In October 2007, two researchers at the Vrije Universiteit Brussel searched the academic literature for papers citing David Allen's Getting Things Done. They found 14 documents, none discussing the method in any detail. The following year Francis Heylighen and Clément Vidal published the first serious treatment of it in Long Range Planning — and opened by admitting there were no empirical studies confirming what GTD promises.
Eighteen years later, little has changed. GTD is the most influential productivity system ever written, one of the least tested, and the most abandoned.
I use parts of it every day. I will say which, and why I threw out the rest.
The five steps, with no mystique
The official site describes five steps, resting on two rules: nothing important lives in your head, and nothing enters a list without a defined next physical action. Capture: whatever has your attention leaves your head and goes into an inbox. Clarify: is it actionable? If so, what is the next physical action; if not, it is trash, reference or waiting. Organise: the result goes to next actions, calendar, projects or someday. Reflect: you review the system frequently — in practice, the weekly review. Engage: you act from the system, choosing by context, time and energy.
What the evidence actually supports
Heylighen and Vidal did not test GTD. They showed something else: that it is consistent with what is known about situated and distributed cognition. The brain uses the environment as external memory, as a trigger for action and as a source of feedback, and GTD is a protocol for building that environment — within Miller's classic limit of about seven items in working memory at once.
The tested part comes from elsewhere. In 2011, Masicampo and Baumeister published six experiments on unfulfilled goals in the Journal of Personality and Social Psychology: people carrying a pending task had intrusive thoughts during a reading task unrelated to it. The finding that matters for GTD came next: writing a concrete plan — how, when and where — removed the effect, without the task being done. Relief does not come from doing. It comes from deciding.
Now the part the method's defenders leave out. The "open loop" idea comes from the Zeigarnik effect, and a 2025 meta-analysis in Humanities and Social Sciences Communications pooled 59 publications on it: the memory advantage for interrupted tasks does not appear, with a recall ratio of 0.99. What survived was the neighbouring finding, the Ovsiankina effect — 67% of people spontaneously return to an interrupted task.
The two gears that demand exactly what is missing
GTD has five steps but runs on two. Without constant capture, the inbox lies. Without periodic review, the lists go stale — and Heylighen and Vidal are explicit about why the review exists: to build and maintain genuine trust in the system. Those are the two gears most expensive in executive function, and GTD is recommended precisely to the people with the least of it.
Capture requires noticing. It works only if you register the actionable thought, interrupt what you are doing and write it down in seconds. In a scattered head, the idea is gone before the "notice" step: the item was not forgotten, it never reached the inbox.
Clarifying punishes whoever captured well. Every item becomes a micro-decision, and a full Friday inbox stalls on the third line.
The weekly review has no owner. One to two hours with no external deadline, nobody waiting and no consequence if you skip it: an exact description of the task an overloaded executive function will not start. Skipping one week costs nothing; skipping three destroys trust in the lists, and a system you do not trust is abandoned in silence.
There is no public data on GTD's abandonment rate; I looked and found none. The failure pattern is predictable, though: nobody quits at the capture step, they quit at the review.
The stripped-down version that survives
I kept four ideas and threw out the rest — contexts, someday lists, the incubator.
What even the stripped-down version leaves
GTD organises; it does not execute. The list can be immaculate and still sit there, because nothing in it comes to you. "Engage" is the only one of the five steps the method does not instrument: it depends on you opening the right list at the right moment — and that is exactly the moment you are busy with something else.
Why this matters for what I built
I ran full GTD for two years and abandoned it three times, always at the weekly review. What I was missing was not a better list app: it was someone on the other side to take a capture in one sentence and come back to ask what was left.
That is what I put into ASB: the inbox is the conversation you already have open, and the review does not wait for Sunday — it comes looking for you.