All use cases

Retain / Save

Policy-cancellation save conversation

A policyholder starts to cancel mid-term — usually with a competitor's quote already open — and the flow becomes a conversation that finds the real reason before the cover lapses.

Next best thingClose the loopBehavioral triggerRisk triggerChatVoiceEmailInsurance

What it does

When a policyholder moves to cancel mid-term, the reason is rarely "I no longer need cover" — it is a premium that jumped, a claim that left a bad taste, or a cheaper quote in another tab. Most cancellation flows never learn which, so a savable policy gets rubber-stamped and the renewal premium walks out the door. The agent turns the cancellation into a diagnostic conversation: it asks why, listens, and offers the save that actually matches — a re-rate against current factors, a coverage adjustment that fits the real need and budget, or a retention measure within policy. When the policyholder genuinely wants out, it processes a compliant mid-term cancellation with the correct pro-rata or short-rate handling. Save offers are tested, and the save rate is measured against a holdout so the number is real.

How it works

  1. 1
    Trigger. cancellation intent appears — a mid-term cancel request in chat, on a call, or by email, or churn language ("I've found cheaper cover") surfacing in conversation.
  2. 2
    Decision. the agent diagnoses the real driver from the conversation plus policy memory — premium change at last renewal, claim experience, coverage fit, a competitor quote — and selects from a policy-bounded ladder of saves (re-rate, adjust coverage, retention measure). A judge gates the exchange for tone, suitability, and regulated language.
  3. 3
    Action. the conversation runs on the policyholder's channel — chat, voice, or email — and the agent makes the matched offer, explaining the trade-off plainly (what changes in cover, what changes in premium); a re-rate or coverage change is quoted in the moment for confirmation, never silently rebound.
  4. 4
    Follow-through. the outcome and stated reason are written back for renewal and product teams; a genuine cancellation is processed with compliant pro-rata or short-rate refund handling and confirmation; anything outside the agent's authority hands to a licensed adviser, and saves are reported as lift over the holdout.
Example
AgentBefore I action the cancellation — can I ask what's prompting it? If it's something on the policy I can look at, I'd rather do that than just close the cover.
CustomerHonestly, I got a renewal quote elsewhere that's quite a bit cheaper.
AgentUnderstood. Your premium did go up this term, partly from a rating change I can walk you through. Two options: I can re-rate your current cover against today's factors, or adjust the policy to better match what you actually need — either could close most of that gap. Want me to quote both so you can compare before deciding?

Configuration

How the agent is wired for this use case.

Triggera mid-term cancellation request or cancel intent — a policy_cancel_requested event from the policy admin system, or cancellation/competitor-quote language detected in conversation.
Tools & actions
  • Policy admin system · read the policy, term dates, premium and rating history, and claim record to diagnose the driver.
  • Rating / quoting engine · produce a re-rate or coverage-adjustment quote against current factors for the policyholder to confirm.
  • Knowledge base · retrieve the regulated cancellation, cooling-off, and pro-rata / short-rate refund rules for the product and region.
  • CRM · log the stated reason, the offer made, and the outcome for renewal and product teams.
  • Messaging channel · run the diagnostic conversation in chat, on voice, or by email.
Autonomythe diagnosis, the matched save offer, and the explanation run unattended behind the judge. Any change that re-rates or rebinds a policy, moves money (pro-rata or short-rate refund), or touches a regulated suitability decision is a confirmation / handoff policy class — quoted and confirmed with the policyholder, executed against policy admin only on agreement, and routed to a licensed adviser where authority or suitability requires it.
Channelschat · voice · email
Escalationa coverage or suitability question beyond the agent's authority, a complaint or vulnerability signal, a regulated decision the agent cannot make, or a refund dispute hands off to a licensed adviser with the policy context attached.

What you need

The inputs this use case runs on. Your channels stay yours; the agent supplies the judgment.

Signals

mid-term cancel requests, cancellation and competitor-quote language in conversation, renewal-premium-change and claim-experience flags.

Data

policy terms and dates, premium and rating history, claim record, coverage and suitability context, customer memory of prior offers, consent state.

Guardrails

cancellation always stays one step away — no dark patterns, no loops; the save ladder is capped by policy and never crosses a suitability line the agent isn't authorized to decide; judge gating on every message for tone and regulated language; re-rates, rebinds, and refunds quoted and confirmed before execution, with money-moving and suitability steps held to a confirmation / handoff class; regulated cancellation, cooling-off, and refund rules applied per product and region; randomized holdout for save-rate measurement; the insurer keeps owning the send channels.

Metrics it moves

  • save-rateup, by matching the save (re-rate, coverage adjustment, retention measure) to the diagnosed reason instead of rubber-stamping the cancellation.
  • churndown, as price- and experience-driven mid-term cancellations get addressed at the moment of intent.
  • renewal-rateup across the book, and even processed cancellations leave a clean reason and a warm path back at renewal.

Related use cases

Retain / Save

Cancel-flow save conversation

the subscription and fintech parent; this carries insurance re-rating and regulated cancellation rules

Retain / Save

Policy auto-renewal and lapse save

the same save conversation anchored to the renewal window

Win back

Lapsed-policyholder win-back

picks up the relationship when the cancellation goes through

See it on your own customer journey

Bring one drop-off, one churn cliff, or one silent segment. We will show you what a proactive agent with memory and judgment does with it.

Book a demo