Lenny's PodcastWhy every company now needs to think and operate like a lab team | Josh Woodward (VP Google Labs)
At a glance
WHAT IT’S REALLY ABOUT
Why companies must adopt lab-style experimentation to win in AI
- Josh Woodward explains why rapid AI capability shifts force companies to operate like labs: continuously testing frontier tools, running many experiments, and converting new technical possibilities into user value before competitors do.
- He describes where strong ideas actually come from—obsessive side projects and surprising demos rather than scheduled design sprints—and introduces an “almost possible” tracking framework for spotting capability phase changes.
- Woodward reframes early product-market fit as an art: watch for human signals like eyes lighting up, iterate through multiple pivots, and stay attached to the problem rather than a specific solution.
- He outlines when and how to kill ideas: teams usually know first, leaders must create safety to voice doubts, and stopping a launch based on weak user pull should be celebrated.
- The conversation covers operating principles for AI-era teams: rolling 6‑month planning windows with ~100‑day milestones, cultivating “explosive endurance,” and building labs that can both innovate independently and successfully transfer wins into the broader company.
IDEAS WORTH REMEMBERING
5 ideasEvery company needs a frontier team that behaves like a lab.
Woodward argues every company needs a dedicated frontier group whose full-time job is to test the latest models/tools, spot capability “phase shifts,” and rapidly prototype before competitors do. The goal isn’t novelty for its own sake—it’s continuously discovering what just became feasible and translating it into user value.
You can’t “microwave” good ideas—design conditions for them to emerge.
Great ideas rarely appear on a calendar via brainstorms or design sprints; they emerge from persistent tinkering by obsessed builders and from leaders staying alert to surprising demos/behaviors. Labs should be designed to let “weird things grow,” then capture and channel those sparks quickly.
Use an “almost possible” list to systematically hunt for breakthroughs.
Google Labs keeps a running inventory of capabilities that are nearly workable (latency, modalities, languages, etc.), organized by future domains (e.g., entertainment, creativity, work). When a capability crosses the threshold from “almost” to “now,” they mobilize a small team immediately.
Early product-market fit is measured in eyes, not dashboards.
In the earliest stages, Woodward looks for visceral reactions (people leaning in, eyes lighting up) rather than retention dashboards. The goal is to detect real human pull before instrumented metrics become meaningful.
Fall in love with the problem; expect 3–5 pivots to get to a hit.
Because it typically takes multiple pivots, teams should commit to the underlying problem rather than a specific UI or feature set. Falling in love with the solution makes pivots feel like failure instead of the expected path.
WORDS WORTH SAVING
5 quotesThey very rarely, at least in my experience, have come from design sprints or times where you set aside on a calendar, "Okay, this is we're gonna go come up with our next idea."
— Josh Woodward
They don't come in conventional places. You can't microwave good ideas.
— Josh Woodward
When you're showing people early prototypes, what I'm always doing in those is you're looking at people's eyes, and, like, that is the metric.
— Josh Woodward
The team usually knows before the leader knows.
— Josh Woodward
I'm actually looking a lot right now for people, what's their unlearning rate? Like, how fast can they learn something and then walk away from it.
— Josh Woodward
High quality AI-generated summary created from speaker-labeled transcript.
