Live · Free
Map the org the way it actually runs. The deliverable is a diff, not a chart.
A guided session that maps your organization the way it actually runs — not the chart you'd show an investor, but the one where you can name whose yes actually moved the last real decision. You inventory the recurring decisions first, then run a two-pass de-facto probe on each to surface every place the chart and the running org diverge, derive your real seats as decision-clusters, and split each seat into throughput, judgment, and trust to read where accountability moves once AI absorbs the throughput. You leave with a Decision Authority Map, a derived Role Inventory, an Accountability Lattice, and a Structural Gaps list — and a clear read on which decisions are convertible to AI today, which are judgment seats, and which are blocked seams. Zero divergence between chart and reality is a failed session — it means you handed over the fiction.
The arc
A short calibration read — your role, the one org or unit in scope (not the holding company), and how decisions actually move in a normal week. Detects the solo signal that changes how the next phases run.
Before any roles, enumerate the recurring decisions across five fixed categories — hiring, spend, product, customer, strategic — each with a real, nameable recent instance. You can't draw an aspirational decision the way you can draw an aspirational org chart.
The longest phase — the diff is born here. For every decision, run the two-pass de-facto probe and the three tells, set the divergence flag, and state the AI-Conversion Read out loud. Never accept the nominal answer alone.
What you'll walk away with
Not the chart you'd show an investor — the one where you can name whose yes moved the last real decision, plus the gaps between how it runs and how it should. Yours to act on.
Also from Framework Friday
NOW name roles — as decision-clusters derived from the map, not titles. Split each seat into throughput, judgment, and trust, then run the accountability-movement test: if AI absorbs the throughput, does the seat survive thinner or dissolve?
Resolve every outcome to one accountable seat or log a named gap. Adversarially gatekeep every 'shared' — HONEST-SHARED only if it carries a named tie-breaker and both-judgments evidence; otherwise it's DIFFUSE, a structural gap.
No score — a binary readiness check. Run the five must-haves out loud and read the diff verdict: zero divergence is a fail, every convertible read carries a rule, and an org that can't say who decides isn't ready to point AI at a workflow.
Hand over the Divergence Map + Conversion Read. Re-confirm the five screens and the diff verdict, state the paired-axis cross-check once, and name the FridayOS bridge as completion, not handoff.