Scaling CSM Coverage Without Diluting Engagement
Your customer success team supports 420 accounts with six CSMs. Segmentation is by annual contract value: 60 strategic, 180 mid-market, 180 small. The health score uses product login frequency and open ticket count; it flags 38% of accounts as red, but last quarter only 9% of red accounts actually churned or downgraded. QBRs are completed for 35% of strategic accounts and 8% of mid-market. Usage telemetry sits in a data warehouse, renewal dates in the CRM, and survey results in a spreadsheet. The board has asked for a retained-revenue forecast and a credible coverage model. CSMs spend most of their time on inbound tickets from small accounts and on custom QBR decks. There is no standard onboarding playbook, and executive sponsor mapping exists for only 10 accounts.
1. Propose a segmentation and coverage model that fits six CSMs and 420 accounts. Justify the design.
Read the worked answer
I would segment by value, complexity, and risk, not ACV alone. A practical model is:
- Strategic: 60 accounts, high-touch. One CSM per 20 accounts, so three CSMs. Coverage includes success plans, executive sponsor mapping, quarterly QBRs, and renewal ownership.
- Mid-market: 180 accounts, tech-touch with pooled coverage. Two CSMs, one per 90 accounts, supported by a digital playbook, standardized QBR, and monthly health review. Accounts with risk or expansion potential get temporary high-touch.
- Small: 180 accounts, digital-led. One CSM as program owner, supported by self-serve onboarding, automated health alerts, and a shared inbox that is triaged away from CSMs.
Justification: the current model lets small-account tickets consume strategic time and custom QBR decks do not scale. The pooled model preserves human coverage where revenue and complexity justify it. I would also define tier movement rules so an at-risk mid-market account can be promoted temporarily. This is a coverage model, not a headcount request. The first proof is that strategic QBR completion rises and small-account inbound ticket time falls without increasing churn.
2. How would you recalibrate the health score to reduce false positives? What data and governance are needed?
Read the worked answer
The current score is too narrow. Login frequency and ticket count measure activity and support load, not health. A 38% red rate with only 9% actual churn or downgrade means the score is not calibrated to outcomes.
I would rebuild it with four components: outcome and adoption, relationship and sponsor strength, support and technical risk, and renewal and commercial context. Each component needs two or three indicators. For example, adoption should include active use of at least one value-driving feature, not raw logins. Relationship should include executive sponsor mapped and last executive contact date. Support should distinguish severity and recurrence, not just ticket count. Renewal context should include days to renewal and contract risk.
Data sources: usage telemetry from the warehouse, renewal dates and account hierarchy from the CRM, survey results from a governed survey tool, and ticket severity from the support system. Governance: a monthly calibration review comparing red accounts to actual outcomes, documented weight changes, and a rule that no score changes without a business rationale. I would also add a manual override with a reason code, reviewed monthly. The goal is not a perfect score; it is a score the team trusts enough to act on.
3. What operating cadence and metrics would you put in place for the next two quarters?
Read the worked answer
Cadence: weekly risk triage for red and renewal-90 accounts, monthly health calibration and coverage review, quarterly QBR cycle for strategic and selected mid-market accounts. I would also add a monthly onboarding review because there is no standard playbook and new accounts are the future renewal base.
Metrics for the next two quarters: gross revenue retention and net revenue retention; renewal forecast accuracy; percentage of strategic accounts with a signed success plan; percentage with executive sponsor mapped; QBR completion by tier; onboarding time to first value; adoption of at least one value-driving feature; red-account conversion rate; and CSM time spent on inbound small-account tickets. I would avoid measuring activity volume such as number of QBR decks.
Governance: one source of truth for account health and renewal dates, with a documented data owner. The board should see retained-revenue forecast with confidence bands, not a single number. I would also set a two-quarter target to reduce false-positive red accounts and increase strategic QBR completion. The operating principle is tiered engagement with evidence, not equal treatment for every account.