Peter Deng on the Five PM Archetypes, Scaling from Zero to Billions, and the Six-Month Rule

Key ideas

Five PM archetypes describe how great product people differ. Consumer PMs are half designer; Growth PMs are half data scientist; GM/Business PMs are half MBA; Platform PMs build tools for other builders; Research/AI PMs are half researcher-engineer. Each demands a different skill profile and thrives in a different product context.
The Six-Month Rule defines the right hiring bar. In six months, if Peter is still telling someone what to do, he has hired the wrong person. The person hired should be telling him what to do by that point — not because they ignore authority, but because they have the context, conviction, and initiative to lead from their domain.
Growth mindset is the single most important interview criterion. Deng screens for it with one question: describe a significant mistake, explain why it happened, and show how it changed how you think and work. Authentic vulnerability and genuine reflection signal the capacity to learn under pressure.
Go slow to go fast in the one-to-one-hundred phase. The zero-to-one phase rewards moving before understanding. The one-to-hundred phase rewards building systems, instrumentation, and rigour that compound. Teams that skip this pay a compounding cost.
Sometimes the product is not the product. Deng’s Uber example: Uber Reserve succeeded not because of a polished interface but because it solved for peace of mind — the customer’s real job was certainty about their early-morning ride. Understanding which element of the offering actually matters is more important than product craft.

Summary

Peter Deng has held product leadership roles at an unusual density of consequential companies: PM #4 at Facebook, Head of Rider Product at Uber, Head of Product at Instagram, VP Product at Airtable, VP Product at OpenAI, and now General Partner at Felicis Ventures. He was involved in building products used by billions of people and witnessed multiple scale inflections from the inside.

This is a career synthesis conversation. Deng covers the five PM archetypes, his hiring principles including the Six-Month Rule, the importance of growth mindset, how to build growth teams at the right time, what he looks for when investing, and several specific product case studies — including Uber Reserve (now a $5B/year business) and a failed Instagram camera app called Bolt.

The Five PM Archetypes

See Five PM Archetypes for the full framework. In brief:

Consumer PM: Half designer. Excels at intuition, taste, empathy with end users. Instinctively thinks in product experiences.
Growth PM: Half data scientist. Fluent with experimentation, funnel analysis, retention metrics. Optimises for measurable outcomes.
GM/Business PM: Half MBA. Thinks in markets, business models, competitive positioning. Often runs a P&L.
Platform PM: Builds internal tools, APIs, or infrastructure for other builders. The customer is other product teams.
Research/AI PM: Half researcher-engineer. Deep technical fluency alongside product taste. Creates roles that rarely existed before.

Deng argues that understanding your archetype prevents force-fitting: you should optimise for a role that draws on your natural orientation, not contort yourself to fit an archetype that does not match.

The Six-Month Rule

Deng tells every new direct report: ‘In six months, if I’m still telling you what to do, I’ve hired the wrong person.’ The rule establishes an explicit expectation: the person should acquire enough context, conviction, and initiative that they are leading their domain by that point. This is not about ignoring guidance — it is about developing genuine ownership. Deng credits this framing with setting a tone that attracts self-directed people and surfaces mismatches early.

Growth mindset interview technique

Deng’s test: ask a candidate to describe their biggest mistake — the more painful, the better — and explain how it changed the way they think and work. The technique works because authentic vulnerability is hard to fake at volume. It also creates an early relationship moment that normalises fallibility if the candidate joins.

Build the growth team early — and go slow in the 1→100 phase

Deng argues for building a growth team earlier than feels comfortable, specifically because it forces instrumentation discipline on the rest of the organisation. Teams that lack instrumentation at scale discover the gap too late.

The broader principle is that the 1→100 phase requires systems-building rather than moving fast. The zero-to-one phase forgives rough edges; the 1→100 phase punishes them. ‘Go slow to go fast’ — invest in infrastructure, rigor, and process during this phase.

Uber Reserve case study

Deng co-built Uber Reserve at Uber, addressing the insight that riders with early-morning flights do not want to request an on-demand ride — they want certainty. The product gave them a scheduled pickup with a visual peace-of-mind signal (‘this is cutting it close — you may miss your flight’). The interface was deliberately simple; the actual job to be done was confidence, not speed. The product is now a $5B/year business at Uber.

Bolt — the Instagram camera app failure

Instagram built Bolt, a camera-first ephemeral messaging app, with the company’s best design and engineering teams. It launched in New Zealand or Australia, failed to generate asymptoting retention curves, and was shut down. Deng’s takeaway: even the best team with the best product taste cannot reliably predict what will hit. The learning is to treat failure as a lesson rather than a judgment, and to salvage whatever technology transfers back to the core product.

Data flywheels as AI moats

When evaluating AI startups, Deng looks for unique data: proprietary datasets that improve models over time in ways that cannot be replicated by competitors with generic training data. A data flywheel is a structural moat. He also looks for crafted workflows — sequences of decisions and interactions that are differentiated from what the underlying model would produce without domain-specific shaping.

Language and thought

A recurring theme in the episode: Deng believes that language shapes thought, not just expresses it. He applies this to product naming, mission statements, hiring phrases (the Six-Month Rule), and management catchphrases (‘say you’re going to do the thing, say you’re doing the thing, say you did the thing’). Precise language encodes behaviour in ways that imprecise language does not.

Notes

See Peter Deng on the Five PM Archetypes, Scaling from Zero to Billions, and the Six-Month Rule (notes layer) for Four Questions analysis and glossary.