The Customer Success Revenue Signals Playbook

Net revenue retention is won in the ninety days before the renewal, not in the meeting. Most CS teams find out about churn at the renewal and miss expansion entirely, because nobody is watching the signals. This is a plug-and-play workflow to score customer health from signals you already have and act early — on both churn and expansion. Google Sheets and Claude. The prompts and the steps are below.

Most customer health systems are a wall of green, yellow, and red that nobody trusts and nobody acts on. The colors feel like progress, but a CSM cannot do anything with “yellow.” By the time an account turns red, the renewal is already lost — the signals that predicted it were visible weeks earlier, in usage that quietly fell off, a champion who changed jobs, a support thread that escalated.

And churn is only half the number. Net revenue retention is just as much about expansion, and the signals that predict expansion — usage hitting a limit, a new team onboarded, the customer's own funding round — are the same kind of triggers you already score for new business. A health system that only watches for risk leaves half of NRR on the table.

What to buy, what to build

Stage 1 — Define the signals

Health is not a vibe, it is a set of named signals pointing one of two ways. Decide which ones you can actually see, then log them consistently.

1. List your risk signals: champion or sponsor departure, declining active users, rising support severity, stalled adoption, payment friction.
2. List your expansion signals: usage near a plan limit, new teams or seats, high adoption, and the customer's own growth (funding, hiring, new leadership).
3. Stand up two tabs: accounts (account, ARR, renewal_date, segment, plan) and cs_signals (account, signal_date, signal_type, source, detail, direction).

Stage 2 — Capture them

Most of these signals already exist in systems you have; the work is routing them to one tab.

1. Usage. Logins, active users, feature adoption, and usage-versus-limit from your product analytics or a usage export.
2. Support & sentiment. Ticket volume, severity, and CSAT from your help desk.
3. Relationship. Champion and sponsor changes — the highest-signal risk there is — tracked the same way the Data Quality Playbook tracks contact job changes.
4. Commercial. Renewal date proximity and any past contract or payment friction, from your CRM.
5. External. Funding, hiring, and leadership changes — the same triggers your new-business motion watches, here pointed at expansion.

Stage 3 — Score account health

Turn the signals into a score, a direction, and a play. Recency and direction are the whole point: a champion who left last week outranks a glowing usage number from last quarter, and an account approaching its plan limit is an opportunity, not a risk.

You are scoring customer account health for net revenue retention. I will paste an account's recent signals. Recent signals count more than old ones.

Signal direction and weight:
RISK (lowers health):
- Champion or exec sponsor left or changed role: very high
- Active users or logins declining over 30+ days: high
- Support tickets spiking or severity rising: high
- Low feature adoption or stalled onboarding: medium
- Past payment or contract friction: medium
EXPANSION (raises health and flags opportunity):
- Usage approaching a plan limit: high
- New teams, seats, or use cases added: high
- High adoption across the account: medium
- The customer's own growth (funding, hiring, new leadership): medium

Recency rule: last 14 days full weight; 15-45 days half weight; older than 90 days, context only.

For each account return, tab-separated:
account  health_score  direction  top_reason  recommended_play
- health_score: 0-100.
- direction: at-risk | stable | expansion-ready.
- top_reason: the single most important signal driving the score.
- recommended_play: one concrete next action.

Ground every score in the signals provided. Invent nothing.

Input, tab-separated, one line per signal, grouped by account:
account  signal_type  signal_date  detail

Rows:
[PASTE SIGNALS]

Stage 4 — Triage by dollars and urgency

A CSM's week is finite. Spend it where dollars and urgency meet: a large account that is at-risk with a renewal inside ninety days comes before everything else; a large account approaching its limit is the expansion to chase. Let the sheet rank it.

// Triage = dollars at stake x urgency. accounts tab: ARR in B, renewal_date in C.
// health tab: account in A, health_score in B, direction in C.
=B2 * IFS(
   AND(VLOOKUP(A2,health!$A:$C,3,FALSE)="at-risk", C2-TODAY()<90), 3,
   VLOOKUP(A2,health!$A:$C,3,FALSE)="at-risk", 2,
   VLOOKUP(A2,health!$A:$C,3,FALSE)="expansion-ready", 1.5,
   TRUE, 1)

// Sort this column descending: the top is where a CSM's hour is worth most this week.

Stage 5 — Run the play

The score tells you which accounts; the play tells you what to do. Match it to the direction — a save play for risk, an expansion play for opportunity — grounded only in the signals, in the same honest voice you use everywhere.

You are drafting a customer success play, grounded only in the account's signals. Match the play to the direction.

If direction is AT-RISK, produce a save play:
- the likely root cause, named honestly from the signals
- the one outcome to re-anchor the relationship on
- a short, honest outreach to the right stakeholder (champion, sponsor, or economic buyer) - no spin, no pretending the problem is not there
- the internal action to take first (e.g. a value review, exec alignment)

If direction is EXPANSION-READY, produce an expansion play:
- the specific signal that opens the conversation (usage near a limit, a new team, their growth)
- the business case in one or two sentences, tied to their outcome, not our SKU
- a short outreach that earns the expansion conversation without pressuring
- who else to multi-thread to (often the economic buyer)

Keep it honest, specific, and first-person. No hype, no false urgency.

Account: [ACCOUNT]
Direction: [AT-RISK | EXPANSION-READY]
Signals and top reason: [PASTE THE ACCOUNT'S SIGNALS]
Stakeholders: [LIST NAMES AND TITLES]

Make it run every week

1. Manual, free. Each week, refresh the signals tab, re-score, and surface the accounts that crossed a line this week — a champion who left, usage that dropped, an account that hit its limit. Act on the threshold crossings, not the whole book.
2. Scheduled, with the API. An Apps Script trigger pulls usage, support, and external signals, re-scores health, and flags accounts whose direction changed — so a CSM opens Monday to exactly the accounts that moved.

What good looks like

• Every account has a health score built from named signals, not a color someone set by feel.
• Risk surfaces 90 days before the renewal, not in the renewal call.
• Champion and sponsor changes are tracked — the single most predictive churn signal is never missed.
• Expansion is watched as closely as churn; usage-at-limit and account growth trigger a play, not just a note.
• Triage is by dollars and urgency, so the team's time lands where the number moves.
• Net revenue retention, not a health-score average, is the metric the system is judged on.

Retention is not a renewal conversation, it is everything that happens in the ninety days before it. Watch the signals in both directions, score them honestly, triage by what is actually at stake, and act before the account knows it had a problem — or before it knows it was ready to grow. That is how net revenue retention gets above one hundred percent and stays there.