Open this lesson in your favourite AI. It'll walk you through the why, explain the demo, and quiz you on the try-it list.
John Cutler coined 'feature factory' to describe teams that ship lots of features but don't move metrics. Hallmarks: no clear strategy, every backlog item is 'high priority,' shipping is celebrated independent of impact, no postmortems. Recognizing the pattern in your team is the first step to fixing it.
Feature factory symptoms + cures.
Use these three in order. Each builds on the one before.
Explain in one paragraph what a feature factory is and how to recognize it.
Walk me through transforming a feature-factory team into an outcome-driven one over 6 months.
Given a CPO who's incentivized on ship count (not outcomes), how do you fix the factory pattern from below?
SYMPTOMS OF A FEATURE FACTORY:
1. EVERYTHING IS "HIGH PRIORITY"
Backlog: 200 items. All P1.
No real prioritization happens.
2. SHIPPING > IMPACT
Sprints celebrate "we shipped X."
No conversation about whether X helped a customer.
3. NO POSTMORTEMS
Features ship. Team moves on.
Did it work? Nobody knows.
4. ROADMAP = WISHLIST
Roadmap is "every stakeholder's wish."
No theory of value behind what's in vs. out.
5. METRICS ARE VANITY
Tracking: pageviews, signups, MAU.
Not tracking: customer outcomes, retention cohorts, revenue per cohort.
6. NO KILLED INITIATIVES
Nothing on the roadmap has ever been deprecated.
Stuff just accretes.
7. PMs ARE TICKET WRITERS
PM job: file Jira tickets. Pass to engineering.
No strategy. No discovery.
8. CUSTOMER FEEDBACK = FEATURE REQUESTS
Customer says: "you should add X."
PM: "great idea, adding to backlog."
No interrogation of the WHY behind X.
COHEN'S 11 SINS ↔ FEATURE FACTORY:
Feature factory IS Sin #11 in modern dress.
"Developing technology rather than developing products."
Shipping features because they're shippable, not because they're useful.
CURES:
1. PRODUCT THESIS:
"We believe X for customer Y, and this matters because Z."
Every roadmap item maps to a thesis.
Items not mapping: cut.
2. STRATEGY ABOVE TICKETS:
Vision → Bets → Initiatives → Tickets.
Multiple layers; tickets at bottom.
Strategy reviews quarterly.
3. POST-LAUNCH METRICS:
Every launch: explicit success metric + post-launch review.
Did it move the metric? Yes/no/maybe.
If maybe: what would we test next?
4. KILL OLD INITIATIVES:
Quarterly review: what's not working? Kill it.
This is harder than starting new things. Most teams can't.
5. CUSTOMER OBSESSION → CUSTOMER QUOTES:
PRDs cite specific customer quotes.
Roadmap items have customer champions.
"5 customers asked for X" beats "customer asked for X" beats "X seems neat."
6. NORTH STAR + INPUT METRICS:
One overall metric the team optimizes.
Each feature has an input metric that should move the north star.
7. PM = STRATEGIST + DISCOVERY LEAD + INTEGRATOR:
Not Jira janitor.
Owns the product theory + customer evidence + scope decisions.
CASE STUDY — A FEATURE FACTORY I FIXED:
Team of 12, shipping every 2 weeks.
Backlog: 350 items. 27% retention. 9% MAU growth.
Diagnosis:
- No theory.
- 9 simultaneous "strategic" initiatives.
- No postmortems.
Interventions over 6 months:
- Wrote a 1-page product thesis.
- Cut 220 backlog items (literally: closed without doing).
- Killed 5 of 9 initiatives.
- Introduced launch reviews with success metrics.
Result after 6 months:
- 36% retention (+9 pts).
- 21% MAU growth.
- Team morale: improved (less work feels weird at first; productivity improves).
THIS MODULE'S FOCUS:
Audit your team. Are you a feature factory?
Pick ONE intervention this quarter.