Salesforce to Acquire Fin (formerly Intercom) for $3.6B

Valuation and deal size

• Many see $3.6B as low for a leading AI support tool, especially given an estimated ~$400M ARR (sub‑10x multiple) and large customer base/brand.
• Compared against “AI-adjacent” valuations (e.g., cursor, Sierra at $15.8B, Decagon at $4.5B), some call it a “steal”; others argue those peers are overvalued.
• There’s debate on what the price implies about the true TAM for AI support and whether it undercuts narratives that AI will replace “millions” of support workers.

Salesforce finances and strategy

• Several comments show Salesforce has ample cash, profits, and free cash flow to afford the deal easily.
• Viewed as a catch‑up move against newer AI support vendors and to prevent independent AI agents from becoming control points outside Salesforce’s CRM.
• Some note prior activist pressure to limit M&A and question why Salesforce is back to big acquisitions despite recent stock underperformance.

Rebrand to Fin

• Intercom rebranded to Fin about a month before the deal; many assume this was coordinated with the acquisition timeline.
• Consensus that Salesforce wanted the “Fin” AI‑agent branding rather than “Intercom,” and will likely rebrand again under its own naming (Einstein/Agentforce).

AI support agents: effective tool or grift?

• Strong split:
  • Critics: current “AI agents” just regurgitate FAQs, don’t handle root causes or edge cases, and often block access to humans; seen as case‑deflection at the cost of user experience.
  • Supporters: report high automation rates of simple tickets, higher CSAT vs humans in some deployments, and real savings by freeing humans for complex cases.
• Risk themes: agents empowered to issue refunds or perform actions can be abused or manipulated; liability and misleading “resolution” metrics (e.g., counting silence as success) are contested.
• Technical claims conflict: some say Fin’s Apex model beats top LLMs on support benchmarks; others assert Fin mostly orchestrates external models (OpenAI/Claude) with RAG, rules, and workflows.

Build vs buy debate

• One camp reports building custom AI support agents or full in‑house helpdesks in days–weeks, arguing SaaS like Fin adds little beyond a UI and is now often more expensive and constraining.
• The other camp stresses that most companies lack the skills/bandwidth to build and maintain reliable, integrated agents at scale, making per‑resolution SaaS economically rational, especially below very high ticket volumes.
• There’s broader frustration about rising SaaS prices and “one‑size‑fits‑all” tools versus highly integrated in‑house systems.

Customer experience and philosophy of support

• Many users dislike AI/chatbots and intentionally seek companies that provide fast human support; some would rank or choose vendors based on this.
• Others have had excellent AI experiences (e.g., fast refunds, immediate account‑aware help) and argue the main issue is poor execution, not the concept.
• Recurrent point: support interactions generate valuable product insight; over‑automation risks losing this signal and degrading long‑term quality.

Perceptions of Intercom/Fin and Salesforce

• Intercom is remembered fondly as an early, elegantly executed “chat bubble + help center” product that defined a category; some feel the heavy AI pivot made it unrecognizable.
• Multiple commenters fear Salesforce will neglect or enshittify Fin, citing Heroku and describing Salesforce products as bloated, confusing, or “enterprise graveyard” material.
• Others simply congratulate the team on a large but perhaps smaller‑than‑expected exit and see this as a logical consolidation moment in AI support.

Leadership and ethics concerns

• Some participants revive past reporting about the CEO’s alleged misconduct and controversial political activity, expressing discomfort that this did not prevent a large financial win.
• Others push back on bringing politics into the thread, but the ethical dimension remains a visible undercurrent in the discussion.