War-game a ticket deflection goal before bands, exception paths, and clawback rules harden into how every support grant gets made.
Ticket deflection goals fail when the steps invent deflection speed the article map never funded, when grants dual-count the same contact as deflected and as resolved, when support still promises custom floors in private channels, and when support ops cannot show which exceptions actually cleared under the published path. A neat deflection flowchart is not evidence.
Lock the goal draft
State the proposed percent bands and owners, which products and regions it covers, who owns exception audit and clawback, what success looks like after two quarters of average deflection and exception rate, and the kill criteria if shadow floors or fairness disputes past a named threshold. Attach the current deflection distribution, sample grant logs, queue samples, and the measured path from contact to posted deflection. If support, product, and finance disagree on whether the goal buys self-serve or just deferred queue fights, reconcile before seating.
Name the decision you will make if the war game finds nothing new, and the hold criteria if any band still lacks a named owner or a verified audit path.
Seats that should attack it
- Angry customer. Fear that "self-serve first" still means a quiet dead end below the published ladder.
- Support director. Queue math that looks clean while exception cohorts cannot be listed.
- Product. Talk tracks that invent article coverage the knowledge base never wrote.
- Support agent. Macros that still host retired deflection language.
- Skeptic. The claim that looks strongest and is least sourced.
Optional legal seat if SLA language binds the form. Same pack for every seat.
Run the private passes
Feed Pingpong the goal draft and exhibits. First pass steelmans the deflection steps. Later passes attack from the seats above. Final pass turns surviving objections into clearer bands, a published clawback path, or a hold. Delete invented "customers already expect these floors" claims and dual-counted deflection rates.
Force a day-after narrative: what happens if a large account screenshots mismatched deflection terms, if agents still teach unwritten floors, or if finance cannot list the open exception cohort. If those stories are stronger than your mitigation plan, fix the package before you publish it. Separate a ticket deflection goal from a support SLA tier split or a support escalation ladder so each decision stays testable.
Ask every seat to mark which grants are optional in practice. Optional clawback audits become skipped within a month if nobody names the owner and the time box. Related: pretend you are the support ops lead, war-game a knowledge base rewrite, and the war-game decisions hub.