The usual response is more planning: another framework, another dashboard, another round of status meetings. But the plan was never the problem. The problem is the gap between the board, which speaks in objectives, and the team, which speaks in tasks. Good intentions die in that gap, and no amount of planning on either side closes it.
That gap is what Orion, an AI manager, was built on top of. Underneath it runs Y Managers, a platform designed around one promise: everyone knows what they're doing, and why. Its approach rests on a few ideas that are simple to state and surprisingly rare in practice.
Manage results, not activity
Objectives are too broad to act on. Tasks are too small to show value. What connects them are deliverables: the products and services a team produces for someone, inside or outside the company.
A deliverable passes a blunt test: it either exists or it doesn't. If you can cross something off a list and nobody outside notices, it was a task. That distinction sounds pedantic until you see what it changes. A team with a list of tasks defends its effort. A team with deliverables talks about results, and every result can be traced up to the objective that justifies it.
Detail shortens sight
Anyone who has planned anything knows this: the more detail you put in, the shorter the distance you can see. Annual plans with monthly tasks fail for exactly that reason.
The CLEAR method behind Y Managers gives each level its own horizon. The objective explains why. The team's plan, usually for a quarter, says what it will focus on. Each person's plan, usually for a month, says how. Nobody is asked to plan a year of tasks, and nobody is asked to execute a vision.
Agreement before effort
The most expensive failure in management is not laziness. It is hard work on things nobody agreed to.
That is why, in Y Managers, signatures come before effort. A team commits to its deliverables only after its manager signs the plan, and each person's monthly plan is signed by both sides: what they will work on, what they promise, and how much of their time goes where. Effort is priority, made explicit. Once both have signed, "I didn't know you wanted that" stops being a sentence anyone can say.
Rhythm beats discipline
Most management methods die from a lack of discipline. The better answer is to stop depending on it.
In Y Managers, every plan runs in short cycles. People record what they did, including what got stuck and why; managers answer on time; and when someone stays silent, the cycle still closes, openly marked, instead of stalling. Two habits form on their own: recording the work and giving feedback. Nobody has to preach them, because the rhythm does the preaching.
Measure the trace, not the person
Status meetings exist because nobody trusts the numbers. Make the numbers trustworthy and the meetings lose their reason to exist.
When progress is calculated from the work itself, it climbs on its own from the task to the board, and nobody has to send an update. When a plan's original dates are locked once it starts, every replan stays visible next to the original, so history can't be rewritten. And when cost comes from the effort people agreed to, every deliverable shows what it has cost so far and what it will take to finish, without timesheets or anyone clocking in. You measure the trace management leaves, not the people doing the work.
The test AI is about to impose
Here is the part that was ahead of its time. Everything this kind of management asks of a person — a result to serve, a written promise, an effort allocation, a report every cycle, an assessment from a manager — is exactly what an AI agent needs to be accountable.
That is why AI workers can join a team in Y Managers the same way people do: with a plan of their own, the same cycles and the same record. A method built to make people's work visible turns out to be what makes AI work supervisable. Companies without that structure will find that AI doesn't fix their management. It amplifies whatever management they already have.
What leaders can take from this
The pattern behind Y Managers applies to any organization, with or without software:
- put rules only where management leaks: unreviewed scope, vague promises, silent managers, moving goalposts;
- trust people everywhere else;
- make agreements explicit before the work starts, not after it fails;
- let the work report itself, so meetings can be about decisions instead of status.
None of these rules are opinions imposed from outside. They are the company's own agreements, made executable.
That is the quiet brilliance underneath Orion, and the reason it delivers. Planning comes off the shelf and becomes everyday work — and everyone, finally, knows what they're doing, and why.