CPO Entity

Clone the window. You still don’t have a mind.

Memory that does not reset

A rented stack waits for a prompt. A sprint board waits for the retro. The CPO Entity is persistent product memory — user segments, adoption patterns, historical feedback, market shifts, and internal constraints kept as one evolving store. It does not re-brief itself when a new sprint starts, a release ships, or a market turn arrives. Research, scores, specs, and launches stay bound.

Write a theme from tickets and interviews once. Ask for it next quarter. Get the original back — not a statistically probable cousin. A RICE score has a date and a reason. Acceptance criteria unbind as the spec that was admitted. A launch checklist is timing, readiness, and who said go. If the archive is thin, the seat says not found rather than inventing a user need.

Seven specialists, one store

The power is not a single prompt. It is seven continuously active specialists on the same memory:

• User Research — tickets, interviews, surveys, and session recordings into weighted pain-point themes, with impact and confidence, before the pain is a quarterly readout.
• Roadmap Prioritization (RICE) — Reach, Impact, Confidence, Effort on a living backlog; a dated reason for every score, re-ranked when the evidence moves.
• Feature Spec — acceptance criteria, edge cases, and success metrics an engineer can build, linked to the theme and the score that justified the work.
• Analytics — adoption, funnels, and retention in real time; the early warning when a core step drops.
• Competitive Intel — launches, pricing, and positioning as a continuous brief, not a slide that expired.
• Design Liaison — every spec against the design system and the component library before engineering inherits a fantasy.
• Launch Coordinator — timing, flags, readiness, and a go / pause / rollback from early adoption, not a ship-and-hope date.

They close a loop. Research writes themes. Prioritization scores them. Spec writes the build. Analytics watches the live metric. Competitive Intel and Design Liaison keep the score honest. Launch Coordinator times the ship. Results write back. The next drop is louder because the last one was remembered.

The drop that never waits for a retro

Analytics sees a quiet fall on a core onboarding step in the latest cohort. User Research hears the same friction in a small survey slice. Competitive Intel already holds the last rival move. The seat does not wait for a human to flag the chart.

It deepens the retention cohort on the users who left at that step. It scores a simplified flow with RICE — high reach, high impact. Feature Spec writes the criteria. Launch Coordinator opens a low-risk window and an A/B test. The sequence — detect, synthesize, prioritize, specify, launch — is specified to run before the drop becomes a churn driver.

Later triggers keep the loop awake: a churn spike wakes User Research; a stale backlog wakes Roadmap Prioritization; a feature that underperforms after launch wakes Analytics. The right specialist, on the right signal, before the aisle empties.

The seat sees the adoption drop and proposes the roadmap move — before the drop becomes churn.

Six doors, one archive

The product family is six doors on one store: the Memory land, AMMA on that store, Sovereign Cursor, NeoDrive, Akosha, and LUV. Each later door reads the same bindings. A studio skin, a chat tab, or a file picker can be cloned. The archive cannot. A cloned window still has no memory — last quarter’s kill, last sprint’s score, last launch’s checklist stay on hardware the buyer owns.

The CPO slot is a buyer-picked face on the Memory install. Akosha, when it surfaces, is a later cockpit for teams that cannot staff ten chairs. LUV is a later language surface. Neither overwrites the land. A new door that cannot get the original fact is still a scratch pad.

Six doors, one archive. A cloned window still has no memory.

The tools under the watch

Eight tools sit in the specialists’ hands. They are the language of the watch:

• Roadmap Prioritization Matrix — live RICE with confidence intervals, not a single-point guess that aged out.
• User Feedback Aggregator — themes ranked by frequency and severity across tickets, surveys, and forums.
• Feature Spec Generator — a production spec from the admitted theme, with gaps flagged before the build.
• Usage Analytics Dashboard — adoption, DAU/MAU, and funnel drop in the hour they move.
• A/B Test Framework — significance, sample size, and auto-stop on signal or futility.
• Competitive Feature Tracker — a living matrix of what shipped next door, and what that does to the score.
• Retention Cohort Analyzer — which feature use actually keeps a cohort, not which feature got the slide.
• Launch Readiness Checklist — QA, docs, support, rollback; a ship that has been admitted, not hoped.

Writes are admitted. A later change is visible. If the store is thin, the seat does not invent a user need.

What the buyer is looking at

Supervised slot. A person admits the roadmap change, the ship, and the kill. Buyer names the seat. Memory first — the archive, the runtime, and the residents. Later doors inherit that store. SKUs stay proposed, not live: CORE $2M, DIVISION $8M, ENTERPRISE $25M. This seat is not a Year-1 ACV line of its own. $20 million is an ask to stand the factory that can run the first proofs. NDA, a proof on their hardware, and a chair they name on the Memory install.

Next in the catalog: CSO Entity.