Salesforce announced a $3.6 billion acquisition of Fin, the support software company that built its brand as Intercom and rebranded around its AI agent a month ago. Fin sells customer service automation, not just a chat widget. The pitch is an agent that can answer questions, pull account context, take actions like refunds or troubleshooting, and escalate when needed. That framed the whole reaction. People were split on whether the deal shows strength or surrender. A lot of readers expected a much bigger outcome given Fin’s profile, its Anthropic deployment, and the broader hype around AI support. Others pointed out that a roughly sub-10x revenue multiple is still a solid exit for a mature SaaS business and says more about today’s disciplined buyers than about product failure.
The sharpest conversation was about whether AI support is actually good now. The consensus landed in a pragmatic middle. When the issue is routine and the bot has access to account data, AI can beat the old support experience by being instant, consistent, and willing to take simple actions. That is why Amazon-like refund flows and straightforward account tasks get decent marks. The failure mode is just as clear. If the company uses AI to front-load denial, hide policy, or block escalation, users hate it fast. Several examples made the point. Bots that loop on
FAQ text, invent shipping answers, or route a locked-out customer to mail a physical letter are worse than no support because they add friction without agency. A recurring theme was that the thing people actually want from support is not conversation. It is permission, exception handling, and action.
That led straight into the biggest strategic divide. One camp argued that support vendors still matter because enterprise-grade automation is hard. The hard part is not summarizing help docs. It is safely wiring an agent into real systems, giving it enough customer context to act, keeping latency low, preventing
prompt injection or jailbreaks, and creating operational tooling so non-engineers can see failures and improve coverage. People with hands-on implementation experience said Fin’s product shines there, especially in the feedback loops around unresolved questions and help center updates. The other camp said those same claims are exactly why this category will get squeezed. In their experience, the real work is company-specific context engineering, tooling, and escalation logic that lives outside the vendor anyway. Once you do that work, the model and
UI feel interchangeable. For teams with decent engineering talent, building a tailored support agent or even replacing the ticketing layer outright now feels much cheaper and faster than it did a year ago.
Salesforce itself drew more skepticism than excitement. Many saw the acquisition as a catch-up move against Sierra and Decagon, and as a defensive way to keep AI support from becoming a control point outside the
CRM. People who liked Intercom worried that Salesforce will fold a startup-friendly product into the usual enterprise maze, much as critics say happened with
Heroku. The broad mood was that Fin probably built one of the better products in the space, and that is exactly why the acquisition is notable. It still did not convince people that AI support is a winner-take-all gold mine. It convinced them the category is real, useful in narrow bands, and headed toward consolidation fast.