Last week at the Retention.com off site in NYC, I presented on how we're using AI across RB2B. Here's what's working and why these approaches transfer to any SaaS business.
Not the flashy stuff, but the actual systems quietly making us more efficient every day.
Three Areas Where AI Actually Helps Us
1. Support Efficiency: Training AI Like a New Teammate
We treat AI like any new support hire. We feed past tickets into GPT to draft KB articles, use Intercom's Suggestions to guide us on knowledge gaps, and organize training through structured Topics and Guidance.
We've built n8n and Fin workflows for tasks like bounce-list removal and event access, which increased our resolution rate by ~10%. We approach this as automating with AI, not just deploying a bot.
2. Support → Product Feedback Loop
We stopped viewing support tickets as costs and started treating them as signals. AI helps us spot recurring pain points faster than manual review.
Real examples that became product improvements:
- Account mergers and user migration flows
- Multi-domain support functionality
- Better webhook error logging
- Clearer subscription details and cancellation flows
3. AI as a Data Lens
Beyond support, we use AI to spot operational trends: proxy and bot detection patterns, email validation issues, and other systemic inefficiencies that would take humans much longer to identify.
What We've Learned
If you document better, train smarter, and listen harder to your data, AI handles the heavy lifting while you focus on strategy.
