Before you change anything
Three systems do the work. When customers churn, escalations pile up, or renewals surprise you, one of these is usually missing or disconnected from the others. Build them once, then run everything through them.
Lifecycle architecture
Defines what success looks like at every stage, from first contact through renewal, and who owns each handoff so no customer falls between teams.
Operational governance
Escalation tiers, severity definitions, and SLAs that keep issues from aging quietly in a queue, with clear ownership from first touch to resolution.
Value reporting
The metrics, dashboards, and business reviews that prove ROI to customers and connect customer experience to revenue for your executive team.
The workflow, in order
Seven steps to stand up or rebuild a customer operation. Work through them in sequence. Your progress saves in this browser.
Run a baseline diagnostic
Before redesigning, measure what is actually happening. Pull twelve months of churn, renewal, ticket, and satisfaction data, then talk to the people closest to the problem.
More detail
- Segment churn by reason, tenure, and customer size. Averages hide the story.
- Find the oldest open tickets. Aging backlog is the fastest signal of broken ownership.
- Ask frontline staff where they work around the process. Workarounds show you where the design failed.
Map the customer lifecycle
Define each stage with entry criteria, exit criteria, an owner, and the handoff to the next team. If a stage has no owner, it has no outcome.
More detail
- Typical stages: sales handoff, onboarding, adoption, value realization, renewal, expansion.
- Write exit criteria as customer outcomes, such as "first report delivered to their executive," not internal tasks.
Define success and health
Agree with each customer segment on what success means, then build a health score that predicts risk early enough to act on it.
Build escalation and SLA governance
Set severity levels, tier ownership, response and resolution targets, and a clear path to executive escalation. Then audit the backlog against it.
More detail
- When a queue is far behind, reach out to affected customers personally. Acknowledging a lapse rebuilds more trust than silence does.
- Track SLA breaches weekly, not monthly. Breaches compound.
Instrument the reporting
Build one source of truth for retention, service, and value metrics. Automate it so the team spends its time acting on data, not assembling it.
Enable the team
Give frontline and customer-facing teams runbooks, troubleshooting guides, and self-service content so good outcomes do not depend on who picks up the ticket.
Run the value cadence
Hold regular business reviews, risk reviews, and expansion conversations. Customers renew what they can see, so make the value visible on a schedule.
Template library
Copy any template into your own tools and fill in the highlighted fields. Each one is a starting structure. Adjust weights, targets, and wording to your business.
Customer health score framework
Use this when you need an early warning system for churn risk, or your current score is not predicting what actually happens.
CUSTOMER HEALTH SCORE FRAMEWORK Segment: [Enterprise / Mid-market / SMB] Review cadence: [Weekly for red, monthly for all] SIGNAL WEIGHT GREEN YELLOW RED Product adoption 30% [>70% of seats active] [40-70%] [<40%] Outcome progress 25% Success criteria Partially met Not met or met on schedule undefined Support experience 15% No open P1/P2 Aging tickets Open escalation Relationship strength 15% Exec sponsor Champion only No active engaged contact Sentiment (CSAT / survey) 15% [>=90%] [75-89%] [<75%] SCORING Green = 3 points, Yellow = 2, Red = 1. Multiply by weight and total. Overall: 2.5+ Healthy | 1.8-2.49 Watch | Below 1.8 At risk REQUIRED ACTIONS At risk: recovery plan within 5 business days, owner named, leadership visibility Watch: CSM outreach within 10 business days, root cause noted in CRM Healthy: identify expansion or advocacy opportunity at next review CALIBRATION Every quarter, compare scores from 90 days ago with actual renewals and churn. If the score missed, change the weights. A health score is a hypothesis.
Escalation matrix and SLA governance
Use this when escalations depend on who is available, tickets age without owners, or customers escalate straight to your executives.
ESCALATION MATRIX AND SLA GOVERNANCE Team / platform: [Team name, ticketing system] SEVERITY DEFINITIONS P1 Critical Service down or data at risk for multiple customers P2 High Core workflow blocked for one customer, no workaround P3 Medium Degraded experience, workaround available P4 Low Question, request, or cosmetic issue TIERS AND OWNERSHIP Tier 1 Frontline support Triage, known fixes, knowledge base Tier 2 Specialists Complex troubleshooting, configuration Tier 3 Engineering/Product Defects, platform changes Exec [Director or VP name] P1s, at-risk accounts, customer executive requests SLA TARGETS Severity First response Update cadence Resolution target P1 [30 min] [Every 2 hours] [8 hours] P2 [2 hours] [Daily] [2 business days] P3 [1 business day] [Every 3 days] [5 business days] P4 [2 business days] [On change] [10 business days] RULES 1. Every ticket has one named owner at all times. 2. A breached SLA escalates automatically one tier up. 3. P1s trigger a customer communication within the first-response window. 4. Weekly review: breaches, oldest open tickets, repeat drivers. 5. Monthly: top three root causes go to Product with customer impact attached.
At-risk account recovery plan
Use this when a customer questions the ROI they are getting, a renewal is in doubt, or an executive sponsor has gone quiet.
AT-RISK ACCOUNT RECOVERY PLAN Account: [Name] Annual value: [$] Renewal date: [Date] Recovery owner: [Name] Executive sponsor (ours): [Name] 1. WHAT THE CUSTOMER BELIEVES In their words, what is not working? [Quote or summary] What did they expect when they bought? [Original success criteria] 2. WHAT THE DATA SHOWS Adoption: [ ] Open issues: [ ] Delivered outcomes to date: [ ] Gap between expectation and delivery: [ ] 3. ACKNOWLEDGE What we own and will say plainly in the first conversation: [ ] 4. RECOVERY ACTIONS (30 / 60 / 90 days) 30 days: [Fix the most visible issue, agree on revised success criteria] 60 days: [Deliver first measurable result, share value report] 90 days: [Business review with their executive, renewal conversation] 5. VALUE REPORTING Metrics we will report, how often, and to whom: [ ] 6. EXIT CRITERIA The account moves back to healthy when: [ ]
Executive business review agenda
Use this when you want business reviews that customers' executives actually attend, and that lead to renewal and expansion conversations.
EXECUTIVE BUSINESS REVIEW (45 minutes) Customer: [Name] Attendees: [Their exec, champion; our exec, CSM] 5 min Their priorities Confirm what matters to them this quarter. Ask, do not assume. 10 min Outcomes delivered Results against the success criteria agreed at onboarding, in their business terms: time saved, cost avoided, revenue supported. 10 min What we learned Adoption trends, support themes, and one honest area to improve. 10 min What comes next Recommendations tied to their priorities, including any expansion that solves a stated problem. 5 min Commitments Named owners and dates on both sides. 5 min Feedback How are we doing as a partner? What would make us indispensable? FOLLOW-UP within 2 business days: summary, commitments, next review date.
AI prompt: customer operations analyst
Use this when you have a ticket export, survey results, or churn notes and want a structured first read before your leadership review. Remove customer names and personal data before pasting anything into an AI tool.
You are a senior customer operations analyst advising a [VP of Customer Success / COO] at a [company type] serving [customer segment]. I will paste [ticket data / survey responses / churn notes] covering [time period]. All personal data has been removed. Analyze it and return: 1. The top five drivers of volume or dissatisfaction, with counts or share of total. 2. For each driver: likely root cause, whether it is a product, process, or people issue, and the evidence. 3. Leading indicators of churn risk you can see in the data. 4. Three fixes ranked by impact and effort, each with an owner function and a metric to track. 5. What the data cannot tell us, and what I should collect next. Rules: - Separate what the data shows from what you are inferring. Label inferences. - Do not invent numbers. If something is not in the data, say so. - Write for an executive audience: concise, specific, decision-ready. Data: [paste here]
Operating cadence
Systems only work if they run on a rhythm. This is the minimum cadence that keeps governance, risk, and value visible.
| Rhythm | Review | Owner | Output |
|---|---|---|---|
| Weekly | SLA breaches, oldest tickets, red accounts | Operations lead | Owners and dates on every exception |
| Monthly | Health score movement, top ticket drivers, renewals due in 90 days | CS leadership | Recovery plans, root causes sent to Product |
| Quarterly | Business reviews, retention and expansion results, health score calibration | VP or executive sponsor | Executive report, updated targets |
| Annually | Lifecycle design, segmentation, team structure, tooling | Executive team | Operating plan for the next year |
Metrics and starting targets
The targets below are the starting points I use when building a customer success framework. Treat them as a baseline to calibrate against your own data, not universal benchmarks.
Definitions
| Metric | How to calculate | Why it matters |
|---|---|---|
| Gross revenue retention | (Starting recurring revenue minus churn and downgrades) ÷ starting recurring revenue | Shows how well you keep what you have, with expansion excluded |
| Net revenue retention | (Starting recurring revenue minus churn and downgrades plus expansion) ÷ starting recurring revenue | Shows whether existing customers grow the business |
| Logo churn rate | Customers lost in period ÷ customers at start of period | Counts relationships lost, regardless of size |
| First contact resolution | Issues resolved on first contact ÷ total issues | A direct read on enablement and knowledge quality |
| SLA attainment | Tickets meeting response and resolution targets ÷ total tickets | Tests whether governance holds under load |
| Time to value | Days from contract start to first agreed success milestone | The earliest predictor of renewal |
Failure modes to design out
Most customer operations break in predictable places. Name the ones most likely in your organization and build a control for each.
Lifecycle
- Onboarding with no agreed success criteria
- Sales-to-CS handoff that loses context
- Renewal work that starts 30 days out
Governance
- Tickets with no single owner
- Escalation that depends on who is online
- Backlog nobody reviews by age
Data
- Vanity metrics in executive reports
- Health scores never checked against outcomes
- Three teams, three versions of churn
People
- Outcomes that depend on heroics
- No runbooks for top ticket drivers
- Frontline feedback that never reaches Product
Customer
- Silence mistaken for satisfaction
- Value delivered but never reported
- Executive sponsor who left unnoticed
Readiness checklist
Before you scale headcount, customers, or tooling, confirm these are in place. Your progress saves in this browser.
Principles behind the playbook
The ideas I come back to every time I build or rebuild a customer operation.
Customer success is an operating system, not a department
Retention is decided by how Sales, Product, Support, and Finance work together. A CS team alone cannot fix a broken handoff.
Design beats effort
If good outcomes depend on people working harder, the system is the problem. Build processes that make the right result the easy result.
Empathy is a strategy that scales
Understanding what customers expected, and saying plainly where you fell short, rebuilds trust faster than any discount.
One size does not fit all
Segment by what customers need to succeed, not only by what they pay. Different segments need different motions.
Silence is not satisfaction
Customers who stop complaining have often stopped caring. Measure engagement, not only the absence of tickets.
Make value visible
Customers renew what they can see. If you deliver outcomes but never report them, you have left the renewal to chance.