---
name: proactive-support
description: Reach customers before they contact you, when telemetry or a known issue means they are about to have a problem. Use when failures are detectable before the customer notices or writes in.
---

# Proactive support

Contacting a customer about a problem they have not reported yet
converts a future angry ticket into a moment of trust. It requires
knowing who is affected and saying something useful rather than merely
apologising for an inconvenience.

## Method

1. **Detect the conditions that predict contact.** Failed payments,
   sync errors, repeated failed actions, and expiring credentials all
   precede a ticket reliably.
2. **Identify the affected precisely.** Messaging everyone about a
   problem affecting some is worse than silence, since it creates
   concern and support load (see notification-strategy).
3. **Lead with what they should do.** The action, then the explanation.
   A proactive message that only informs is an interruption.
4. **Be honest about cause and scope.** Under-describing an issue to
   limit alarm destroys credibility when the full extent emerges (see
   agent-crisis-comms).
5. **Time it before impact where possible.** An expiring card warned a
   week ahead is a non-event; the same message after the failed charge
   is damage control (see failed-payment-recovery).
6. **Give a path to a human.** Proactive contact generates questions,
   and dead-ending them undoes the benefit.
7. **Measure prevented contacts and sentiment.** The value is tickets
   that never arrived, which needs a comparison rather than a raw count.

## Boundaries

Proactive contact uses behavioural data, which needs a privacy basis and
should not feel like surveillance (see data-minimization). False
positives erode trust faster than reactive support does. It cannot
substitute for fixing the underlying failure.