Use Gainsight's AI Health Scores to Catch At-Risk Accounts Early

What This Does

Gainsight's AI-powered health scoring continuously monitors customer usage, engagement, support activity, and relationship signals — then surfaces at-risk accounts automatically with suggested next best actions. Instead of manually checking 50 accounts every week, Gainsight alerts you when something changes, so you can be proactive before a customer is already unhappy.

Before You Start

• Your company uses Gainsight (confirm with your CS manager — Gainsight is enterprise-priced)
• Health scores and data integrations are configured by your CS Ops team
• You have at least a CSM-level user account in Gainsight
• Time needed: 15 minutes to learn the dashboard; 10 minutes/week for ongoing monitoring
• Cost: Company license — typically $30,000+/year enterprise; individual CSMs don't purchase it

Steps

1. Access your account dashboard

Log in to Gainsight at your company's Gainsight URL (usually [company].gainsight.com). Click My Accounts or Accounts in the left navigation to see your portfolio.

What you should see: A list of your assigned accounts with colored health score indicators — typically Green (healthy), Yellow (at risk), Red (critical). Troubleshooting: If you don't see health scores, your CS Ops team may need to assign you to the accounts. Contact them or your CS manager.

2. Sort by health score to find at-risk accounts

Click the Health Score column header to sort accounts by score (lowest first). All red and yellow accounts appear at the top. This is your weekly priority list — accounts that need attention right now.

What you should see: Your at-risk accounts sorted with the lowest scores first, with health score breakdown by category (Product Usage, Engagement, Sentiment, etc.).

3. Review a health score breakdown

Click on a yellow or red account. In the account view, find the Health Score section. Click to expand it — you'll see which factors are pulling the score down:

• Product Usage: Login frequency, feature adoption depth, daily/monthly active users
• Engagement: Time since last CSM call, email response rates, QBR completion
• Relationship/Sentiment: NPS scores, executive sponsor engagement
• Support: Open ticket volume, severity, resolution time

This tells you exactly what's wrong — not just "the score is red" but "usage dropped 40% in the last 30 days and they haven't responded to 2 emails."

4. Use AI-suggested next best actions

After reviewing a health score, look for Suggested Actions or Call to Action section (labeled differently by Gainsight version). Gainsight's AI analyzes the account's situation and suggests:

• "Schedule a check-in call — last contact was 3 weeks ago"
• "Send product adoption resources for [unused feature]"
• "Escalate to your CS manager — revenue at risk exceeds $X"

Click Create CTA to turn the suggestion into an assigned task. This is your action item queue — work through it each morning.

5. Set up a weekly health score review habit

Every Monday morning, open Gainsight → sort by health score → scan your reds and yellows. For each at-risk account, either:

• Act immediately: Create a CTA (call to action) to reach out
• Note and monitor: Add an internal note about why the score is low and when you expect it to improve (e.g., "customer is in implementation freeze for 2 weeks")
• Escalate: Tag your CS manager if the account is high-value and score is critically low

Total time for weekly health review: 10–15 minutes for a 40-account portfolio.

6. Use Journey Orchestrator for automated at-risk outreach

If your CS Ops team has configured Journey Orchestrator (Gainsight's automation engine), at-risk outreach emails can be triggered automatically when a health score drops below a threshold. Ask your CS Ops team if there are existing playbooks you should activate for your accounts.

Real Example

Scenario: Monday morning, you open Gainsight and see that one of your $120K renewal accounts dropped from Green to Yellow over the weekend.

What you see in Gainsight: Health score breakdown shows Product Usage dropped from 72% to 41% active seats. Last CSM contact was 18 days ago. An open P2 support ticket hasn't been resolved in 6 days.

Suggested action: "Contact account — usage decline detected. Open support ticket requires follow-up."

What you do: Create a CTA to call them today. Before the call, use the account timeline to see what changed 2–3 weeks ago (when the drop started). Prepare a check-in script that acknowledges the support ticket and asks what's driving the reduced usage. Total prep time: 12 minutes.

Tips

• Your health score is only as good as your data — if Gainsight's product usage integration hasn't been set up by CS Ops, lobby for it; it's the most valuable signal
• Don't rely solely on the score — it's a lagging indicator for relationship signals. You often know an account is at risk before Gainsight does because of conversation tone or escalation frequency
• Use the Notes section in each Gainsight account record to log what you know about account context that the score can't capture (e.g., "budget freeze in Q2 — don't count low engagement against score")

Tool interfaces change — if a button has moved, look for similar AI/magic/smart options in the same menu area.