// how we work
A ship cadence, not a slide deck.
The 90-day wave is how a subscription runs. Each wave ships a written diagnosis, shipped changes on both stores, a recommendation panel, and a wave-close review. The weekly cadence is the compounding engine underneath.
01
the 90-day wave
The wave, week by week — with the deliverable attached.
01
Diagnosis
Full listing audit across the five practices — intent, native metadata, visual + CPP, reviews, discovery. Baseline pulled on rank, conversion, and assistant recommendation panel.
→ Written diagnosis + prompt panel v1 + 90-day plan
02
Intent + metadata
Keyword corpus rebuilt against real intent clusters. Title, subtitle and keyword field rewritten for capability-clarity across both stores in the primary market.
→ Metadata deck + rollout plan + change log
03-04
Visual + CPP
Screenshot narrative rebuilt around the decision the user is making, not the product tour. First Custom Product Pages authored against paid campaign intent.
→ Screenshot arcs + first CPP set + A/B test lane opened
05-06
Reviews + discovery loop
Review response cadence installed, sentiment themes converted into listing and product inputs. Assistant recommendation panel measured weekly with prompt drift tracked.
→ Review loop installed + first recommendation panel report
07-10
Compound + iterate
Every ship gets a release-diff line: what changed, what moved, what didn't. Losers are rolled back inside the same week. Winners are pushed to secondary markets.
→ Weekly release diffs + winners promoted
11-12
Review + wave close
Written wave review: what compounded, what didn't, what the next 90 days must ship. Portfolio-level roadmap re-planned against product and marketing calendars.
→ Wave review + next-wave brief
02
the weekly cadence
What ships weekly, bi-weekly, monthly, quarterly.
Weekly
- →Rank scan across tracked keyword set
- →Store-surface + recommendation panel monitoring
- →Release diff + Slack async update
- →One shipped change minimum
Bi-weekly
- →CPP ship or screenshot A/B rotation
- →Review-response batch + theme review
- →Strategy call (Growth and above)
- →Paid ↔ organic reconciliation
Monthly
- →Full performance report + release log
- →Recommendation panel v2/v3 refresh
- →Roadmap review with product + marketing
- →Boundary-strip update on active cases
Quarterly
- →Description-body rewrite + long-form refresh
- →Category-share brief + competitive read
- →Portfolio review (Scale + Portfolio)
- →Wave close + next-wave scoping
03
the decision surface
Who owns what, and how decisions travel.
Who owns the ship
The senior operator on the intro call. No handoff, no pod rotation.
Where changes live
A shared release log with a line per ship — date, surface, change, expected mechanism, observed movement.
When a test dies
If a test doesn't move the target in its intended window, it is rolled back that week. The next test ships the same cycle.
How escalation works
One Slack channel, one weekly async, one strategy call cadence. Anything urgent gets a same-day written response, not a meeting.
What gets published externally
Nothing without your written approval. Case pages stay under NDA until you say otherwise.
04
the operating rules
Four rules the practice never breaks.
One operator per account.
The person on the intro call runs the engagement. No junior handoff, no pod rotation, no account manager sitting between decision and ship.
Every ship gets a receipt.
Slack update, dashboard delta, or a written release-diff line. No deliverable disappears into a doc.
Nothing gets parked.
If a test fails, the next test ships in the same cycle. The cadence is the compounding — not the roadmap document.
Evidence beats opinion.
We publish written diagnoses, release diffs and boundary strips. If we can't show the mechanism, we don't claim the result.
// ready?