Control Tower: the panel I built because fifteen open fronts isn't a discipline problem
Discipline wasn't the problem — there was no single place where fifteen open fronts showed up at once. How I adapted classic GTD with two blocks it doesn't account for, and why.
Every time a task falls through, the obvious explanation is a discipline problem. It almost never is. In most cases, what slips away is whatever never had a written next action, or never had a date attached to chase it.
Running several fronts at once — business, product, infrastructure, plus whatever personal stuff can’t wait — I got to a point where the list of open projects passed ten and no single place showed all of them at the same time. The feeling of drowning didn’t come from volume. It came from not having one place to look.
I fixed it with GTD (Getting Things Done), but not the textbook version. Two pieces the classic method doesn’t account for, because the real bottleneck wasn’t task volume — it was waiting on someone else and pending decisions.
Three leaks explain almost everything that disappears
A project with no next action. A project name on a list isn’t a task — it’s a label. As long as the line says the project’s name instead of the next physical, concrete move, nothing moves. “Sort out client X” isn’t an action; “send the email asking Y” is.
Delegated with no follow-up date. Handing work off gets it out of your head, but if nobody marked the day to ask “how’s it going?”, it dies quietly. This is the leak that kills the most projects — not because the person failed, but because nobody defined when to chase it.
A decision with no deadline. A decision only you can make blocks an entire team while it sits loose in your head. It needs a scheduled slot, like a meeting — otherwise it competes with everything else urgent that day, and always loses.
Five lists, and only five
Pure GTD is exactly this, and the temptation to add more lists is the most common way to kill the system by week two:
- Inbox — everything that arrives, unprocessed. Emptied once a day, never worked directly from inside it.
- Projects — anything with more than one step. Every line requires a next action, no exceptions.
- Next actions — grouped by context (decide, write, call…), not by project. This is the list you open to work.
- Waiting on — delegated to someone, with an owner and a follow-up date. The single most valuable list in the whole system.
- Someday / maybe — alive, but out of this week’s focus. Out of mind without being out of the system.
Five steps, each with a fixed time
Capture, clarify, organize, reflect, engage — GTD’s sequence works because each step only exists if the one before it actually happened. That’s why each one got a fixed time slot instead of happening “whenever”:
- Capture — all day, zero cost. Anything that comes up becomes a loose sentence, no organizing or judging yet.
- Clarify — ten minutes, right after the focus block. Actionable in under two minutes? Do it now. Anything more becomes a next action, a waiting-on item, or someday/maybe.
- Organize — automatic, goes straight from triage into one of the five lists, with context and, if delegated, a follow-up date.
- Reflect — the weekly review. The one non-negotiable step — skip it and the system turns into a graveyard of lists in three weeks.
- Engage — a protected block, chosen by context and available time, not by whoever shouted loudest.
The two blocks classic GTD doesn’t account for
In the textbook version, chasing people and making decisions both live inside the weekly review. That wasn’t enough for my case, and probably isn’t enough for yours if the bottleneck looks similar: a week’s delay chasing someone costs more than ten undone tasks.
So the standard week picked up two extra blocks, with a fixed time like any meeting:
| When | Block | For |
|---|---|---|
| Mon–Fri, 7–9am | Focus block | One front only, no meetings |
| Mon–Fri, 9:00am | Triage | Empty the inbox, pick the day’s priorities |
| Tuesday, 2pm | Chasing | Walk the entire Waiting-on list, line by line |
| Wednesday, 3pm | Decision desk | Only what’s blocked waiting on a decision from you |
| Friday, 4pm | Weekly review | Walk all five lists, pick next week’s focus |
The rules that hold it up
- Five active projects, maximum. For a sixth to enter, one has to leave. It’s the limit that turns a list of fifteen into actual work.
- Every project has a physical next action. If you can’t write the next move, it’s still a decision disguised as a task.
- Delegated becomes a Waiting-on line with a date. No date means it wasn’t delegated — it was abandoned with a witness.
- Under two minutes, do it now. During triage, don’t write it down: resolve it.
- Friday’s review doesn’t get rescheduled. If one day slips, the system holds. If two weeks in a row slip, you’re back to square one.
None of this is sophisticated. What matters is that every rule exists because it answered a specific leak I’d already watched happen — not because it was in a book.
I built the panel itself together with AI, which in the end is the same subject as every technical post here: a system built for one real, defined problem, not a generic off-the-shelf tool.