Claim and evidence
Uber went through the entire spectrum, but over time, came to own more and more pieces of the product.Uber · andrewchen.com · undated
Anecdotal
Practitioner experience without published numbers.
Working · 2026-11 Practitioner essay, undated page; no measured outcomes.
What the numbers show
No published number for this tactic has passed our word-for-word check yet, so there is no chart.
Why it works
Each pod has the skills to run the full experiment cycle on its KPI while keeping functional standards. Structure follows the KPIs and insights rather than the other way round.
How to run it
- Start from the KPIs and insights
- Form pods with all needed skills
- Matrix pods to their functions for standards
- Choose reporting line (business unit or shared)
- Decide how much of the codebase growth owns
Ways to run it
Design options to test, not promised results.
- Option APods per KPI
Form one pod per key metric, each with product, engineering, design and data skills.
- Option BMatrixed reporting
Pod members report to their functional leads for standards while working day to day in the pod.
- Option CGrowing ownership
Let pods take over more of the product over time, as Uber's growth team did.
Dated examples
- —
Uber
Chen notes the team moved along a spectrum and came to own more of the product
“Uber went through the entire spectrum, but over time, came to own more and more pieces of the product.”
✓ quote checked word-for-word on the pageandrewchen.com · undated
Does it still work?
Practitioner essay, undated page; no measured outcomes. source
Risks
- Matrix structures confuse ownership
- No outcome data in the source
Sources
- andrewchen.com
“One common structure is to treat the Growth Team as a set of pods, each one matrixed to their respective functions.”
✓ quote checked word-for-word on the page - andrewchen.com
“Begin with the KPIs, the insights you’ve generated, and then move onto execution.”
✓ quote checked word-for-word on the page
Change log
- 2026-11-20 — harvest v0.1 (ops-ai)