EOS isn't working: ten patterns and the fix

Write down what is actually happening in one sentence. Meetings run long. Rocks miss. Nobody trusts the numbers. The room is polite and nothing real gets said. Those four look alike from the outside and have almost nothing in common underneath, and a team that skips this step usually buys more of whatever it already has too much of.

The reflex when a system stops working is to add to it: another board, another report, another standing meeting about why the standing meeting is failing. Almost every fix below removes something instead. Most of these ten patterns are caused by volume, and adding volume to a volume problem makes the next quarter worse.

Give the diagnosis a week. Sit through one leadership meeting with a notebook and record where the time went and which decisions got deferred for the second or third time. That page will point at one of the patterns below, which beats a team debate about whether EOS® is worth continuing.

Each row below pairs a symptom with the cause we find behind it most often and the first thing worth trying. Work one row at a time. Teams that try four fixes in one quarter cannot tell which one helped, and they usually abandon all four when the quarter comes in flat.

The most common version is a ninety-minute meeting with sixty minutes of reporting in it. The agenda exists to get through the numbers fast so the team can argue about the two or three things that matter. If your meeting inverts that, the fix is subtraction: shorter updates, a real issues list, and permission to skip a segment nobody needs this week.

The second version is quieter and worse. Everyone agrees, the meeting finishes early, and the real problem never reaches the table. The test is simple: when did somebody last tell the founder they were wrong in front of the leadership team, and what happened to them afterward?

The Level 10 Meeting™ is worth defending as a format because it removes the negotiation about what happens next. Same order, same time, same day. Teams that let the order drift end up back in a status meeting, because status is the comfortable thing to do with a room full of people.

Two symptoms, one root. Rocks that miss every quarter and a Scorecard nobody trusts both come from work that was assigned rather than accepted. If a leader did not say the number out loud and agree it was achievable, they will report on it rather than own it, and reporting is what you get in month three.

Start with the count. Three to five company priorities a quarter suits most leadership teams, and teams carrying ten rarely finish more, and feel worse about the quarter. Cutting the list is unpopular for one meeting and obviously right for the rest of the quarter.

Distrust in the weekly numbers almost always traces to definitions. Two people compute the same row differently, and within a month the team argues about the measurement rather than the result. One written definition per row, one name against it, and no changes until the quarter ends fixes more than a new reporting tool will.

If a leader misses commitments and nothing happens, or two departments keep working around each other, the fault is usually structural. Somebody is accountable for an outcome that also belongs to somebody else, or a seat was drawn around a person who was already there. Neither gets fixed by a conversation about commitment.

The test that exposes it takes an afternoon. For each outcome the company depends on, write the one name that owns it. Where you cannot write a name, or write two, you have found the argument the team has been having in code for the last year. That is what the Accountability Chart™ is for, and it is the reason drawing one is uncomfortable.

Structural problems also masquerade as effort problems. A leader who is good at their old job and out of their depth in the new one will work harder rather than say so, and the team reads it as a work-rate issue for months. Rating people against a written seat separates the two quickly.

This is the pattern nobody writes on the whiteboard. Direction arrives between sessions, decisions get reversed in corridors, and the leadership team learns to wait rather than commit. EOS Worldwide points at a leader who will not be held to account by their own team as the failure mode it sees most often, and nothing in the toolkit routes around it.

The mechanical part is fixable. New ideas go on the issues list like everything else, get priced by the team, and either become a priority with an owner or get dropped in ten minutes. Founders who adopt that routine tend to find it a relief.

The harder part is the reversal habit. If a decision made in the room can be undone in a hallway on Thursday, the room is theater and everybody knows it. The Integrator™ seat only works when the owner has said, in front of the team, that operational calls end elsewhere.

Bring in an outside facilitator when the problem is a person in the room rather than a process on the wall. A self-implementing team can fix agendas, counts and definitions on its own. It cannot referee a fight between two owners, tell a founder something unwelcome, or make a seat conversation happen that everyone has been avoiding for a year.

The other trigger is repetition. If you have run the same fix twice and the symptom came back both times, the thing you are fixing is not the cause. Somebody who has watched dozens of leadership teams will usually name it in one session, which beats another two quarters of experiments.

It is also fine to conclude the system is wrong for the company rather than badly run. Under about ten people the overhead does not pay. If the real problem is strategy, capital or a market that moved, no meeting cadence touches it.