Finishing the Boring Tail — Why the Last 20% Is Hardest

The short answer: the last 20% has different brain-economics than the first 80%

Anders Ericsson's deliberate-practice research (source) explains part of why finishing is so much harder than starting: the late stages of any project have no novelty dopamine, the visible progress is small (everything looks 95% done forever), and the work that's left is often the parts the brain quietly avoided through the easy middle. You're not lazy at the end; you're working without the reward system that carried the start. The intervention isn't more discipline. It's external structure that compensates for the missing internal pull — scoped finishing windows, deliberate reward pairing, a public reveal date with teeth. The boring tail finishes when the system is built for it.

Why "just push through the last bit" rarely works

The last 20% looks small from outside the project but contains the highest concentration of accumulated procrastination — the small tasks you deferred during the exciting middle. Those small tasks aren't small now; they're crowded into the same window and they're the parts you weren't excited about even when motivation was high. Telling yourself to push through assumes you have what you had at the start. You don't. The willingness budget is lower; the dopamine economy is different. Either acknowledge that and engineer for it, or watch the project sit at 90% for months.

Six moves that close projects

• Set a public reveal date. Announce when the thing ships, to someone who'll notice. The external date converts the internal-only finish line into a real deadline, and the social cost of slipping is the enforcement the late stage can't generate internally.

• Break the tail into named deliveries. "Finish the project" is too big. "Send the contract draft Monday, fix the three remaining bugs Wednesday, write the launch email Friday" is shippable. Each named delivery has its own ship-or-not signal, and the small wins keep the system running.

• Reward the boring parts deliberately. Temptation-bundling logic applied at the project scale: a specific coffee shop you only go to when you're doing the boring tail, the audiobook you only listen to while doing tax-equivalent admin. The novelty doesn't have to come from inside the work if you can stack it from outside.

• Time-box the tail. Decide how many weeks the finishing phase gets. Two weeks, four weeks, six weeks. Past the box, the project closes whether it's perfect or not. Parkinson's law guarantees that work expands to the time given; bound the time and the work bounds itself.

• Identify the one part you're avoiding and do it first. There's usually a single hard task hidden in the tail — the conversation, the legal review, the rewrite — that everything else can't proceed without. Doing it first unlocks the rest; deferring it stalls everything. Spending one bad afternoon on the hard thing usually saves three frustrating weeks.

• Document the project as part of finishing. Write the short retrospective — what worked, what didn't, what you'd do differently — as part of the project's last 5%. This sounds optional and isn't: it produces the visible artefact of finishing, which the brain reads as completion. Without it, the project quietly continues in your head.

Why this pays double for ADHD

ADHD reward systems are unusually sensitive to novelty and unusually weak at the late-stage carry. The exact phase where neurotypical workers can grind through, ADHD readers reliably stall. The system-based approach reverses this: the public date provides external pressure, the named deliveries provide multiple novelty-reward moments, the temptation-bundling reintroduces dopamine from outside the work, the time-box ends the open-ended dread. None of these requires you to be a different kind of brain; they all compensate for the brain you have.

FAQ

What if the project genuinely isn't ready?

Then the time-box and the public date will force the conversation about what's worth shipping vs. what's perfectionism. Most often the not-ready version is actually ready, and the gap is the maker seeing flaws the audience won't. If the gap is real and load-bearing, time-box the next iteration; don't drift the current one.

What if I keep finding new things to add at the end?

Classic late-stage scope creep. Add a parking lot list — "good ideas for v2" — and put each addition there instead of into the current project. Most additions look less essential a week later; the parking lot lets you discover this without spending the time first.

What if I'm finishing but it feels anticlimactic?

Normal. The finishing moment for most projects feels small from inside even when the project itself is significant. The retrospective and the visible artefact help — they make completion legible. Some readers also benefit from a deliberate finish ritual (close the laptop, take a walk, name what's done) that marks the transition.

How do I avoid this in the next project?

Set up the external structure earlier. The public date can be set at the start of the project, not at the end. Named deliveries can be planned from the beginning. The tail gets easier when the system was designed in from week one, not bolted on in panic at week ten.

Smallest move today?

Tell one person today when the thing ships. Specific date. Don't make it too far away. That single text message converts the project from "someday" to a real deadline and starts the finishing system without any other infrastructure. The rest builds from there.