Leading Indicator Models

Leading indicators are signals that predict health score changes before those changes occur. They are the advance warning system within the health model — not the health score itself but the inputs that will move the health score in the next 30-90 days. The most reliable leading indicators vary by industry and business model, but three categories are nearly universal. Behavioral leading indicators: changes in how the customer interacts with you (email response speed, meeting participation, inbound requests). Product leading indicators: changes in how the customer uses what you delivered (feature adoption, usage frequency, support ticket patterns). External leading indicators: changes in the customer's business environment (leadership stability, financial health, competitive pressure, organizational restructuring).

1. Identify Your Leading Indicators Analyze the signal data from accounts that churned or expanded in the past two years. Which signals changed first — before the health score started moving? Those are your leading indicators. Common findings: support ticket sentiment shifts 60-90 days before health score decline; executive meeting cancellations begin 45-60 days before; usage decline begins 30-45 days before. The lead time tells you how far ahead each indicator can warn you.
2. Build a Leading Indicator Dashboard Track leading indicators separately from the health score itself. The dashboard shows the health score (lagging composite), the leading indicators (individual signals that predict future health changes), and the predicted trajectory (where the health score is heading based on the leading indicators). The dashboard gives the account manager both the current picture and the forward view.
3. Set Alert Thresholds on Leading Indicators Define the threshold at which each leading indicator triggers an alert — before it has time to affect the composite health score. Support ticket sentiment drops below a threshold: alert. Executive meeting cancellations exceed two in a quarter: alert. Usage declines more than 15% month-over-month: alert. The alert fires when the leading indicator crosses its threshold, giving the account team intervention time that the lagging health score cannot provide.