The reason things get lost is rarely forgetfulness in the moment. It is that the task arrived somewhere that is not the list: mentioned in a phone call, buried in an email thread, thought of while driving, or agreed in a message. Each of those locations is a place where an obligation can sit indefinitely without anybody noticing, and the fix is capture rather than memory.
Make capture effortless or it will not happen. A note application on your phone that opens in two seconds beats a well structured system requiring a laptop, because the moment a task appears is rarely a moment you are at your desk. A list of forty rough entries is considerably more valuable than a beautifully organised one containing six, and tidying can happen later.
Then process rather than only capture. Once a day, move everything from wherever it landed into the single list, and while doing it decide what each item actually is. Many turn out to be two minute tasks better done immediately, several are not tasks at all, and a few are projects disguised as single items, which is why they never get started.
The item that has been on your list for three weeks is telling you something. Usually it is not one action but several, and the first is undefined. Rewriting it as the specific next physical step, such as email the supplier about the delivery date rather than sort out delivery, is frequently all it needs. The second most common reason is that it does not actually matter, and deleting it is a legitimate outcome.
Review weekly at a fixed time rather than continuously. Fifteen minutes scanning for anything without a next action, anything with a date that has passed, and anything you have been avoiding. That review is what catches the work you were unknowingly abandoning, and it is the habit that separates a list that works from one that becomes a graveyard.
Keep commitments to other people visible separately, because those are the ones with real consequences. Something you promised a customer, a supplier, or an employee carries a cost if dropped that an internal task does not, and mixing them into one undifferentiated list means the important ones compete with intentions.
Resist elaborate systems, which are the most common failure mode among people trying to solve this. Tags, projects, priorities, and contexts feel like progress and add friction to the two operations that matter, which are adding something quickly and looking at the list. Nearly every abandoned productivity system was abandoned because maintaining it became a task.
Then accept that the list will never be empty, and that this is normal rather than a failure. The purpose is not completion. It is that nothing is lost and that you can see what is there, which is what allows you to decide deliberately rather than reacting to whatever surfaced most recently.
Handle anything under two minutes immediately rather than capturing it, because the overhead of writing it down, reading it later, and deciding on it again exceeds the task itself. This single rule removes a surprising proportion of a typical list, and it keeps the list populated with things that genuinely need thinking about rather than with items you could have finished while recording them.
Review the list somewhere other than your desk if you can, since the physical change makes it easier to delete things rather than deferring them again.