Rituals that compound.
Rituals that don’t.
Most product orgs run ten weekly rituals. Three of them produce decisions; the rest produce artifacts. The Cutler/Goodhart test sorts them.
(Goodhart: when a measure becomes a target, it stops being a measure. Cutler: a ritual is only a ritual if it changes a decision.)
Three rituals AI killed. Did you replace the rigor?
The PRD review meeting
Killed because Claude writes a passable PRD in 90 seconds. The rigor it produced (other senior PMs catching scope drift, framing problems, naming risks) didn’t move anywhere. Most teams just stopped reviewing. The drift now lands in production three cycles later.
The competitive teardown
Killed because every AI agent has competitor research as a first-class feature. The rigor it produced (someone on the team forming a real opinion about why we’re different) doesn’t happen automatically. The agent gives you an even-handed summary; opinion is a human job.
The status update Friday memo
Killed because Linear webhooks now emit real-time progress. The rigor it produced (one human filtering signal from noise, naming what was actually at risk) is worth ten dashboards. Most orgs replaced it with raw feeds. Now everyone is drinking from a firehose without context.
Cycle A shipped 14 features. Cycle B shipped 8.
Same team size. Same stack. AI-assisted delivery either explained the gap or it did not, and the answer matters.
If the variance is spec quality, it shows up as a Context dimension dip. If the variance is individual heroics, it shows up as a Team Operations dip. If the variance is unmeasured, the next quarter goes chasing a heuristic instead of the cause.
Six months ago you did not need to measure this. AI-assisted delivery removed the headcount-explained excuse for variance.
How long does a customer signal take to reach a roadmap change?
If a real risk surfaces in week 1, when does it land in a cycle? Two weeks is a continuous-improvement org. Six weeks is a reporting org. Eleven weeks means the signal died in a quarterly review deck.
DAC traces every roadmap change back to the signal that triggered it. When latency drifts past two weeks, the named pattern fires and the move surfaces in the next cycle plan. You stop hearing about the gap from customers and start seeing it before the next standup.
Right now, do you have a named cross-framework tension and an owner for closing it?
When team capability is ahead of what the product reflects, that is a pattern with a name and a closer. When the product is ahead of what the team can support, that is a different pattern with a different closer. When all three frameworks line up at the same depth, the org compounds without you in the room.
If nobody can name the current pattern, that is the read. The system surfaces it before you have to ask.
Are you the operating system, or have you built one?
Take two weeks off. Don’t answer Slack. When you come back, can you tell what happened from the system, or only from people’s memory?
A real operating system tells you: what shipped, what slipped, which decisions got made by whom, where the team scored last cycle, what tension showed up, what was already coached, what your lead deferred. A title that conceals three unmanaged orgs cannot do this. The vacation test is the cleanest version of "does the rigor survive me."
DAC scores the system, not the leader. The score is the read; the named patterns are the diagnosis; the moves are the prescription. Cycle over cycle, the gap between you and the system narrows. Eventually, you can take the vacation.
If you also score yourself: For senior PMs. If you brief the board: The 60-second board story.
See which rituals are leverage and which are theater.
Get your read in 2 minutes. Free. No sign-up. Just your product's website.
Free for 30 days. Sign up in 60 seconds. Day 1 starts when you finish.