My 12 Week Year

#

TL;DR

#

I run The 12 Week Year and Getting Things Done together on one printed sheet, a time log, an execution scorecard for outcomes and actions, and a GTD capture box, reviewed daily, weekly, and monthly. Vision comes first, actions are scheduled day of week commitments instead of tallies, and the target is 80%, not 100%.

Want to try it yourself? LukeWerks generates and prints the sheet.

Every January I set goals for the year, and by March I had forgotten most of them. The problem was never the goals, it was the horizon. A year is long enough that “I’ll get to it” always sounds reasonable, right up until it’s December.

What A 12 Week Year Actually Is

#

The 12 Week Year comes from Brian P. Moran and Michael Lennington’s book of the same name. Their argument is simple. Most people and most companies run on an annual planning cycle, and a year is long enough that real urgency never shows up until the last few weeks of it. Treat twelve weeks as a full year instead, and that urgency shows up in week one.

Four of these blocks fill a calendar year, each one short enough that there’s no room for “later,” and long enough to actually finish something worth finishing.

Two Systems, One Sheet

#

On its own, the 12 Week Year answers “what am I trying to achieve.” It doesn’t answer “what do I do with the fifteen random things that land on me before lunch.” For that half of the problem, I layered in Getting Things Done , David Allen’s system for capturing everything as it happens instead of trying to hold it in your head.

Getting Things Done handles the input. The 12 Week Year handles the direction. Together they turn one printed page into both a scorecard against my twelve week goals and a capture net for the day’s noise.

Vision Before Anything Else

#

Before a block starts, I write vision at two horizons. A longer one, three to five years out, that stays fairly stable, the kind of person I’m trying to become rather than what I’m doing this quarter. And a nearer one that translates that longer vision into what this specific block should accomplish.

The vision is what keeps a block from turning into an arbitrary checklist. Without it, an hour of coding is just a task. With it, that hour is one rep toward something I actually decided mattered.

Outcomes And Actions, My Way

#

The 12 Week Year borrows its core vocabulary from sales. A lag indicator is the number you’re chasing, revenue for example. A lead indicator is a countable activity known to move that number, calls made for example. That works fine for “get more sales.” It works badly for “get healthier,” because there’s no obvious “calls made” equivalent for something like that.

I kept flipping which was which. Lag sounds like something trailing behind, an activity. Lead sounds like the main event, the goal. That’s backwards from what they actually mean, and it tripped me up enough times that I stopped fighting it. From here on I’m calling the goal an outcome, and the scheduled thing that drives it an action. Same ideas, borrowed from the same lag and lead vocabulary above, just named the way my own head actually parses them.

My adjustment is to make actions scheduled time block commitments rather than tallies. Instead of counting how many times I did something, I decide in advance which days of the week I’m doing it, then track whether I actually followed through on those specific days. A blank day isn’t a missed one, it’s a planned exception, the same way a cheat day is planned rather than an accident.

Here’s a real slice of one block, grouped under an outcome I’ll just call “Lose 5 Pounds.”

Saturday is the one day “No Food After 7pm” has no letter. That’s not a slip I’m hiding from myself, it’s a cheat day I chose on purpose before the week even started.

The Daily Sheet

#

Each day starts from one printed page split into three parts, a time log on the left, the execution scorecard on the right, and a notes section along the bottom.

The Time Log

#

The left side of the sheet is a plain time log, half hour rows running from six in the morning to ten at night. Its purpose is honesty. At the end of the day I compare what actually happened in each block against the action commitments from the scorecard. It’s easy to feel productive without that feeling being connected to anything I said mattered, and the log is what closes that gap. It also surfaces patterns, some controllable and worth fixing, some not controllable and better off scheduled around instead.

A sample stretch from an actual sheet looks like this.

The log isn’t only a record after the fact either. Action commitments can be blocked in ahead of time, before the day even starts. Any block that keeps proving valuable tends to graduate into a recurring calendar commitment for the rest of the twelve weeks.

The Execution Scorecard

#

This is the table shown above under “Outcomes And Actions, My Way.” One row per action, grouped under its outcome, with the day of week pattern next to it. At the end of the day I cross off any letter I earned. Anything left uncrossed once its day has passed was a miss, plain and simple.

The outcome and action structure itself doesn’t change day to day. It’s set once at the start of the block and held fixed for all twelve weeks, so the daily routine stays simple, log time, hit commitments, capture noise, review honestly at night.

Notes

#

Along the bottom of the sheet is a wide open space labeled Notes. This is the Getting Things Done inbox. Anything that shows up during the day, a task, an idea, a reminder, gets written down there the moment it appears instead of being trusted to memory.

A sample from an actual sheet looks like this.

• Call dentist, reschedule Tuesday appointment
• Sarah asked about the Q3 report, follow up
• Idea, look at the pricing page copy again
• Order anniversary gift before Friday
• Renew car registration

At the end of the day, anything on that list still open gets copied by hand onto tomorrow’s sheet. That copying is deliberate friction. It would be easy to build a system where unfinished items roll forward automatically and silently. Writing each one again by hand forces a real decision every single day, keep carrying this, schedule it properly, hand it off, or drop it outright. Nothing lingers by accident.

Reviews, Nested

#

Four review cadences exist, each one rolling up the cadence below it.

Daily

#

The end of day review described above.

Weekly

#

At the end of each week, I collect the week’s sheets, score every action’s completion rate, roll that up into a per outcome read, and flag anything consistently over or under performed for more than a week running.

Grading The Week

#

Every Friday, at the top of that day’s sheet, I write my grade for the week. Something like 18/20, which is 18 of the week’s 20 targeted action days actually crossed off, worked out to 90%.

That math is only quick on Friday because I don’t save it all for Friday. Each day I jot down that day’s own tally right on the sheet, how many of today’s targeted letters actually got crossed off out of how many were targeted. By the time Friday shows up, the weekly grade is just adding five small numbers together instead of paging back through an entire week to recount everything by hand.

What that score should actually be, and why chasing 100% backfires, is worth its own explanation, see the example below.

Monthly

#

Roughly monthly, about four weekly reviews get rolled up into a month level view per indicator, since a twelve week block runs close to three months.

End Of Block

#

The final week of a block is left open on purpose rather than scheduled like the rest. That week is for reviewing the block as a whole, honestly re-questioning the vision behind it (usually the answer is still yes, but not automatically), and setting the next block’s outcomes and actions before day one of the next twelve weeks starts.

From Vague Idea To Real Commitment

#

Here’s a walkthrough of how a fuzzy idea actually turns into something trackable on the sheet.

Say the idea is “Learn AI.” That’s broad. It has intent, but no focus yet. This step is really about how any vague idea gets turned into concrete steps, whatever the idea happens to be.

What does “Learn AI” actually mean to me? Is it the mathematical underpinnings? Is it just learning to use ChatGPT more effectively? Maybe there’s a book, “Playing With Reality ,” that caught my eye. Narrowing it down to something an action can actually attach to takes a little work, and that’s part of why this methodology builds in real thinking time between blocks instead of rushing straight from one into the next.

In this case, I decide I want to read that book, and I also want to learn about harness engineering. “Learn AI” becomes the outcome I’m after, and those two things become actions under it.

Before I schedule anything, I write down everything Learn AI could reasonably ask for. That’s the reference list, more candidates than any single week could actually hold.

Nothing on that list has a day yet, and it shouldn’t. Trying to run all four as real weekly commitments at once is exactly how a block gets overloaded before it even starts. Paring that list down into an actual week is the real skill here, not writing the list itself.

The book comes first, and it gets its own lane. It’s short, so there’s no reason to let it compete with something that’s going to run most of the block.

Thirty minutes of reading, Monday through Thursday, and a standing Friday task to add at least one page to my notebase with links, so notes have somewhere to land instead of staying in my head. That Friday task was never really about the book. It’s going to outlive it, whatever I’m learning under “Learn AI” gets a note written about it that week.

I’m guessing I can get through the book in a couple of weeks. Once it’s done, harness engineering moves into that lane. It’s messier than a book, its resources are scattered and the field itself is still emerging, so I settle on a study and apply approach for it, 2 hours a week on research and 6 hours a week on application. That’s roughly a 1 to 3 split, research to application, closer to 70/30 than half and half, and it’s my own rule of thumb rather than a number I’d point to a citation for. It does lean the same direction as UC Berkeley’s Center for Teaching and Learning , compared to lecture alone, active learning raises performance and lowers failure rates. Spending more time doing than being told is the well supported side of that argument, the exact split is just mine.

That works out to an hour of research on Monday, another hour on Tuesday, defining that week’s application from those results Tuesday evening, then 2 hours a day Wednesday, Thursday, and Friday building small proof of concept work and extending my notes. The notebase task carries over unchanged, it was never tied to the book to begin with.

Okay. All set. Then the first Monday of the harness engineering lane shows up and throws the whole plan off track.

Because I’m dutifully logging every 30 minute block, when I get to the end of the day and realize almost none of it got done, I can see exactly why. Looking back at the log, it’s all fairly typical Monday stuff, meetings and interruptions I didn’t plan around. Maybe Monday just isn’t a good research day.

Sometimes that’s what a pattern like this means, you notice it and adjust the plan. That’s fine. What you don’t want is to be constantly tweaking your outcomes and actions week to week. So in this case, I move that block permanently, one 2 hour research block, once a week, on Tuesdays instead.

That’s the fix for a real pattern. A single bad Monday is a different problem. If Monday is normally a fine research day and this particular one just went sideways, that’s a one-off, not a trend, and a one-off gives you two honest paths. Catch up that evening and tick the box before bed. Or let it go.

Here’s what the lane looks like once harness engineering has taken over and that Monday adjustment has stuck.

That’s the swap in practice. Neither of the two scheduled tables was ever meant to be the final word, both are just slices of the full list at the top, worn down into whichever week actually fit. When the book finished, its lane opened up and harness engineering, the next thing on the list, moved in. The notebase task is the one row that just carried straight through every version unchanged.

The target for either table is 80%, not 100%. If you’re like me, that number might drive you a little nuts on its own. But chasing 100% in this system is what will actually drive you nuts, and 80% is the number I’m holding myself to at the top of my Friday sheet every week. This is about consistent execution toward a goal, not a flawless scorecard. If the actions are sized right to begin with, 80% still gets you there, without one part of your life bulldozing every other part.

That’s really the whole point of the system, it’s there to support the goal, not become a second job. If tracking a two week book starts costing more thought than reading the book would have, something’s gone wrong.

Paper Or Spreadsheet

#

I’m not a luddite. I run my calendar, my email, and my task list digitally like everyone else. But for this particular kind of tracking, paper just works for me. Writing something down and physically crossing it off carries a weight that clicking a checkbox doesn’t.

Try It Yourself

#

LukeWerks is the tool that generates this sheet, set up your own outcomes and actions and print your own twelve week block.

References

#

Brian P. Moran and Michael Lennington, “The 12 Week Year ”

David Allen, “Getting Things Done ”

Kelly Clancy, “Playing With Reality ”

UC Berkeley Center for Teaching and Learning, “Active Learning ”