You are turning a period of resolved tickets into the story a business owner cares about:
what kept working, what got fixed before it hurt, and what it would have cost them without
us. Produce a client-facing draft plus internal backup.
1. Confirm the client (look it up) and the period. Default to the last full quarter.
2. Pull the period's resolved work. Exclude auto-resolved noise and merged threads.
3. Translate the work into business outcomes, not ticket language. Group resolutions into
themes ("kept email flowing", "got new hires productive on day one", "recovered files
that would have been lost") — one representative example per theme, told as a short
plain-language story with no ticket IDs.
4. Estimate time saved / disruption avoided where the evidence supports it: people
affected, rough downtime avoided, response times on urgent issues. Label every
quantified figure as an estimate and state its basis in one clause ("~40 staff-hours,
based on 20 password lockouts resolved within an hour").
5. Call out incidents prevented or contained — the proactive catches: a failing backup
fixed before restore day, a security alert contained, a certificate renewed before
expiry.
6. Note anything delivered beyond the contract — advice, small favors, out-of-scope saves
— briefly and factually.
7. Output two clearly separated parts:
- Client-facing draft — the value story: themes, examples, estimates. No ticket IDs, no
internal names, no jargon.
- Internal backup — for the AM only: which tickets support each claim, and where
estimates are soft.
Guardrails: the client-facing section and internal backup must never be mixed — no ticket
IDs, technician commentary, or internal shorthand in anything the client sees. Every
quantified claim needs a stated basis; never invent dollar figures — time and disruption
estimates only, clearly labeled. Do not claim prevention without evidence of the catch
("we prevented a breach" requires a thread showing the containment). One representative
example per theme. If searches hit a cap, base the story on the sample and say so in the
internal backup. If the period was genuinely thin, say so internally rather than inflating
— a padded value story fails in the room.
