An offsite for the people who can fund or stop a bet earns its place when a decision is stuck. It wastes the calendar when the room has a theme and nothing to decide. You can tell which booking you are making before anyone drafts an agenda, and the tell is the decision itself, not the quality of the venue.
Jeff Bezos set the distinction out in Amazon's 2015 letter to shareholders. Some decisions are consequential and hard to reverse. He calls them one-way doors, and says they should be made slowly, with deliberation and consultation, because you cannot walk back through. Most decisions are two-way doors. If the choice is wrong, you can reopen the door, so a small group with good judgement should make it quickly. His warning is that a large organisation starts using the heavy process on almost everything, including the reversible choices. The result he names is slowness, unexamined risk aversion, too little experimenting and less invention. A footnote notes the opposite risk: treating a one-way door as easy to reopen.
When the day is the right tool
A reversible bet that has sat in steering committees for a quarter is a two-way door being treated as a one-way door. A new offer you could withdraw, a pilot you could end, a price you could change. The day is worth running when the people who can approve a cheap test are in the room and will leave with that test written down. If they will not stay for the decision, do not book the day.
A one-way door is a different booking. A public commitment, a spend you cannot claw back, a market you will exit, a system you will turn off. Run the day when that commitment is close and the room has not written what would make it a mistake, who must be consulted, and what happens if the early evidence is bad. Slow down, name the consultation, and record the point at which you would still stop.
Skip the day when there is no live bet. A theme such as growth, culture or digital gives the group nothing to decide. Skip it when the decision is already made and the session is there to announce it. Skip it when the people who can say yes are sending delegates, because a delegate cannot close the door. What one sponsor should be able to sign is a different question. This note is about the team that holds several bets at once.
The page at the end
The output is one page. Each bet is marked reversible or not, in a sentence a director can repeat. For a reversible bet, name the cheapest test that could change the team's mind, the owner, the date and the result that ends the work. For an irreversible bet, name who must be consulted, what evidence would delay the commitment, and who may stop it if that evidence shows up in time. A bet with no owner comes off the page. Taking items off is part of the result.
Agree the marks while the people who disagree are still in the room. A page written the next morning, by the person who liked the bet, will drift back toward it. Read the page aloud before anyone leaves, including the stops. If the group cannot hear a stop said out loud, the session is not finished.
Collective Campus is building high-performance culture and teams for the AI world. The lean startup and product sprint is the room for an executive team that already has a live bet and needs a test, a stop line and a name. What one sponsor should be able to sign is in the sponsor's decision. What to put under test, once the bet is named, is in a lean startup workshop for corporate teams. A pilot with no end date is the subject of pilot purgatory.
What to bring
Bring this month's bets, and the names of the people who can fund or kill each one. If you cannot name those people, you do not need a facilitator.
