Yes, from evidence rather than a gut-feel score: usage signals, support patterns, commitment status, and the warning signs that preceded past churn, connected across systems and surfaced before the crisis.
Health scores fail when they live in one tool. Heidi's read is cross-source: the CRM says fine, the tickets say frustrated, the usage says declining, and she connects the three.
Pattern memory is the differentiator: 'Acme shows the same warning signs Meridian did before churn' is a real Heidi answer.