For 26 years I worked inside one institution. It supplied the things that make work happen without you noticing: deadlines, review, and someone who spoke up when a task stalled. When I left and had to rebuild all of that from nothing, my hardest problem turned out not to be a shortage of ideas. It was that ideas arrived faster than I could finish anything. Many people with ADHD know this trap, and the fix that worked for me was counterintuitive: ration starting, not ideas.
My working formula is simple — Belief × Thinking × Action — and this article is about the Action term when your mind runs ahead of your hands.
How do you stay productive with ADHD when ideas outrun execution?
Stop rationing ideas and start rationing starts. Capture every idea in one unlimited inbox, but cap the tasks you have actively started at two per project. ADHD brains generate options quickly; the bottleneck is finishing, not thinking. Protecting your finishing capacity, rather than blocking new ideas, is what actually raises output.
Adult ADHD is common and under-recognised. Roughly 4.4% of U.S. adults met criteria for ADHD in the National Comorbidity Survey Replication (Kessler et al., American Journal of Psychiatry, 2006), and a 2024 CDC report estimated that about 15.5 million U.S. adults had a current ADHD diagnosis as of 2023 (CDC, MMWR, 2024). If starting is easy and finishing is hard for you, you are not lazy — you are working against a known executive-function pattern.
Why cap started tasks instead of trying to have fewer ideas?
Because idea generation is not the problem — abandoned work is. A backup scan of my own files once turned up 41 tasks abandoned mid-flight. I capped work-in-progress at two started items per project. Ideas stayed unlimited. Only starting got rationed, and the pile of half-built projects stopped growing.
Suppressing ideas fails for two reasons. First, it wastes the one thing an ADHD mind does effortlessly. Second, unfinished tasks do not go quiet — they keep intruding on attention, an effect Bluma Zeigarnik documented back in 1927. Every started-and-dropped task is a small open loop taxing your focus. Fewer open loops, not fewer ideas, is the goal.
What is a work-in-progress limit and how do you set one for ADHD?
A work-in-progress (WIP) limit is a hard cap on how many tasks you may have started but not finished at the same time. The idea comes from Kanban manufacturing. For ADHD, set it low — one or two per project. To start a third, you must first finish one or formally park it.
The cap does the deciding for you, which matters when task-switching is expensive. The American Psychological Association reports that switching between tasks can cost as much as 40% of productive time (APA, ‘Multitasking: Switching costs’, 2006). A WIP limit removes most of those switches by making ‘start another thing’ impossible until something closes.
How do you keep idea capture unlimited without creating chaos?
Separate capturing from committing. Every idea goes into one plain inbox with zero filtering, so nothing feels lost — which quiets the urge to act on it immediately. Committing is the gated step: an idea becomes started only when a WIP slot is open. Capture is free; starting costs a slot.
- Inbox: one list, no categories, no priority — just a holding pen so the idea stops circling in your head.
- Commit gate: pull from the inbox only when you have fewer than two started items in that project.
- No guilt about inbox size: a 300-item inbox is fine. A 20-item started list is not.
What does an open-items ledger do that a to-do list doesn’t?
It records why each item is stuck, not just that it exists. I keep one ledger as the single source of truth for stalled work, tagged by what each item is blocked on: a decision, another person, an external reply, or me. The tag tells you what action will actually unblock it.
A normal to-do list flattens everything into ‘not done’. The ledger separates a task that needs two hours of focus from one that needs a single yes/no answer. Those are completely different kinds of work, and treating them the same is why lists rot.
Comparison: endless to-do list vs. rationed starting
| Dimension | Endless to-do list | Rationed starting (WIP cap of 2) |
|---|---|---|
| Idea capture | Mixed in with tasks | Unlimited, in a separate inbox |
| Tasks started at once | Unbounded | Two per project |
| What a stall looks like | Invisible until a backup scan finds it | Logged with a blocked-by tag |
| Finishing rate | Falls as the list grows | Protected by the cap |
| Emotional load | Guilt from a list you can’t clear | Guilt limited to two open loops |
How do you decide what to work on next?
Clear the cheapest progress first. Items blocked only on a one-line decision are the fastest wins available — answer the question and the task moves. Next, handle items you are personally blocking. Items waiting on other people or external replies need a nudge and a follow-up date, not a work session.
| Blocked by | Example | Cheapest next action | Priority |
|---|---|---|---|
| A decision | ‘Which export format?’ | Answer the one-line question | Highest — clear first |
| Yourself | Draft half written | Schedule one focus block | High |
| Another person | Waiting on a review | Send a nudge | Medium |
| An external reply | Vendor quote pending | Set a follow-up date | Low — just track |
How does writing down every failure point help you finish hard things?
It turns confusion into a list you can work through. I shipped working software without being able to read code by logging every point of failure the moment it happened — each error, dead end, and wrong assumption. The written record, not debugging skill, carried the project to done.
One detail makes or breaks this: when you file a failure note, the index line has to describe the symptom, not the topic. ‘Build fails after adding login’ is findable at the moment you need it again. ‘Auth notes’ is not. A knowledge system only compounds when you can retrieve the right note under pressure.
Why it works with ADHD:
- It externalises working memory, which is the exact resource under strain.
- It makes progress visible even when the task is not done — each logged-and-solved failure is a checkpoint.
- It stops the ‘I’ve been stuck on this forever’ feeling by showing what was actually tried.
What if a task sits untouched for weeks?
Treat abandonment as maintenance, not failure. Anything untouched for 30 days gets surfaced once a week as a discard candidate, and most of mine get discarded. A shorter active list is a feature, not a confession. The idea still lives in the inbox if it turns out to matter later.
This step protects the WIP cap from quiet erosion. Without a scheduled cleanup, started items accumulate that you will never return to but also never officially drop, and the cap becomes fiction.
How do you rebuild external structure after leaving a job?
Recreate the four things an employer supplied: a deadline, a reviewer, a record of what is stalled, and a trigger that notices stalls. The open-items ledger is the record. A weekly review is the trigger. Deadlines and review you now set deliberately, because no one else will.
If you want this loop to run without you maintaining a spreadsheet, the VisionDream app turns goals into daily habits and keeps the next action in front of you. The thinking behind it is in the book Why Do I Fail While Others Succeed?, which is where the Belief × Thinking × Action formula comes from.
The short version
Ideas are not the constraint. Finishing is. Keep capture unlimited so nothing feels lost, cap started work at two so finishing capacity is protected, tag stalled items by what blocks them so you always know the cheapest next move, and write down every failure as it happens so confusion becomes a checklist. None of this requires a better brain — only a container the brain can lean on.
This is a personal system, not medical advice. If ADHD is affecting your work and life, a qualified clinician can help with diagnosis and treatment.