Scaling AI Across Support and Sales: Fin Now Sells Itself
Summary
The episode explains how Intercom’s Fin — an AI customer agent that already automates roughly 81% of support interactions — is being extended from support into sales as Fin Sales Agent. Fin Sales Agent acts like a consultant: it opens sales conversations, qualifies prospects, profiles them into buckets (e.g., enterprise), and creates MQLs that feed the sales funnel. The guests describe a careful rollout strategy (A/B tests, closed betas, incremental integrations) and highlight product improvements planned like booking/calendar integration and direct Salesforce sync. The conversation emphasizes aligning sales and support around a unified customer experience while navigating differing KPIs and trust concerns about AI handling early-stage sales tasks.
Key Takeaways
- 1Reusing support AI learnings to power sales accelerates adoption and capability.
- 2Fin Sales Agent functions as a consultative qualifier that generates MQLs and routes leads.
- 3Practical, metric-driven rollout (A/B tests → closed beta → expand) reduces risk and proves value.
- 4Aligning sales and support requires focusing on the holistic customer journey despite differing KPIs.
- 5Trust and governance are central concerns when AI handles prospecting and qualification.
Notable Quotes
"It's designed to essentially be a consultant as part of the sales team. So we have full control over how we want the sales agent to respond to our customer or prospect questions."
"We're after reaching 81% automation which means that out of everything that comes into our customer support organization... 81% of it ... is being resolved by fin service agents at the moment."
"This really feeds into the vision that our CEO Owen spoke about... this concept of a customer agent that does many different jobs but isn't many different agents that are fragmented across different teams."
Episode questions
What is Fin Sales Agent and how does it differ from Fin Service Agent?
Fin Sales Agent is an AI 'consultant' that opens sales conversations, gathers qualifying data, profiles prospects into buckets (e.g., enterprise), and creates MQLs for the sales funnel; Service Agent focuses on resolving customer support queries or routing them to the right team. The sales agent emphasizes data collection, qualification and pipeline progression rather than fast resolution.
How does Fin Sales Agent determine and track qualified leads (MQLs)?
Sales Agent uses predefined criteria and buckets to profile prospects based on their answers; when criteria are met it marks a prospect as an MQL and that lead is tracked in top-of-funnel systems and eventually in Salesforce for downstream sales stages.
What evidence did Intercom use to justify expanding Fin into sales?
They started with internal A/B tests and a closed beta; the tests showed improvements in metrics sales care about (MQL volume and transition rates), which motivated broader rollout and integrations like booking and later Salesforce sync.
What are the near-term product improvements planned for Sales Agent?
Plans include integrating direct Salesforce sync for seamless data flow and reporting (better tracking from MQL to S1/S2 and close outcomes), adding booking/calendar integrations for demos, and exploring additional channels beyond web such as email.