CONTEXT(CLONE)
Sign in

Renewal risk review across the whole customer book

For: Head of customer success or renewals manager preparing the quarter's renewal plan

Pattern: Map, verify, reduceNeeds a connectorDesigned for 30 to 600 agents
Get early accessSign in to use

The pain today

Health scores are a colour in a dashboard. The real signals sit in tickets, call notes and emails that no manager has time to read for every account, so risk is noticed at the cancellation notice.

The ask

“For every account renewing next quarter, read the connected tickets, call notes and success plans. Tell me which accounts show real risk, what the evidence is in the customer's own words, and where the notes disagree with the health score.”

Plain words, as you would say it to a colleague. Edit it to fit your case before you send it.

What you attach or connect

  • Customer records with renewal dates and health scores
  • Tickets per account
  • Call notes and transcripts
  • Success plans and usage summaries

The unit of work

One worker task per one account with its recent history.

Why a swarm fits

Accounts are independent. A worker needs one account's recent notes and tickets, never the whole book, and risk claims can be checked against quoted customer statements.

Not for

One strategic account: its success manager already knows more than any document trail.

The decision tree

5 typed decisions, each with an action for every answer

At fixed moments in a run, the engine puts one narrow question to a decision model. The decision model never writes text: it answers yes or no with a probability, picks from listed options, or gives a score, about a small slice of the material. The engine then does exactly what this tree says, which is what makes the run auditable. The thresholds are the template's design values, not measured results.

  1. Planner, while planning

    Scope checkA choice among options

    Before work starts on a unit

    Does this account renew in the period asked about, and is there recorded contact in the review window?

    Sees only: The account's renewal date and its dated activity list

    Why: Silence is reported as unknown, never as healthy.

    • Renews in the period, with recent contactthenAccept
    • Renews in the period, little or no contactthenMark unresolved
    • Renews outside the periodthenSkip this unit

  2. Before workers, before a task runs

    Small worker or strong modelYes or no, with a probability

    Before a task runs

    Is this account's recent history short enough, and in one language, for a small worker to read whole?

    Sees only: The account's record count, total length and languages

    Why: Large, tangled accounts get a strong reader; the rest of the book stays cheap.

    • Yes: 0.60 or higherthenAccept
    • Unsure: 0.30 up to 0.60thenEscalate to a strong model
    • No: below 0.30thenEscalate to a strong model

  3. After workers, the judge checks

    Evidence checkYes or no, with a probability

    After a worker answers

    Is the quoted sentence from the customer, dated within the window, and does it express the risk the worker named, such as a budget cut or a competing evaluation?

    Sees only: The quoted sentence with speaker, date and source record, and the worker's risk label

    Why: A risk list built on the success manager's own worries tells the manager nothing new.

    • Yes: 0.85 or higherthenAccept
    • Unsure: 0.50 up to 0.85thenEscalate to a strong model
    • No: below 0.50thenReject and retry

  4. Reconciler, while merging

    Conflict checkA choice among options

    While reconciling

    How does the quoted evidence compare with the recorded health score?

    Sees only: The account's verified signals and its health score

    Why: The gap between dashboard and documents is the finding, so it is shown, not resolved.

    • Evidence and score agreethenAccept
    • Score is healthy but evidence shows riskthenMark unresolved
    • Score is poor but evidence shows nonethenMark unresolved

  5. Accountable person, before anything is settled

    Person decidesYes or no, with a probability

    Before anything is reported as settled

    Does this account summary point towards a commercial step, such as a discount, a contract change or an executive call?

    Sees only: One account's verified summary

    Why: The account owner decides the play and any offer.

    Accountable: The account owner decides the play and any commercial offer. Notes contain personal data and need access limits.

    • Yes: 0.35 or higherthenAsk a person
    • Unsure: 0.10 up to 0.35thenAsk a person
    • No: below 0.10thenAccept

The fleet: who does what

Model tiers by role, not brands: you choose the models. Strong reasoning models plan and reconcile, small fast models do the wide work, and the judge is a decision model from a different family, so it does not share the workers' blind spots.

  1. Planner

    A strong reasoning model defines the risk signals to look for and scopes each account's history to the relevant period.

    Decisions here:1. Scope check

  2. Workers

    Small fast workers from an open-weight family each read one account and report signals with dated quotes.

    Designed for 30 to 600 agents, one worker task per one account with its recent history. Each worker receives only its own unit.

    Decisions here:2. Small worker or strong model

  3. Judge, from a different model family

    A decision model from a different family checks each risk claim is supported by what the customer actually said.

    Decisions here:3. Evidence check

  4. Reconciler

    A strong reasoning model orders accounts, contrasts evidence with the health score and marks thin-evidence accounts.

    Decisions here:4. Conflict check

  5. Accountable person

    The account owner decides the play and any commercial offer. Notes contain personal data and need access limits.

    Decisions here:5. Person decides

Checked before anything is accepted

  • Each risk signal carries a dated quote and its source record
  • Accounts with little recent contact are reported as unknown, not healthy
  • Disagreement between notes and health score is shown, not resolved

What comes back

  • Accounts at risk with quoted evidence
  • Accounts where notes and health score disagree
  • Accounts with too little contact to judge
  • Recurring risk themes across the book, with cited accounts

What to measure

  • Flagged accounts the owner agrees are at risk
  • Churned accounts that were not flagged
  • Manager hours per book review
  • Cost per account reviewed

Names of measures only. No result is claimed for this template.

Templates open in the workspace chat with the ask filled in. Nothing runs until you send it.

Get early accessSign in to use

More in Customer support and success

A quarter of support tickets clustered into defects

For: Head of support or product operations lead reporting product problems to engineering

Tags are applied in a hurry and mean different things to different agents.

Pattern: Map, verify, reduceNeeds scale5 decisionsDesigned for 50 to 1,000 agents

Help articles re-checked when the product changes

For: Knowledge manager or support enablement lead owning the help centre

Every release quietly breaks a few help articles.

Pattern: WatchtowerNeeds a connector5 decisionsDesigned for 10 to 300 agents