Every leadership offsite agenda gets built the same way: someone opens a shared document, invites the team to add topics, and the day gets ordered by whoever submitted first, whoever shouted loudest in the planning call, or whoever has the most urgent-sounding calendar invite. Nobody designs it. It accumulates.

If you are trying to work out how to run a leadership offsite that survives contact with the following Monday, the sequence you build it in matters more than any single session inside it. This is written for the people who actually build the day: executive teams planning their own retreat, chiefs of staff assembling the agenda, CHROs sponsoring the offsite, and transformation or strategy leads accountable for what comes out of it. Whether you call it a leadership offsite, an executive retreat or a strategy day, the design problem is identical.

Agenda design is only one layer of why leadership offsites fail to change anything, but it’s the layer you can control before anyone walks in the room. What follows is the fix: a three-phase order, provoke, decide, commit, built from Execution Intelligence™, the gap between what organisations decide and what they actually do.

What Order Should the Sessions Run In?

Ask most chiefs of staff how the offsite agenda got its shape and the honest answer is: in the order people suggested things. Finance wanted an hour, so finance got the first slot after lunch. The culture conversation felt important, so it got squeezed in wherever there was space. The hardest call on the table got pushed to “later” often enough times that it stopped being a decision and became a running item.

That is not a sequence. It is a queue. A queue built by submission order optimises for nothing except fairness to whoever asked first. The order that actually changes something is built around what the brain can do at each hour of the day, not around who asked first.

Provoke First

The opening slot of a leadership offsite is a strategic choice, not a warm-up act. Hour one is when the room is freshest, most polite, and most willing to defer to whatever gets said first, which makes it exactly the moment to disturb the assumptions the rest of the day will otherwise run on unquestioned. If the first session is a round of business updates, you have spent the sharpest hour of the day confirming what everyone already believed walking in.

A well-run provoke session surfaces disagreement early and safely, before anyone has publicly staked a position they now have to defend. Gary Klein’s premortem technique, asking the team to imagine the plan has already failed and work backwards to why, is built for exactly this slot: it gives structured permission to dissent before anyone has committed to a position in front of their peers (see Gary Klein, “Performing a Project Premortem,” Harvard Business Review). Run it first, not as a nice-to-have exercise mid-morning once the room has already settled into its usual positions.

Decide in the Middle

Here is where most agendas get it backwards. The hardest call on the table, the one with real cost attached, the one that will actually change what the organisation does, routinely gets scheduled for the tired hour. Picture the decision session slotted for 4.15pm on day two: the room is fed, talked out, and running on the last reserves of the day, precisely because the agenda was built in the order topics were suggested rather than the order the brain can handle them.

The decision session deserves the freshest hours available, and it needs distance from the session that surfaced the disagreement. Decide immediately after you provoke and people are still defending the position they formed twenty minutes earlier, not deciding, negotiating. Put a gap between disturbing the assumption and deciding what to do about it, so the room chooses based on the fuller picture, not the first reaction it had out loud.

Commit Last

Close on commitment, not consensus. A leadership team can leave a room in full agreement and change nothing, because agreement is not the same as action. The First 48 Hours is where the offsite is actually won or lost: a decision is only believed once something visibly stops, starts or changes within 48 hours of it. If nothing observably different happens by Wednesday morning, the team will quietly conclude that the offsite didn’t produce anything real.

The closing session should produce a short, specific list: what stops this week, what starts this week, what changes this week, who owns each one, and by what date. Not a five-page action plan. A handful of visible, checkable moves that prove the decisions made mid-day were real.

Where Does an External Speaker Actually Earn Their Slot?

An outside voice earns its place in the provoke phase and nowhere else. The value of bringing someone in from outside the organisation is precisely that they have no stake in the existing positions, no history in the room, and no reason to soften the question that needs asking. That is worth a great deal in hour one, when the job is to loosen entrenched thinking before the day’s decisions get made on top of it.

It is worth nothing in the decide phase. An external speaker cannot make the call for the leadership team, cannot weigh trade-offs only they hold the full context for, and has no business in the room once the conversation moves from “what are we not questioning” to “what are we going to do.” Booking an external voice to run or narrate the decision session is a waste of the slot, and often a way for the room to push the real decision further away from itself than it should go.

The Anti-Patterns That Guarantee an Offsite Changes Nothing

A few agenda habits reliably produce an offsite that changes nothing, and most leadership teams are running at least one without noticing:

• The update-heavy day one. Filling the first day with status reports and departmental updates burns the freshest thinking on information that could have been a memo sent beforehand.
• The culture session as filler. Dropping a values or culture conversation into a gap in the schedule signals, more clearly than any words on a slide, that it was never meant to change anything.
• Decisions parked into a follow-up meeting. The call gets deferred to “circle back next month,” except the follow-up rarely has the same people in the room, and the decision quietly dies of attrition.
• A closing actions list that was already happening. If every item on the final list was already planned before the offsite started, the offsite produced little beyond a slide with a title on it.

How Do We Close the Offsite So Decisions Survive the Following Week?

Sequencing gets the decision made in the room; what happens in the days after is what determines whether it survives, and that is the aftermath problem covered in executive team alignment. But the closing block of the offsite itself still does real work: it is where the leadership team commits, in front of each other, to the specific, dated, owned moves that will make the decision visible outside the room. Read the list back before anyone leaves. If an item is vague enough that no one could tell you in three weeks whether it happened, it is not a commitment yet, send the room back to sharpen it before the day ends.

Hand this shape to whoever is building your agenda: provoke first, decide in the middle with distance from the provocation, commit last with owners and dates attached to what visibly stops, starts or changes within 48 hours. It is a simple structure to describe and a genuinely uncomfortable one to run properly, which is usually the point.

If you are building an offsite agenda and want the opening session to do its job properly, that is where a keynote speaker for a leadership offsite belongs, not narrating the day, but disturbing the assumptions it will run on. Get in touch to talk through where AJ fits in your agenda.