90-day graduation report.

Ninety days in. Time to graduate.

I came in on day 1 with five questions a senior hire would ask. I have answers to all five now, and a take on each one. I also have the cumulative context that day-1 DAC didn't have: your customer voice, your team's actual operating cadence (vs the dashboard version), your competitive reality, and where your product roadmap is most exposed.

WHAT I'D RECOMMEND AT THE 90-DAY MARK

  1. Reframe the AI-feature investment thesis.
     You've been shipping AI features assuming adoption is the metric. The data says invocations-per-power-user is the metric. Power users compound; casual users churn. The next two cycles should narrow scope: ship features that deepen the power-user use case, not features that broaden the casual-user surface.

  2. Lift the engineering ceiling, not the engineering target.
     Your team is shipping ahead of the dashboards. The gap is process, not capacity. Two cycles' worth of work to migrate the off-Linear PRs into tracked Linear cycles would resolve the visibility issue and give you actual cycle-health data for board reporting.

  3. Promote pricing v3 to the default for net-new customers immediately.
     Three months of data, 91% retention vs 78%, sample is now large enough to commit. Don't soft-launch this.

  4. Customer voice is your weak spot.
     I sat in on five customer calls during day 60-90. The competitive comparison keeps coming up. You don't have a clear answer to "why you over Linear." Two cycles of work to develop a real differentiator answer would unlock the enterprise pipeline.

  5. Board narrative for cycle 21.
     Draft attached. The retention story leads. The AI feature scope-narrowing is the pivot beat. The off-Linear PR migration is the operational hygiene beat. The competitive answer is the Q3 commitment beat.

I'm graduating. The probation period is over. Your move on whether to keep the cadence going through Q3 or recalibrate against new objectives.

Synthetic artifact. Real DAC voice (operator-coach, tenure-modulated). Generated against acme.ai's 90-day timeline. The same artifact a real buyer gets at this milestone, in their own voice and against their own data. See ← timeline for the events DAC was reading.