Why Projects Stall at Ninety Percent, and How to Finish Them

Somewhere in your home there is a stool that needs one coat of varnish, and it has needed that coat for over a year. You own the varnish. It is under the sink, behind the thing you bought to clean the oven, which has also not happened.

Ask most people why they don’t finish things and you get a speech about discipline. No follow-through. A starter, not a finisher. The attention span of a golden retriever at a tennis ball factory. It all feels true and it is all useless, because none of it says what to do on a Tuesday evening.

Go and look at the graveyard instead. It is more specific than the speech.

The shelf of nearly

A typical shelf: the stool. A website that is nine tenths built. A short story that needs one scene rewritten and the file names sorted out. A photo book of a trip, fully laid out and never ordered, because ordering it means choosing a paper stock.

A spice rack, built and never labeled, so it is just eighteen identical jars of brown.

Notice what is missing from that list: anything abandoned at the start. This is not a pile of projects that died on day two. These died on day forty, after the hard thinking was done, after the problem was solved, after the interesting part had been fully mined.

Because think about what the last ten percent contains. Varnish. Proofreading. Renaming forty files so they match. Writing the paragraph you already know how to write. Choosing between two shades of nearly identical gray. Sending the thing to the person. Tidying up.

There is no discovery in any of it. Nothing surprises you. Meanwhile a brand new idea, shiny and entirely theoretical, is sitting right there, offering a full paycheck for doing nothing but imagining it.

So you never decide to stop. You just get outbid.

Finish lines pull, but only the ones you can see

The odd part is that the pull is measurable. Research on rewards suggests people tend to speed up as a goal gets close. In a 2006 study of reward programs, customers with a café loyalty card bought coffee more often the closer they got to the free one. People handed a 12-stamp card with 2 stamps already filled in completed their 10 purchases faster than people handed a plain 10-stamp card. Same distance, more visible progress, more speed.

That is coffee, not carpentry, and the jump to hobby projects is an inference. But it suggests what the stool is missing. A loyalty card has a last box. “Finish the stool” has no boxes at all, only a mood. Nothing is pulling, because there is nothing to see.

So build the card, starting with the last box. Before starting anything, or right now for the thing on the shelf, write one sentence defining done. Dull, literal, checkable.

“Done means the stool is sanded, varnished, dry, and standing in the kitchen.”

“Done means the story is 4,000 words, proofread once, and emailed to one reader.”

If the sentence will not come, that is information. Usually the project has no end, only a feeling, and the honest move is to give it an invented end or admit it is a hobby and let it be one. Some things are for the doing. It is pretending they are projects that produces the guilt.

Six dull tasks, each with a time and a place

Then write the tail as its own list. Tiny, stupid, specific. For the photo book:

• Pick the paper stock (matte, don’t deliberate, matte)
• Fix the two blurry photos on pages 9 and 14
• Check the spelling on the cover
• Put in the shipping address
• Order it
• Move the file to the archive folder

A list like that takes a few minutes to write, and the chores on it are each smaller than the dread that was covering them. It turns a vague, oppressive feeling called finishing into six things you can do while half watching something. Specificity is the difference between a task and a dread, which is also why rewriting a to-do list feels productive and isn’t.

One upgrade is worth the extra thirty seconds: give each line a when and a where. “Tuesday after dinner, at the laptop: paper stock and blurry photos.” A meta-analysis of 94 tests of this kind of plan, the kind that spells out in advance when, where and how you will act, found a medium-to-large positive effect on whether people reached their goals. Among other things, the plans helped people get started and shielded work in progress from unwanted influences. A shiny new idea is exactly that kind of influence.

Tell one person, once

The last piece is a date, given to exactly one person.

No public announcement, no group-chat pledge with a forfeit attached. Just: “The story is coming to you on the 20th.” Now abandoning it is no longer a silent drift. It is a sentence you would have to say to someone.

There is decent evidence for getting the finish line out of your head. A meta-analysis of 138 studies with 19,951 participants found that prompting people to monitor their progress helped them reach their goals, and that the effect was larger when the progress was reported or made public, and when it was physically recorded. The summary does not say what kinds of goals those studies covered, and it is a safe bet that none involved varnish. Still, the direction is clear enough to act on: tick the six boxes on paper, and let one person know the date.

They do not need to chase you. The date exists outside your head, and that turns out to be most of the job.

So go and find the oldest ninety-percent thing you own, the one that gives a pinch of guilt every time that drawer opens. Write the done sentence. Write the six dull tasks, each with its Tuesday. Then finish it badly this week, on purpose, at a standard you would be slightly embarrassed to defend.

A streaky stool you can sit on beats a flawless one you are still going to varnish someday.