Customer success ops

Customer success is the function most often built on good intentions and bad systems. One CSM with 40 accounts, a Google sheet, and a lot of goodwill. The teams that scale separate the human work (relationships) from the process work (automation) and build both deliberately. The goal: the CSM spends their time on the conversations that drive retention and expansion, not on the busywork that could be automated.

The CSM's job, clarified

A CSM's job is to drive three outcomes:

1. Adoption, the customer actually uses the product and gets value from it
2. Retention, they renew
3. Expansion, they expand (more seats, higher tier, additional products)

Not: support tickets, billing issues, onboarding logistics, QBR slide-making. Those are support, billing, onboarding, and RevOps respectively. When CSMs get consumed by those, adoption + retention + expansion suffer.

Segmentation, the foundation

Not all customers deserve the same CSM attention. Segment by revenue and strategic value:

• Enterprise ($100K+ ARR), named CSM, QBRs, executive sponsor, custom success plan
• Commercial ($25K–$100K), pooled CSMs, quarterly check-ins, standardized playbooks
• SMB (<$25K), digital-only: in-app, email, shared CSM queue
• Strategic (revenue-agnostic), key logos, case study partners, analyst references, elevated treatment regardless of ARR

The common mistake: giving SMB customers enterprise-level attention. They churn anyway; you just can't afford it.

The success plan

For every named-CSM account, write a success plan within 30 days of signing:

• What success looks like, the customer's definition, quantified
• Milestones. 30, 60, 90 days, and the first annual goal
• Stakeholders, executive sponsor, economic buyer, champion, technical lead, end users
• Risks, what could derail adoption or retention
• Expansion path, how this account grows if things go well

Review it at every QBR. Update as reality evolves.

The customer health score

A composite number combining:

• Usage, monthly active, key feature adoption, trend
• Sentiment. NPS, CSAT, support ticket patterns
• Relationship, executive engagement, champion tenure, QBR attendance
• Commercial, paid on time, contract health, renewal timing

Output: Red / Yellow / Green. Updated weekly. Red = escalation; Yellow = proactive outreach; Green = expansion play.

Simple scoring example.
Each dimension 0–10. Weighted sum. >75 = Green, 50–75 = Yellow, <50 = Red.
Low usage + sponsor just left + open P1 ticket = 28/100 = Red. Trigger a save play today.

The renewal playbook

Renewals shouldn't be a surprise:

• Day -180, health score review; early-warning on at-risk accounts
• Day -120, value recap document; quantify ROI delivered to date
• Day -90, renewal conversation kickoff; begin expansion discussion
• Day -60, commercial terms discussion
• Day -30, contract in legal
• Day 0, renewal signed; kick off year 2 success plan

Renewals worked in the last 30 days are expensive (concessions made under pressure). Renewals worked at -120 days are cheap (you have leverage).

The CS–Sales handoff

Most churn is born in the sales cycle. If sales over-promised, CS inherits an impossible customer. Fix with a structured handoff:

• Sales completes a handoff doc: customer goals, stakeholders, commitments made, known risks, expansion signals identified
• CSM + AE do a 30-minute handoff call with the customer included
• First CSM meeting within 10 days of signature

Tooling

The minimum stack:

• CRM (Salesforce/HubSpot), accounts, opps, renewal forecasting
• Customer success platform (Gainsight, ChurnZero, or spreadsheets at small scale), health scores, playbooks, alerts
• Product analytics (Mixpanel, Amplitude), usage telemetry feeding health score
• Shared inbox / ticketing (Zendesk, Intercom), support + CS visibility

You don't need all of them early. You need the health score visible somewhere by the time you have 50 accounts.

What good looks like

• Every named account has a success plan, reviewed quarterly
• Health scores exist and trigger action
• Renewal motion starts 180 days out, not 30
• CSM to managed-ARR ratio is known and defended ($2M–$5M per CSM is typical)

Related: Churn diagnostics · Pipeline design · SLAs + SLOs