Your week, in decisions.
The weekly intelligence page for Crux AI members. Personal. Cumulative. Yours alone.
You caught yourself anchoring on the first number three times this week. Twelve weeks ago, you caught it once. The gap between your confidence and your outcomes is the smallest it has been all quarter.
TRAILING 30 DAYS
Delay the launch two weeks to fix onboarding, or ship on time.
Move the team to a four-day week for one quarter.
Switch the main supplier after years on the same contract.
CLARITY + REALITY / TRAILING 30 DAYS
The two decisions you framed weakest are the same two where your assumptions broke later. How clearly you frame a problem predicts how well it turns out. Your sharpest catch this month: spotting yourself anchor on the first number during the launch call, and stopping to check it.
UNCOVER / TRAILING 30 DAYS
The two cruxes you stopped at why-two are the two where outcomes underperformed. Drilling matters. Your confidence at commit averages 71%, but your owned outcomes are scoring 8s and 9s. You are calling these too cautiously.
EXECUTE / TRAILING 90 DAYS
Three premortems you wrote in March arrived this month: two of them Find failures, both linked to assumptions you flagged but didn't fully test. The pattern: you spot belief gaps in Reality and under-weight them at eXecute. Of the seven If-Then triggers that fired this quarter, you followed your pre-committed response on five.
REFLECT / TRAILING 30 DAYS
The gap between how confident you felt and how decisions actually turned out has closed by nineteen points this quarter. The flat grey stretch is before the habit took hold. Once you started testing assumptions before deciding, the line began climbing toward true.
Both decisions you closed out scored well. The results came in 12 points ahead of how confident you felt going in: you were tougher on yourself than the outcome deserved. Fastest decision this month: the launch call, 4 days. Slowest: the pricing review, 23 days.
Delay the launch by two weeks to fix the onboarding flow, rather than ship on time with a known drop-off.
The launch date had become the thing everyone was protecting. You framed the real question: is the date the goal, or is a launch that actually lands the goal? Six assumptions surfaced; five became things you could test, one you dropped.
The constraint isn't headcount, it's where the headcount is positioned.
Why Not: because the date had been promised so often that questioning it felt like failing, not thinking.
“What would prove this crux wrong?”
If a quick test shows users get through onboarding fine as it is, the flaw is not the real problem and shipping on time is the right call.
Four pre-mortems written, one per verb category. Two if-then plans set in advance: “if a quick onboarding test comes back clean, ship on the original date.”
The call held up. Confidence going in 73%, outcome 9/10. Carry: two weeks of fixing the flaw was worth far more than the date you gave up.
Should we end the side project that is quietly draining the team?
Should we go after the bigger client, or stay focused?
The roadmap is locked for the quarter, so there is no point raising it.
Finance will never approve a phased rollout, so do not bother asking.
Stalls compound. Constraints unchallenged become infrastructure. The longer you leave them, the more they cost.
4-WEEK COMMITMENT COMPLIANCE
Of the commitments you logged in prior Reviews, the rate you've held.
Let one big number from an earlier decision set the frame for an unrelated one.
Re-test whether the roadmap is really locked this week.
The next Review reports back.