In this article
- The 90-day pattern we kept seeing
- Why week three is where fixes actually die
- Build one checkpoint, not five
- Watch for the quiet rollback
- What to do when a fix genuinely isn't working
I tracked 14 process changes across six client teams between January and June 2025, everything from new stand-up formats to revised approval chains. Eleven of them were still running at the 90-day mark. Three were quietly abandoned, and in every one of those three, nobody noticed the change had died until I asked about it directly.
The 90-day pattern we kept seeing
Here's the thing. A new habit doesn't fail all at once. It fades. Week one, everyone follows the new format because it's new and someone is watching. Week two, attendance or compliance is still high, maybe 90 percent. By week four, if nobody has checked back in, that number often drops to 60 percent without a single conversation about whether the change was working.
One product team I worked with adopted a Monday async update instead of a live meeting. For three weeks, everyone posted. In week four, two people missed it and nobody followed up. By week six, the async update was effectively dead, and the team had drifted back into pinging each other individually, which was the exact problem the change was meant to fix.
Why week three is where fixes actually die
Plus, week three is usually when the initial motivation wears off but the habit hasn't become automatic yet. That's the gap. BJ Fogg's research on behavior design puts the automaticity point for a simple habit at roughly three to ten weeks of consistent repetition, and teams are no different from individuals here. Skip the check-in during that window and you're relying on willpower alone to carry a process that hasn't become routine yet.
In my experience, the teams that made a change stick almost always had one thing in common: a scheduled check-in at week three or four, with one question. Is this still working, and if not, what needs to change.
Build one checkpoint, not five
Surprisingly, adding more tracking makes this worse, not better. A logistics team I advised in 2024 tried to monitor a new handoff process with a daily compliance spreadsheet. The tracking itself became the burden, and the spreadsheet got abandoned in ten days flat, well before the actual process did. A single checkpoint conversation, scheduled once, beats five metrics nobody updates.
A fix that nobody revisits doesn't fail. It just quietly stops existing, and everyone finds out three months later at the same time.
Watch for the quiet rollback
That said, a rollback rarely announces itself. It looks like one missed update, then two, then a team that's back to the old pattern without ever deciding to go back. If a manager notices two consecutive lapses in a new habit, that's the signal to ask what's getting in the way, not to assume the team simply forgot.
What to do when a fix genuinely isn't working
But not every abandoned change deserves to be saved. If the week-four checkpoint surfaces a real problem, the format doesn't fit the team's actual workflow, the meeting time conflicts with something recurring, then the answer is to redesign the fix, not force compliance with a broken one. The difference between a habit that needs more time and a habit that needs to change is usually visible by week four if anyone bothers to look.
The bottom line
Most team fixes don't fail because the idea was bad. They fail because nobody scheduled the one conversation that would have caught the drop-off in week three or four. Pick one checkpoint, put it on the calendar before you start, and ask a single honest question when it arrives: is this still happening, and why or why not.
Frequently asked questions
How long should we wait before deciding a team change failed?
Give a genuine process change 8 to 10 weeks before judging it. Most changes we tracked either held or collapsed somewhere between week 4 and week 8, not in the first two weeks.
Is it normal for a new habit to feel forced at first?
Yes. Every habit change we've tracked felt forced for the first two to three weeks. The teams that stuck with it past that point kept the habit. The ones that dropped it in week two usually called it a bad fit instead of an unfinished one.