This tech giant is betting big on AI. But will companies pay for it?

0
1

San Francisco: Salesforce is going all in on its flagship artificial intelligence product – Agentforce – which the software giant’s billionaire boss says is leading a “digital labour revolution”, but the jury is out on whether customers will open their wallets for the service.

Salesforce is the firm behind the customer relationship systems used by companies worldwide and Agentforce represents its bid to automate business processes using AI “agents” – digital workers that can handle tasks autonomously, from answering customer service queries to managing sales leads and personalising product recommendations.

Unlike traditional chatbots that follow rigid scripts, these agents can reason through problems, pull data from multiple sources and take action without human supervision.

Marc Benioff, chief executive officer of Salesforce., speaks during the 2025 Dreamforce conference in San Francisco.

Marc Benioff, chief executive officer of Salesforce., speaks during the 2025 Dreamforce conference in San Francisco. Credit: Bloomberg

There are questions about the degree to which the agents will replace human workers, and more immediately for Salesforce, how many companies will be willing to pay for them.

At Dreamforce, Salesforce’s flashy annual AI summit, the company announced it has amassed 12,000 Agentforce contracts since launching the product roughly a year ago. Yet only around 6000 of those are paying customers – barely 4 per cent of its 150,000-strong total customer base – highlighting a gap between experimentation and monetisation that’s making Wall Street nervous.

Salesforce’s stock has fallen 25 per cent over the year to date even as the broader tech sector climbs.​

“The rate of innovation is so fast it has outpaced the rate of customer adoption,” Salesforce chief executive Marc Benioff told this masthead via text message. He recently announced the company’s third-quarter fiscal 2025 results, which showed revenue of $US9.44 billion, up 8 per cent year-over-year.

“Agentforce is at the heart of a groundbreaking transformation,” he said on the earnings call.

Investors remain sceptical. Investment bank Jefferies, while maintaining a “buy” rating on Salesforce with a $US375 price target – nearly 50 per cent above current levels – warned the stock would not get the re-rating investors hope for until the company proves more customers are paying for AI products.​

Advertisement
Attendees outside Moscone Center ahead of the 2025 Dreamforce conference.

Attendees outside Moscone Center ahead of the 2025 Dreamforce conference.Credit: Bloomberg

At Dreamforce, executives made their most aggressive pitch yet for the enterprise, where AI agents handle complex workflows autonomously. Customer showcases highlighted early wins: PepsiCo says it will be “agentic-first” by 2026, while American homewares brand Williams-Sonoma has deployed agents to personalise recipes and product recommendations.

“I’m shocked that more people aren’t adopting these things,” Williams-Sonoma CEO Laura Alber said.

New Zealand’s largest telco OneNZ said it has deployed nearly 100 agents across sales, service and network operation and seen five-times return on AI investment within months.​

“Customers are engaging with the agentic journey of plan changes four times more than the old journey,” OneNZ CEO Jason Paris told journalists. The telco achieved 60 per cent faster marketing campaign creation and 20 per cent cost savings on mobile network energy consumption, he said.

There have also been challenges. OneNZ allocated 25 per cent of its AI budget to employee training and acknowledged some staff were “freaking out” as AI consumed up to 50 per cent of their roles.

“AI is coming to your role, not for your role,” Paris said, though he admitted the company was still working out exactly how its workforce structure will need to change.​

Paris said building OneNZ’s first agent took eight hours, but deployment took two weeks “because our data wasn’t great and we had to integrate it into our existing IT infrastructure. It wasn’t the Agentforce tool that wasn’t ready – it was the organisation’s ability to consume it.”

Robin Washington, Salesforce’s chief operating and financial officer who joined six months ago, acknowledged the technology’s innovation was “outpacing adoption”.

“We see our customers at early stages, moving down the flywheel,” Washington said. She pointed to Salesforce’s own use of Agentforce in customer support, which now handles 1.8 million inquiries monthly with a 77 per cent resolution rate.​

Pricing complexity hasn’t helped. Salesforce initially charged $US2 per conversation but introduced multiple models this year: a $US0.10-per-action credit system, $US125-per-user monthly subscriptions for standard editions and $US150 for premium industry solutions. The company also offers pay-as-you-go, pre-commit and pre-purchase options.

Salesforce president Robin Washington.

Salesforce president Robin Washington.Credit: Suppllied.

Jayesh Govindarajan, Salesforce’s executive vice president of AI, said Agentforce adoption had been “unprecedented in any enterprise software”, noting the 12,000 customer figure represents deep enterprise implementations, not superficial deployments.

“If you’d asked me a year ago where we’d be, I wouldn’t have been able to say we’d have 12,000 customers,” he said.

Still, Jefferies research found 70 per cent of customers are pushing back on Salesforce’s recent 6 per cent price increases, with many questioning whether AI add-ons justify the extra spend. The firm also noted Salesforce’s core Sales Cloud and Service Cloud products are growing at just 8 per cent annually, well below newer offerings but still accounting for most revenue.​

To shore up confidence, Salesforce announced a $US7 billion share buyback over the next six months and $US15 billion in San Francisco investments over five years. The company also expanded partnerships with OpenAI and Anthropic to embed advanced AI models into Agentforce, and unveiled a bullish forecast of over $US60 billion in revenue by 2030 – a projection that would require reaccelerating growth above 10 per cent annually after months of sluggish single-digit expansion.

Fierce competition looms. Microsoft, Google, OpenAI and a wave of AI-native start-ups are all vying for the same enterprise AI budgets.

“It’s the fastest growing product in our history. There’s never been a faster growing product that we have,” Benioff said.

“This is a moment where technology innovation is outstripping customer adoption. Our job is to get those customers into adoption mode.”

David Swan travelled to San Francisco as a guest of Salesforce.

Most Viewed in Technology

Disclaimer : This story is auto aggregated by a computer programme and has not been created or edited by DOWNTHENEWS. Publisher: www.smh.com.au