The SaaS-pocalypse can wait, Salesforce still has customers where it wants them

https://www.theregister.com/headlines.atom Hits: 46
Summary

SaaS The SaaS-pocalypse can wait, Salesforce still has customers where it wants them AI coding agents may make software cheaper to build, but switching off major platforms remains expensive, risky, and deeply annoying Salesforce CEO Marc Benioff appeared on the VC podcast All-In last week to share his vision for the future of AI and software. Among the usual bonhomie – he skipped Trump's trip to China; he's not a Democrat or Republican, just "here to support the country" – he revealed the CRM giant he co-founded may spend circa $300 million with Anthropic in 2026 to harness its coding agents. "Everything's gonna be cheaper to make, it's more efficient," he told the show. "I can do things that I just could not do before. I can go faster than ever before. I can implement my software and sell it at the same time. I can break through obstacles that I've had because I have coding agents and humans working together."There are other consequences of this rush to the LLM-guzzling future. Despite its growth, Salesforce did not hire any more software engineers in 2025. It has also cut around 4,000 support staff, although it has been hiring elsewhere. So what's Salesforce doing with the money it's saving? It’s not going to the customer, it seems. In 2025, Benioff told investors he saw a "very high margin opportunity" in users adopting its AI agent platforms. Later, Miguel Milano, president and chief revenue officer, said he was happy to make a loss on a capped-price deal for AI agents in the short term, as the company had "20 years to monetize that customer." Gartner later warned Salesforce users that a capped enterprise agreement for its AI and data platforms may not be available when they come to renew, potentially meaning customers could struggle to predict costs and understand value. Salesforce responded: "The claim that we are moving away from capped agreements is inaccurate. Renewals remain flexible, and because AI compute costs may actually shift over time, we focus on t...

First seen: 2026-05-26 10:32

Last seen: 2026-05-28 08:05