Escalation Triage Protocol (Delivery)

Escalation Triage Protocol (Delivery)

1. 01

   Define the escalation tiers in writing before monitoring goes live

   Agree on three tiers with the client: routine mention logged weekly, reputational risk answered within 4 business hours, and crisis routed to a named client sponsor within 60 minutes. Tier definitions belong in the retainer scope, not in a shared doc nobody opens.

2. 02

   Set the alert threshold so the inbox stays actionable

   Cap automated alerts at the volume one analyst can read in 20 minutes per shift. A platform watching 25 million sources will surface thousands of mentions a week; unfiltered alerts train the team to ignore all of them.

3. 03

   Route each alert to a single named owner, never a shared queue

   Assign one delivery lead per client account and one backup. Shared queues stall on Fridays and during holiday weeks, which is exactly when a negative thread compounds.

4. 04

   Draft the response options before the client asks for them

   Prepare three pre-approved responses per tier: no action with a logged rationale, a public reply template, and an internal briefing note. Response planning is the deliverable clients pay for, not the alert itself.

5. 05

   Log every escalation with timestamp, tier, action taken, and outcome

   A running log turns a chaotic month into evidence at the quarterly review. It also exposes whether the query set is catching noise or missing the conversations that actually moved.

6. 06

   Run a monthly calibration on tier accuracy

   Pull five escalations at random and ask whether each was over- or under-classified. Adjust thresholds and query terms from that sample rather than from gut feel.

7. 07

   Report escalation volume and response time in the client review

   Show counts by tier, median time to first response, and any tier-3 event with the resolution. This is the number that justifies renewing a monitoring retainer when the client questions the line item.