Zone Defense for Product Work
Spread taste-makers out to cover the whole field instead of clustering on the same problem
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 3
- Confidence
- 87%
With ideas being thrown everywhere and top-down annual planning broken, Ambrosino runs product like a zone defense: if two product people are working too closely, that's a bad signal. You deliberately create space between people so a small number of high-taste operators cover the entire company's surface area, filling gaps rather than overlapping.
Origin
Andrew Ambrosino credits ongoing conversations with his colleague Alexander for the zone-defense analogy applied to the Codex org.
Core principles
- 01Two product people working too closely is usually a bad signal
- 02Curation, steering, and alignment replace top-down long-range planning in a high-chaos environment
- 03You want full company coverage from a few taste-makers, not redundant effort
- 04Fill the gaps: figure out who's best at what and create space between them
How to run it
- 1
Run a force-directed spacing exercise
Treat product people like nodes that repel each other — actively push them apart to find where the gaps in coverage are.
Pro tip If you notice two PMs converging on the same problem, treat it as a coverage bug to fix, not collaboration to celebrate.
- 2
Assign by strength to maximize coverage
Work out who is best at what, then place people so that between them they cover the whole product surface with minimal overlap.
Watch out Top-down, year-long planning will not work in this environment — coverage has to be dynamic.
- 3
Fill gaps with product-minded engineers
Hire engineers who are product-minded so features arrive coherent, rather than needing a full team to review everything for product coherence after the fact.
Pro tip Aim for everyone to have product skills, but let people go deep on different things.
Watch out You still can't have everyone work on everything — breadth and depth limits are real, which is why managers don't disappear.
In the wild
In a world where 90 people throw ideas around, Ambrosino needs taste-makers to guide things from inception to the right product. The team spreads out for 'company coverage' so nothing important is left unowned.
→ A double-digit engineering team (about half that in design, a few product people) covers a product OpenAI treats as the culmination of everyone's work.
Common mistakes
Letting two product people work too closely
Overlap means wasted coverage and duplicated steering; the same taste could have been spread to cover an uncovered part of the product.
Keeping top-down annual roadmaps as the coordination mechanism
When features' viability depends on model capability that shifts monthly, long-range top-down planning produces false precision and leaves chaos uncoordinated.
Is it for you?
Best for
Product leaders coordinating small numbers of high-agency operators across a fast-moving, idea-saturated org
Not ideal for
Stable, slow-moving orgs with well-defined long-range roadmaps and clear feature ownership
From the transcript
“if two product people are working too closely that's often not a good signal”
“you sort of want to do this like forced directed activity where you're like where are the gaps”
“you basically want company coverage. And so you spread out”
From the episode
OpenAI Codex lead on the new shape of product work
Andrew Ambrosino