February 2026 Scorecard: Four Commitments, One Short Month
On the 1st I wrote down four commitments for February: keep writing daily, take Sundays off tech, narrow my work, and get seven hours of sleep. Twenty-eight days later, here’s how they held up. Two held fully, one held mostly, and one held only partly, in a way that tells me willpower was the wrong tool.
This is the same exercise as the January close-out, with a shorter month and a sharper list.
The scorecard
| Commitment | Target | Result |
|---|---|---|
| Daily writing | A post every day | Done: 29 posts in 28 days |
| No-tech Sundays | Weekly, rescoped to monthly on the 6th | Done (against the revised monthly target): two, both worth it |
| Deeper focus on fewer things | AI consulting plus family presence | Mostly: one new engagement, two noes; better attention at home |
| Better sleep | Seven hours minimum | Partial: averaged 6.5 hours |
Daily writing: done
A post every day: 29 posts in 28 days, with two on the 15th. Added to January’s run, the streak that started on 1 January is still unbroken.
What I noticed this month: the writing got easier but not shallower. I expected the posts to thin out once the novelty wore off. They didn’t. The constraints (daily, shorter, honest) forced more precision, not less. When there’s no time to pad a post, the only option is to say the one thing clearly.
I wrote about the less obvious side effects a few days ago in Fifty-Five Days In, mostly about being wrong in public and writing on the days the tank is empty.
No-tech Sundays: done
I did two of them. The 1st said every Sunday. After the first one I’d settled on one a month, with extras counted as a bonus. Against that, two is done. Against the original plan, it’s half. I’m counting it as a win either way, because the two that happened answered the question I actually had: is this worth protecting?
They were different from each other in ways I didn’t expect. The first, on the 1st and written up on the 6th, was uncomfortable: four phone reaches in the first hour. The second felt almost natural: two reaches, and the kids didn’t complain when the tablets went away. Two attempts are thin evidence, but the direction was clear.
The plan on the 1st flagged a clash: a post every day and a laptop-free Sunday don’t fit together, and I said February would show which one gave way first. Neither visibly gave way. Every Sunday in February has a post dated to it, including the 1st, which was itself the first no-tech Sunday. Scaling the rule back to monthly is what made the clash small: only two days had to carry both commitments, not four.
It keeps going in March.
Deeper focus on fewer things: mostly
The plan named two focus areas: AI consulting and family presence.
On the work side, I took on one new consulting engagement and said no to two others. The January report card also recorded two turned-down gigs, so the count of noes isn’t the progress on its own. The filters I wrote up in Three Filters I Use Before Saying Yes to a Project are the test I’ll hold future noes to, so a no decided once doesn’t need to be decided again.
The plan said Azure OpenAI and agents; the client work that actually filled the month was the Fabric migration.
The slip was the reactive work. I still got pulled into more of it than I should have. The root cause is a habit I know well: treating the inbox as the agenda. Whatever arrives first gets my best hours, which is backwards. Turning down projects is a single decision. Not opening email first thing is a decision I have to make every morning, and I’m not making it reliably.
On family presence, the clearest result came from my wife, not from me. After the second no-tech Sunday she pointed out that I ask better questions on those days: I listen to the answer long enough to ask a follow-up. That’s the half of this commitment that held, and it showed most on the days with a rule behind it.
Better sleep: partial
I averaged 6.5 hours. Better than the five to six I was running before. Not the seven I committed to.
On the 1st I admitted I had no mechanism for this one, only intent, and said I’d find out by the 28th whether intent was enough. It wasn’t. Without a fixed point, bedtime gets set by when work feels done, and work never feels done. The fix is structural: a fixed bedtime that doesn’t negotiate with the to-do list. The AASM and Sleep Research Society consensus of seven-plus hours for adults was the reason I set the target. Half an hour short every night adds up over a month, and the CDC links regularly sleeping under seven hours to higher rates of chronic conditions (CDC MMWR, 2016).
The report-card goals I didn’t carry over
The January report card promised to grade three goals at the end of February: five minutes of meditation a day, a weekly phones-away family activity, and a monthly no-tech day. The monthly no-tech day is covered above. The other two changed shape on the 1st, so here’s what happened to them.
Meditation: dropped, not graded. I took it off the list on the 1st because nothing forced it, so there’s no result to report. If it comes back, it comes back with something forcing it, like everything else that held.
The weekly family activity went monthly. On the 1st I folded it into no-tech Sundays, so when those scaled back to monthly, it did too. That meant two phones-away family days in four weeks, not four. Both of the days we had were worth repeating, which is an argument for adding more later, once monthly is a habit rather than an event.
What went well
Client work was substantive. The Fabric migration project I’m on is interesting work: a good team, a real problem, and the kind of engagement I want more of.
Writing surfaced blind spots. Several posts started as confident takes and ended as more nuanced ones once I had to think them through on the page. The Expert Trap and The Code Review Comment That Wasn’t About the Code were the clearest examples.
Family time was better protected. The no-tech Sundays were as good for me as for the kids. Coming back to work on Monday feels different after a day fully off.
What didn’t work
Afternoon energy. The 2 to 3 PM window is consistently low-output. I’ve been forcing demanding work through it instead of working with it. The fix is to schedule lighter work in that slot rather than fight it.
Side project momentum. I have a small tool I’ve been building for a few months. February added maybe three commits. That’s not enough to keep the context in my head between sessions. At three commits a month I’m maintaining the project in my head, not moving it forward.
What February taught me
Short months are clarifying. Twenty-eight days isn’t much time. The constraint made it easier to see which commitments were real and which were wishes.
Focus means saying no to things I enjoy. Not just the things I dislike. Declining work I’d be bored by is easy. Declining work I’d find interesting, because it would split my attention, is the hard part, and it’s the part that creates focus.
Repetition turns effort into default. The first no-tech Sunday took effort; the second felt almost natural. That’s why I’d rather give the March sleep rule a few weeks before judging it than drop it after one bad night.
There’s a pattern across the two months now. The commitments that held were either a single decision (saying no to a project) or something with a hard deadline and a visible record (a post every day). The ones that drifted, sleep and the inbox, depend on a fresh choice at the moment I’m most tired or most distracted. That’s the design problem for March.
Into March
Better execution than January, and this time the misses share a cause I can design around. So March turns both of them into fixed rules: bed at 10:30 PM with the alarm at 7 AM, and email checked at 9 AM and 1 PM, then not again until the next morning. Afternoons get time-blocked for lower-energy work. Daily writing, the monthly no-tech Sunday and the Fabric project carry on unchanged, and the side project gets a narrower goal: ship the minimum version I’ve already built.
If you’re reviewing your own month, sort your misses by type before you set next month’s goals. A miss that needs more effort and a miss that needs a different structure look the same on a scorecard, but only one of them gets fixed by trying harder.
The full plan is in March 2026: A Fixed Bedtime, Two Inbox Checks, One Tool to Ship.