The problem
A SaaS company's customer success team of six was responsible for approximately three hundred accounts. The team was supposed to monitor account health, engage proactively with at-risk customers and prepare for renewal conversations. In practice, they were reactive — responding to issues as they arose rather than anticipating and preventing them.
The problem was not capability. The team knew what they should be doing. The problem was that monitoring three hundred accounts manually was impossible. By the time a customer showed visible signs of churn risk — declining usage, reduced engagement — they were often already disengaged and difficult to recover.
What we built
We built an AI customer success assistant that monitored account health across the entire customer base:
- Health monitoring. The assistant tracked usage patterns, engagement signals and support activity across every account, flagging those showing early signs of risk.
- Renewal preparation. For upcoming renewals, the assistant compiled account history, usage data, value delivered and any outstanding issues into a briefing for the success manager.
- Proactive engagement suggestions. The assistant identified accounts that would benefit from proactive engagement — new feature adoption, best practice sharing, executive check-in — based on their usage patterns and stage.
- Administrative automation. Routine administrative tasks — meeting scheduling, follow-up tracking, account notation — were handled by the assistant.
The results
Five months after deployment:
- At-risk accounts identified and engaged before churn signals became visible, improving retention
- Success managers spent more time on strategic customer conversations and less on administrative monitoring
- Renewal preparation became consistent and thorough rather than rushed and variable
- The team successfully managed a growing account base without adding headcount
How it worked
The assistant did not replace the customer relationships. It provided the visibility and preparation that made those relationships more effective. The success managers still had the conversations, built the trust and solved the problems. They just did so with better information and more time for the work that mattered.
The assistant's health scoring improved as it learned which signals actually predicted churn and which were noise. The initial scoring was based on industry patterns. The refined scoring was based on the company's specific customers and their actual behaviour.
What we learned
The most important lesson was that proactive customer success requires visibility that manual monitoring cannot provide. The team knew which customers to worry about — the ones who were already complaining. They did not know about the ones who were quietly disengaging until it was too late. The assistant provided the visibility to engage before disengagement became churn.
We also learned that customer success automation works best when it supports rather than replaces the human relationship. Customers valued their success manager. They did not value waiting for routine information that the assistant could provide immediately.
For a broader look at how AI supports SaaS operations, see AI automation for SaaS companies. For a practical example of automated onboarding, see automating customer onboarding with AI.
This case study describes a composite of real implementations. Results vary based on the specific process, team and context.