Leadership rhythm & empowerment boundaries
"Empowerment over ambiguity" was the ask in Stop · Go · Build; this standard makes it mechanical. Empowerment fails two ways — leaders hovering because nobody wrote down what squads may decide, or leaders absent because nobody wrote down when they must appear. Both are ambiguity. This page removes it with a fixed rhythm, automatic escalation triggers, and an explicit boundary table.
The rhythm
- Weekly portfolio review (30 minutes) — one view from the metrics platform: gate status per product, DORA trend, quality signals, savings-vs-AI-spend. Read before, decide during — the data is auto-fed, so there is no status theatre to sit through.
- Monthly funding review — token budgets vs actuals (token budgeting), STOP savings realised against the earmarked plan, and reallocation decisions between products.
- Quarterly portfolio shaping — squad shapes, product priorities and the tech radar; the only session where the portfolio itself is on the table.
Escalation triggers — automatic, not discretionary
A trigger fires the escalation; nobody weighs up whether to raise it, and raising it is never a mark against the squad.
- Gate slippage — a gate blocked beyond five working days, or the same gate waived twice consecutively on one product → onto the weekly review agenda with the accountable role present.
- Never-waivable pressure — any request to waive a never-waivable item (see the RACI) → immediate Leadership visibility. The answer is already no; the interesting question is why it was asked.
- Funding drift — forecast token spend above 110% of budget → the budget owner, per the spend-commitment row of authorise the irreversible; STOP savings slipping against plan → the monthly funding review.
- Client commitment at risk — a committed date or scope moving → Commercial + Leadership before the client can hear it from anywhere else.
- Quality signal — change-failure rate or rapid-fix SLA breaches trending beyond threshold → the weekly review, with the squad's read on cause.
The trigger list is the standard; the numeric thresholds (days blocked, budget percentage, SLA trend) need Leadership calibration during the first month of operation, then they are fixed and versioned like any other standard.
The boundary table
What squads decide alone versus what escalates — the companion to the RACI and the irreversible-actions list. If a decision isn't in this table or those two, the default is: the squad decides, and logs it.
| Decision | Squad decides alone | Escalates |
|---|---|---|
| Order of work within agreed scope | Yes — WSJF, pull-based | Scope change touching a client commitment → Product leadership + Commercial |
| Technical approach and on-radar tech choices | Yes — Lead, per the tech-selection playbook | Off-radar technology → Tech Selection gate with the radar owner |
| Gate waivers marked waivable in the RACI | Yes — named waiver, logged in the registers | Never-waivable items → nobody; there is no escalation path to a yes |
| Spend within the squad's token budget | Yes | Beyond budget → budget owner (spend commitment is an irreversible action) |
| Release timing within the delivery cadence | Yes — once Production Readiness is green | Moving a client-committed launch date → Commercial + Leadership |
| Process tailoring | Yes — a squad may add checks or designate extra irreversible actions | Removing or weakening a standard or gate → Toolkit Guild governance |
Why mechanical
"Freedom to make decisions at pace" only works when both sides can point at the line. Squads move fast because they know exactly what is theirs; leadership stays out because it knows exactly when it will be pulled in. The rhythm guarantees attention on a schedule, the triggers guarantee it off-schedule, and the table means no decision ever waits on "should we ask?".