A working session and a steering committee are both governance meetings, and treating them as interchangeable wastes both. A working session runs weekly, seats the workstream leads, and makes tactical, operational calls, is this analysis on track, does a task need re-sequencing, is there a resourcing conflict the workstream can solve itself, and can tolerate genuine problem-solving in the room, because everyone present shares the technical context. A steering committee runs monthly or at defined milestones, seats the sponsor and senior stakeholders, and makes strategic, resource-committing calls only that level has the standing to make; its participants are busier, less immersed in daily detail, and each answers to their own boss or board for whatever gets decided in that room. Confusing the two produces two opposite failures: a working session run like a SteerCo (formal, slow, ceremony instead of problem-solving) or a SteerCo run like a working session (an unstructured discussion that consumes senior time and decides nothing).
Advertisement
How do you prepare a steering committee meeting?
Agenda design, scaled to the audience, follows one principle: every item is decision-oriented ("decide whether to approve the retrofit business case," not "discuss the plant options"), time-boxed in advance, and ordered so the RAG status and the one or two real decisions get the bulk of the time, never buried after status theater the pre-read already covered. This is where the answer-first discipline CN1.03 and M10.03 teach in full becomes load-bearing at the meeting level, not just the deck level: a pre-read that opens with the governing thought and the specific ask does for a meeting what an answer-first recommendation does for a written deliverable, while a narrative pre-read asks busy stakeholders to do the synthesis themselves, live, in the room.
Pre-wiring is the discipline of never letting a room be the first place a stakeholder hears a recommendation, this module's answer to CN0.02's promise. Before a contentious decision reaches the steering committee, the program lead holds short 1:1 conversations, fifteen to twenty minutes, three or four days ahead, with whoever is most likely to object, walking them through the recommendation and genuinely listening while pushback is still cheap to hear. An objection raised privately is a data point the recommendation can absorb; the identical objection raised live, in front of peers, tends to harden into a position the stakeholder must now defend publicly. Worked Example 4 quantifies exactly what this buys a program, using the plainest measure available, how long the meeting ran, and whether its decisions needed a second meeting to close.
How do you facilitate a meeting to a decision?
Facilitating to a decision moves a stuck, opinion-driven debate to a structured, defensible choice, and the most reliable technique is to separate criteria from positions: agree explicitly, before anyone argues for an option, on the weighted dimensions that should decide the choice (cost, risk, workforce impact, whatever actually matters here), then score every option against them in the open, so any remaining disagreement is about a specific score on a specific criterion, not a clash of instincts. A parking lot, a visibly maintained list of real but off-agenda topics, explicitly promised a future home rather than silently dropped, protects the time-box without the facilitator looking dismissive. Before the meeting ends, the facilitator's last job is to say, out loud, exactly what was decided, who owns the next step, and by when, a meeting everyone nods through but no one can accurately repeat back has not decided anything. Worked Example 5 walks a stuck plant-closure-versus-retrofit debate through this exact technique and the minutes it produces.
What should the minutes and actions record?
Minutes and actions make a decision durable past the room it was made in, and they fail in one of two ways almost every time: recording what was discussed rather than decided, or logging an action with no named owner and no date, which quietly guarantees it will not happen (a program that tracked this found that of five undated, unowned items in one quarter only one closed on schedule, versus seven of eight the following quarter once every item carried a named owner and a due date; this is the difference between an action and a wish). A decision log (decision, rationale, date) and an action log (owner, due date) are worth keeping separately, because a reviewer later needs to answer two different questions, what was decided and why, and whether the follow-through is actually happening, that one blended narrative answers poorly. Circulate both within a day or two, while memory is fresh; minutes that surface a week later, once details have blurred, invite exactly the "I never agreed to that" dispute a steering committee exists to prevent.