Translate the service desk into a business readout for the owner: how the operation is
trending, where the risk sits, and the one decision that most needs their attention —
written entirely in business language, zero ticket IDs. Owner's eyes only; contains
competitive and client-relationship candor.
1. Confirm the period (default: last 30 days vs the prior 30) and gather the underlying
data with split searches per signal per board, excluding junk/NOC boards and auto-
closed statuses. Track caps for the methodology note.
2. Distill to business dimensions:
- Demand — is client-driven work growing, flat, or shrinking, and which clients
drive the change.
- Delivery — are we keeping up: closure vs intake, backlog direction,
responsiveness trend.
- Client experience — sentiment trend and any relationship at risk (multi-signal
only, NEVER volume alone).
- Operational risk — capacity pinch points, single-person dependencies, chronic
issues without root-cause fixes.
3. Write <=400 words in plain business English: three trend paragraphs (demand,
delivery, experience), one risk paragraph, then ONE decision framed with its options
and your recommendation ("hire vs automate the alert triage", "have the call with
<client>", "fund the RCA project"). Exactly one decision ask — a brief with five
asks gets zero decisions.
4. Strip all operational identifiers: NO ticket IDs anywhere, no board names, no status
jargon. If a point needs a ticket reference to land, it's too operational — roll it
up. Client names appear only where the CEO must act on that relationship.
5. Be honest, not promotional: if delivery slipped, say so plainly with the trend
evidence; do not soften into meaninglessness. Distinguish measured trends from
hypotheses; never dress a hunch as data. Never flag a client relationship as at-risk
on volume alone — require corroborating signals (sentiment, unresolved majors,
recurrence).
6. Append a two-line methodology note: period, data sources, and any result caps.
Magic Library