Ran an audit across a handful of SaaS/coaching stacks recently and found the same 4 leaks almost every time:

1. No usage-drop detection. Most tools track logins, not engagement depth. A member can "log in" and still be functionally gone. If you're not tracking feature-level usage decay, you're finding out about churn the day they cancel — too late to save them.

2. Automation coverage gaps at the exact moment it matters. Welcome sequences are usually solid. Win-back sequences after a usage drop? Almost never exist. That's the highest-leverage gap to close first.

3. Zero personalization in cancellation flows. Generic "sorry to see you go" flows convert saves at ~2-4%. Segmented flows (by usage pattern, tenure, plan tier) that reference why someone is likely leaving convert at 15-20%+.

4. Tool sprawl inflating cost per retained user. More tools ≠ better retention. In one case study, cutting from 9 tools to 4 and consolidating data into one source of truth actually improved detection speed because signals weren't scattered.

Quick framework if you want to self-audit: score your stack 1-5 on detection speed, automation coverage, personalization, and cost efficiency. Anything scoring under 3 in two or more categories is actively leaking revenue right now.

Happy to go deeper on any of these in the comments.