Turning a vision into something a team can actually do on Tuesday
Project management is not the ceremony. It is the translation. A founder says a sentence about where the business should be in a year, and somebody has to convert that into work that a specific person can finish this week without asking what it means. These four demos are that translation, running: the ladder from vision down to a task, the board it lands on, a meeting turned into work in front of you, and the capacity check that decides who gets it. Invented client, invented team, roles instead of names.
A vision is not a plan. It is a sentence that has to survive being cut into pieces.
Most operations do not fail at the vision and they do not fail at the task. They fail in the gap between, where a twelve-month ambition never gets small enough for anyone to start on a Tuesday. Five levels close that gap. Every level down adds an owner, a date and a definition of done, and nothing is allowed to exist at one level without a parent at the level above it. Click any level to see it decompose.
Which framework, and why not just one
Hybrid by default. Forcing a single methodology onto every workstream is how teams end up doing ceremony that does not fit the work. The framework is chosen per workstream, not per company.
| Workstream | Framework | Why that one |
|---|
And the loop back up
The ladder runs downward, but the information runs up. Every two weeks the retrospective asks one question that most retros skip: not did we finish the sprint, but did the sprint move the rock. A sprint that closed every task and moved nothing is a planning failure, not a delivery success.
Structure first, then the views that sit on top of it.
Spaces by function, folders by client or program, lists by workstream. Predictable enough that a new hire guesses correctly on their first day. The statuses describe reality, including the unflattering ones, because a board that cannot say waiting on client will say in progress instead, and then it is lying to you.
The layout
The three filters I run on a Monday
Three saved filters find almost every problem in a delivery operation before it reaches a client. They are also the three things a founder can check themselves in ninety seconds, which is the point. Click one.
| Task | List | Owner | Due | Status |
|---|
A meeting is not notes. It is four different things wearing the same coat.
Every transcript contains actions, decisions, risks and information, and they need different homes. Actions become tasks. Decisions get recorded so nobody relitigates them. Risks get an owner before they become issues. Information gets filed. Miss the sort and you end up with a page of bullets nobody can act on. The transcript on the left is invented. Press the button and watch it get taken apart.
Raw transcript
What comes out
Nothing yet. The sort runs first, then the tasks.
Before anyone gets the next piece of work, look at what they already owe.
Capacity is not a feeling and it is not a headcount. It is open tasks against the hours those tasks actually need, held against the hours that person has. The two failure modes look identical on a status call: the person who is drowning says they are fine, and the person who is idle says they are busy. The board settles it.
Over capacity: nothing new goes to them this week. Something already assigned moves, or a date moves, and the client hears about it from me on Monday rather than from a missed deadline on Friday.
Under capacity with everything on track: that is where the next piece of work goes, and it is also worth asking what they are not being asked for.
Under capacity with things overdue: that is not a capacity problem. That is a blocker nobody has said out loud, and it is the single most useful thing this view surfaces.