Rhys Fisher sits down with Simon Leyland, CEO and Co-Founder of Cloud Interact, and Zeus Kerravala, Founder and Principal Analyst at ZK Research, to dig into one of the biggest CX deals of the year — Salesforce’s agreement to acquire Fin for $3.6BN.
The reaction from both guests is telling. For Kerravala, the move was unsurprising given Salesforce’s aggressive acquisition streak — upwards of ten companies in 2025 alone — but he’s quick to point out what makes this one different. Fin brings Salesforce a proven, out-of-the-box AI customer service platform with 30,000 customers and resolution rates hovering around 76%, compared to AgentForce’s baseline of roughly 60%. It’s not just a headcount win; it fills genuine product gaps.
Leyland, speaking from the coalface, frames it as market confirmation. A year ago, 80% of client conversations centred on contact centre migration. Today, that ratio has completely flipped, with 80% now talking about AI. Salesforce, he argues, is scrambling to meet demand that already exists, not create it.
“The conversations that we had last year, maybe 80% were talking about migrating contact centres, 10, 15, 20% were talking about AI. This year, that has totally flipped,” he explained.
“It’s 20% migration, 80% AI.”
But the discussion goes well beyond the headline number. Both guests push back on how AI deflection rates are reported, drawing a sharp distinction between queries that simply need information and those that require action — think prescription refills involving identity verification, medical compliance, and logistics. Resolution rates that look impressive on paper may be heavily skewed by the simpler end of the spectrum, and Leyland shares real-world field data that puts that in sharp relief.
Data quality, arguably the unglamorous engine behind any AI platform’s performance, is pointed out by Kerravala as the chronic problem of poor CRM data built up over years of manual, low-effort input. As a result, this has dragged AgentForce’s numbers down through no real fault of its own.
One thread that didn’t get much mainstream coverage post-announcement is the Apex model, Fin’s proprietary small language model. Both guests make a compelling case for why specialized, vertical-specific models are the future of AI in CX, and why Apex could quietly become one of the most strategically valuable parts of what Salesforce just bought.
The CCAS implications, the SaaS valuation squeeze, and where the hyperscalers fit into all of this round out a conversation that covers far more ground than the deal itself.
