Insiders and VCs expect M&A wave from machine learning

  • As corporate valuations plummet during the market downturn, insiders and VCs predict a wave of mergers and acquisitions.
  • Machine-learning startups in the “$10 million ARR club” are the prime targets, they say.
  • Data giant Snowflake says it is monitoring strategic acquisitions, further pointing to a shakeout.

Company valuations are being massively lowered as the market sours, with many raises and layoffs of employees. Insiders and investors are now anticipating a major wave of acquisitions, with many machine learning startups being the main target.

Some of the startups most likely to be picked up are part of what investors and insiders sometimes refer to as the “token $10 million ARR club,” which refers to companies that have picked up a few big first-time customers, but haven’t yet moved in. the mainstream have broken through. With an impending downturn, their customers could try to cut costs quickly – to the detriment of the startups who rely on their deals.

As with most emerging technology, new machine learning startups typically find a sweet spot in the rapidly evolving field and build a product like a wedge with customers. Those who are successful can then launch additional products and gradually expand until they own a large portion of their customers’ machine learning workflows. Those who cannot build a sustainable business are taken over or eventually close.

With the current market conditions and lower price tags to startups, this means there will likely be plenty of opportunities for acquisitions, either as an “acqui-hire” or to buy up key technology. Snowflake, in particular, will likely be eyeing acquisitions after spending $800 million on a machine learning platform called Streamlit.

“I think over the next six months, if things stay where they are, there could be some interesting M&A opportunities. Not necessarily major M&A, but I do think there will be some valuation resets for some of the private companies out there that could create interesting opportunities,” Snowflake chief financial officer Mike Scarpelli said during the company’s most recent earnings call.

Scarpelli went on to explain that there were some areas on the company’s roadmap where it may make sense to consider acquisitions for both additional staffing and technical purchases.

“We’re not looking for revenue, but good teams and technology at a more reasonable valuation,” he said.

The same tools that lead to billions of dollars worth of valuations lose their luster

The data catalog space, which includes startups such as the $1.2 billion company Alation and the $5.25 billion company Collibra, is one of many areas in the machine learning industry that sources say can be challenging to navigate. prove it as a convincing standalone product, that makes it ripe for acquisitions. Another common part of that workflow is feature stores.

Function stores allow developers to avoid unnecessarily massive recalculations when implementing a product’s machine learning component. The biggest player is Tecton, which operates the open-source feature-store tool Feast. Tecton was founded in 2019 by the makers of Uber’s Michelangelo machine learning tools.

Tecton has since moved beyond feature stores to other products, and like many open-source tools, Feast serves as a stepping stone to a more advanced (and more lucrative) tool. But insiders question whether a feature store — which at the time was enough to bring Tecton in $60 million in funding from investors like Sequoia and Andreessen Horowitz — can be a standalone product. Both Tecton and Rasgo, another startup that launched on the momentum of feature stores, have since broken new ground.

“That terminology was a little tricky for us. It’s very easy to hear the word ‘store’ and think of a database table,” Tecton CEO Michael Del Balso told Insider. “What we’ve seen, and we’re seeing this time and again, teams taking machine learning into production. They’re underestimating this problem.”

It is in many ways a return to the age-old question of whether something is a feature or a product. The machine-learning startup Dataiku, for example, has a feature-store component, while Tecton has quickly sought to grow beyond feature stores. Both are backed by Snowflake after Tecton raised $100 million in a round earlier this month that also included Databricks.

Snowflake and Databricks could be developing the same features as these multibillion-dollar startups

While Snowflake and Databricks are both betting on Tecton and others, there is a shadow over whether they will launch their own products. Insiders say that as long as they serve as a way to boost the use of Snowflake and Databricks, they expect the companies to remain supported. But Snowflake and Databricks may also see certain parts of the workflow, such as a feature store, as an element that they can add to their own products.

Not all machine learning startups are in this position. Many investors and insiders have said that there are several startups that have probably built up enough momentum to avoid being part of a rollup. Hugging Face, recently valued at $2 billion, is one that is common due to its large community, in addition to the startup Weights & Biases following experimentation.

At the same time, acquisitions that occur can yield lucrative results for investors, who rarely see the lightning-in-a-bottle results of a WhatsApp or Red Hat. Smaller acquisitions and IPOs largely deliver the expected results with a sufficiently high volume.

“Machine learning is really exciting, but sometimes it’s hard for some of these companies to match the ROI to,” said an insider close to Tecton and other startups. “Who knows – with this market environment, consolidation could happen very quickly.”

Leave a Comment