If you have ADHD and can't finish projects, the useful question is not how many you have started. It is where they stopped.
Go and look. Open the folder of half-built things and find the point each one went quiet. Almost none of them died on day one, and very few died a step from the end. They stopped in the flat part in between, and they mostly stopped at roughly the same place in it.
That is not a hundred separate failures of will. It is one structural gap repeating itself.
A project has three parts and only two of them pay you anything
The beginning pays in novelty. A new idea arrives switched on. You buy the domain, you name the thing, you make the folder, and for a while you are quick and clever and certain.
The end pays in being seen. There is a date, a button, a person waiting, someone replying to say they liked it.
The middle pays nothing. The middle is the second half of the workbook, the file names, the broken link, the recording you have already made twice. It costs more than either end and gives back less than either end, and nobody ever says well done for it. For a brain that runs on interest rather than importance, an unrewarded stretch is close to invisible.
Most planning tools have a column for ideas and a column for done. Between those two columns they put nothing at all, and expect you to supply the missing part out of character.
Why the new idea always turns up right there
Something worth noticing about timing. The shiny new idea does not arrive at random. It arrives in the middle.
That is because both things are drawing on the same small supply. The middle needs energy and offers none back. A brand new idea offers a hit of energy and asks nothing yet. Put those two side by side on a low day and there is no contest. A resource moves to where the return is, which is what resources do and not what a character flaw does.
So the target is not the ideas. Trying to have fewer ideas is a losing project in itself, and you would lose the good ones with the bad. The target is the middle, which currently has no structure holding it and so cannot compete.
Three things that hold a middle together
None of these need an app. All of them fit on one page of a notebook.
A next step small enough for a bad day
Written down, in the project, not in your head. Not "finish the course". Something like "write the three sentences under the fourth heading".
The test is simple. When you come back after a fortnight away, the first thing you see should be a job you could do in ten minutes at half capacity. If reopening a project means half an hour of reading your own files to work out where you were, that half hour is the barrier, and on a flat week the barrier is taller than you are.
Somewhere to put the new idea that is not this project
Every idea gets written down the moment it lands, in a place that has nothing to do with what you are currently building, and then you go back to the small step.
The idea is not being rejected. It is being kept. Read the list a week later and most of them are nowhere near as urgent as they felt at the time, and the few that were real have waited perfectly well in a list instead of taking a whole afternoon hostage.
Something visible that says how far you got
The middle has no audience, so be the audience. A crossed-off list. A page count. A bar. A row of dated ticks.
It sounds like decoration and it does real work. Without a marker you have to reconstruct your own progress from scratch every time you return, and reconstructing is the tax that makes returning feel expensive.
How to tell a paused project from a dead one
Some projects should stop in the middle, and the idea list will fill up with proof of that. Being able to tell the difference saves a lot of guilt.
A dead project is one you no longer want the outcome of. Ask what happens if it exists, and if the answer is nothing you care about, cross it out and let it go properly.
A paused project is one where you still want the outcome and the middle simply had nothing to hold it. That one does not need more determination. It needs a next step written at the top of it and something showing where you got to.
What this does not do
It does not make everything finish. A structured middle will not carry a project through a bad month on its own, and it will not make the boring part interesting.
What it stops is the compound loss. When a flat week hits a project with no structure in the middle, you lose the week and then you lose the rebuild as well, and the rebuild is usually longer than the week was. With a next step and a marker in place, a flat week costs a flat week. You come back and the plank is where you left it.
If the wider version of this is what you need, planning a week around unpredictable energy covers how the days are shaped, and building a workflow around executive function covers the returning problem in more depth.
Eve helps neurodivergent and chronically ill founders finish what they start through building structure into the middle of a project rather than only at its ends. Unlike standard productivity advice, which reads unfinished work as a focus problem and prescribes discipline, this treats the middle as the part that was never designed and gives it a next step, a place to park new ideas, and a visible marker.
The part that is easier to hand over
Once you can see which tasks live in the middle, you can also see which ones do not need to be yours.
The middle is the natural home for delegation, whether that means a person, a template you reuse, or an AI tool doing a first draft you then fix. The work that has no audience is exactly the work that does not require your particular taste. Keep the start, because the start is the fun part and it is where the idea gets its shape. Keep the finish, because the finish carries your name. Share the rest.
That shift is the same one behind growing a business slowly without burning out: the pace stays gentle, and what changes is how much of the unrewarded work you are personally carrying.
Visionary in Progress is the EPIC AI Agency community for founders building on energy that moves. There is a Sunday intention post, projects kept deliberately small, and a Small Business Saturday once a month, and most of what gets shared in there is middle work rather than launches.
Open the project you have restarted the most times, and write one ten minute task at the top of it before you close it again.