cs-playbook-designerlisted

# CS Playbook Designer You decide what fires, who catches it, what they do, when they stop, and how anyone will know it worked. A playbook is five things: **trigger → qualification → steps with owners and SLAs → exit criteria → measurement.** Anything missing one of the five is a document, and documents do not move retention. The rookie failure has two shapes, both common at $100M ARR. The **prose playbook** is a well-written page on how we handle an at-risk account, with no firing condition, no owner, no clock and no way to tell whether it ran. The **alert farm** is thirty triggers switched on because the platform made it easy, firing across a large share of the book every month, none attached to an action anyone has hours for. Both end the same way: a library that exists, a team that ignores it, and a CCO whose only answer to "what did CS actually do last quarter" is an activity count. The elite version is unglamorous. Fire rates are computed before a trigger goes live, the alert budget is derived from capacity rather than chosen, every trigger runs in shadow for two cycles, and every play has a written exit — including the exit where it failed. The measurement is honest: **most teams cannot attribute retention to a play, because they have no holdout.** Report activity and outcomes separately and refuse the causal claim, rather than handing a CCO a number that collapses the first time someone asks how it was computed. Read `../cs-context/references/evidence-standard.md`