For years, Partners of the Bessemer enterprise has put a lot of effort into promoting its brand around all the things cloud-related. The company has Nasdaq-traded Cloud Index and a large portfolio of well-known cloud startups.
So when Bessemer proclaims its newest vintage Cloud condition report “praise for legacy cloud”, deserves attention.
In its latest report, released Thursday, the company argues that startups are moving away from what it describes as legacy SaaS business models and adopting a latest paradigm: vertical artificial intelligence. While this is not a death knell for the enterprise software corporations that have been its bread and butter for years, it does mean that they are going to have to evolve over time.
“It’s a big deal, especially from Bessemer. I think it reflects how big a movement and paradigm shift this is,” said the company’s partner Sameer Dholakia, referring to the rise of large language models and AI-enabled tools that are changing the enterprise software landscape. “This event spells extinction for companies that are not early adopters.”
All about vertical AI
If cloud has been a buzzword for years, this time Bessemer is pushing to popularize the concept of vertical AI as the next big thing. He defines it as applications “targeting high-cost, repetitive, language-based tasks that dominate many industries and large economic sectors.”
Bessemer, which cites greater than a dozen of its own portfolio corporations as vertical examples of artificial intelligence, is not the only company highlighting this trend. For example, at the end of last 12 months Cowboy ventures 1 published by A Vertical AI market mapAND Gray castle listed outstanding startups in the healthcare, finance and skilled services space.
According to Dholakia, some of the most promising vertical AI startups are already well beyond the seed stage. Many of them have secured seed funding well in advance of their late 2022 launch ChatGPT established the power of great language models for the masses.
While we’re likely at least a few years away from the emergence of large-scale vertical AI IPOs, Dholakia is already seeing early entrants disrupting dominant business models in the enterprise software space. He added that one of the changes enabled by AI is the tendency to charge customers based on work moderately than the number of users using the software.
“We see that people value the delivery of their capabilities differently,” he said. “They solve the task and charge for the work product.”
Dholakia cites a legal technology AI portfolio company Compensate for example, a company may currently charge personal injury attorneys for demand letters created as part of its offering.
Exits and valuations
While Bessemer is excited about the promise of a vertical AI offering, the company is also aware of the reality that startups in the space are largely the way out. Meanwhile, there are a huge number of still private, later-stage enterprise software corporations that were founded before the pure AI-focused era.
In particular, the backlog of not-yet-exited unicorns is accumulating during a period of slow IPOs and large M&A deals involving venture-backed private software corporations. And given the traditionally lazy summer IPO season, it appears the slow streak will proceed.
Dholakia, for his part, expressed hope that the pace of offers will pick up, possibly by next 12 months. He said that for the past few years, software startups have been holding off on going public as a result of market contraction after 2021, subsequent belt-tightening for several quarters and pressure to update artificial intelligence capabilities.
As a result, there is now a “backlog of once-in-a-generation, maybe once-in-a-lifetime” corporations that need to go public but are still private, he said.
Hopefully we’ll see them on their technique to market soon.