported. The New York–based VC firm is said to be raising a $200 million growth-stage fund that would write small checks to price down rounds. If this happens, it will be another confirmation that down rounds are becoming business as usual.


Cast your ballot

The TechCrunch Podcast Network has been nominated for two Webbys in the Best Technology Podcast category. You can help TechCrunch win by voting for Chain Reaction, which digs into the wild world of crypto. Or Found, which brings you the stories behind the startups by sitting down with the founders themselves. Please take a few moments to vote for your favorite at vote.webbyawards.com. You’ll find them in the podcast category under technology. Voting closes Thursday.