Week 9 of the quarter. The Rock owner says “on track,” and it’s true. They’re confident, they have reason to be, and nothing in the room is wrong.
At the quarterly, the Rock gets marked done and someone asks what changed. The answer is surprising because what got built isn’t quite the thing written down in week 1. It’s defensible. Three decisions along the way moved it, and each one was correct when it was made.
The definition had no way to keep up with reality.
A Rock is defined once and the work runs for 89 more days
EOS defines a Rock carefully, and the discipline behind it is real and helpful. Three to seven per person. SMART. One owner. Ninety days. Reported every week in two words. That definition does the job it was built for, which is deciding what matters this quarter and putting one name against it.
It gets written on one day. Then the work runs for 89 more.
Inside those 89 days the Rock changes. A dependency slips and the scope narrows. The team tries the first approach, it doesn’t work, so they go with a second. A customer commitment turns a launch into a pilot with two accounts. Every one of those adjustments is a decision made by capable people with good reasons.
None of them attach to the Rock. They live in a Teams thread, a hallway conversation, a call nobody recorded, and the owner’s memory.
What the weekly report is built to do
The Level 10 asks for on track or off track. Off track drops to the issues list and gets solved. That process works, and it is one of the better things EOS does.
The resolution lands on the issues list, which is a record of issues. The Rock’s definition stays as it was written in week 1.
So by week 9 the Rock and the work are three decisions apart, and everyone close to it is fine, because they were all in the conversations. The distance shows up for the people who weren’t: the new hire who joins in week 6, the leader reviewing at the quarterly, the person who inherits the Rock when the owner moves into a new role in October.
Next quarter’s Rock gets written from memory
For most companies the second quarter’s Rock continues the first one. It gets written in the next planning session, by the same people, from what they remember.
The three decisions that shaped the first 90 days are in that memory or they aren’t. Usually there’s at least a partial memory of what occurred along the way. Then someone reopens a question in good faith, because nothing in the room says it was decided, and the quarter spends its first two weeks re-deciding what the last quarter already handled.
the quarter spends its first two weeks re-deciding what the last quarter already handled.
Three lines under the Rock
The gap is small and so is the fix. It needs no new meeting, no new tool, and no change to how Rocks get set.
Put the decision where the Rock already lives. Ninety, the shared doc, the board, whatever holds your Rocks today. One section under each one. If it needs its own home, people will skip it.
Write the decision the day it’s made, in one line, with the reason. “Mar 14: pilot with 2 accounts instead of full launch, because the data migration slipped 5 weeks.” Date, decision, why. Thirty seconds.
Whoever makes the call writes the decision. Not the Rock owner catching up on Friday, not a scribe taking notes in the L10. The person who made the decision, because they are the only one who has the reason.
Six to eight lines by the end of a quarter is typical. That is the whole practice.
At the quarterly, the Rock owner reads that list instead of reconstructing the quarter from memory, and “what changed” has an answer that takes 40 seconds. The next Rock gets written from what happened.
A Rock names what the quarter is for. The decisions made inside it are what the quarter produced, and right now they get lost and don’t inform future work.
If your Rocks keep finishing somewhere other than where you expect, the diagnostic takes 10 minutes and shows you where the record is breaking: https://amykennedyleadership.com/choose-your-diagnostic/