Build Zero-to-One in the Dark
Inside a big company, new products die of attention. Ask for less scrutiny, fewer resources, and room to fail.
- Difficulty
- Expert
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 90%
Liu's counterintuitive claim from building multiple billion-dollar businesses inside Facebook: big companies kill new products by loving them to death. Weekly executive check-ins on strategy and progress destroy the iteration process that zero-to-one actually requires. Her deliberate posture is to take minimal resources, stay out of the limelight, and buy the freedom to fail — because attention converts every failed iteration into a case for cancellation.
Origin
Deb Liu, from building Facebook's games/payments business, the first direct-response mobile ads product (a billion-dollar business within ~18 months), the mobile ad network, and Facebook Marketplace (now 1B+ monthly users) — none of which were the company's core products at the time.
Core principles
- 01You can love a new product to death — attention is the enemy, not the reward.
- 02In a startup you fail fast or succeed fast; in a big company the long slog gets you cut.
- 03Zigg where others zag: work on what nobody in the core is competing for.
- 04New products are seeds in a portfolio, not the most important thing — and should be resourced accordingly.
- 05A very successful company's new-product hit rate is about 50%. Plan for the coin flip.
How to run it
- 1
Pick the zag
Deliberately choose the surface nobody else wants. At Facebook the amazing people worked on feed, photos and video; Liu took payments, games, and — while the company was overwhelmingly brand-advertising oriented — direct response. Notably her ads product wasn't even built on the ads team; it was built on the platform team.
Pro tip Being on the 'wrong' team for the product you're building is a feature — it keeps you off the executive review calendar.
- 2
Ask your customers what they actually need
Liu had deep relationships with game companies from her payments years, so she asked them directly what ad product they wanted. Their answer: our biggest challenge is the shift to mobile — build us a mobile acquisition engine. That was the spec.
- 3
Negotiate for less, not more
Explicitly take fewer resources and less visibility in exchange for the freedom to iterate and fail. Liu: with every new product she'd rather do it out of the limelight with minimal resources and have the freedom to fail. Resources buy scrutiny; scrutiny kills iteration.
Pro tip Sell it to leadership as portfolio hygiene and patience, not as secrecy.
Watch out The trade is real — you get attention OR resources, and both come with a review cadence you can't survive at v1.
- 4
Iterate through many versions, expecting near-death
Liu shipped five or six versions of the ads product before it took off; it took months and was on the verge of death multiple times. Prune fast, pivot fast, try the next thing — the point is the iteration loop, not the plan.
Watch out Weekly 'what's your strategy, where are you this week' check-ins are incompatible with this. Push them out to a longer cadence.
- 5
Hedge your own downside
When leadership told Liu the ads product wasn't going to work and she should go run her old team again, she did both — she took the payments job back while the gathered team kept building. She held two jobs until the product took off.
Pro tip Keeping a credible fallback role makes it politically cheap for the company to let the bet keep running.
- 6
Time the big swing in your own career
Do not do this in your first years. Learn the core-product skills first, where growth is stable and moving a metric 5%→10% is a legible resume win. Then, once ready, take two or three years for a big swing — knowing that if it fails you go back to core products with lessons and nothing else to show.
Watch out Expect that in a year you may 'literally have nothing to show for it but the lessons'. If you can't afford that outcome, don't take the swing yet.
In the wild
Facebook was almost entirely brand advertising. Liu, sitting on the platform team rather than the ads team, asked game companies what they needed and heard 'mobile acquisition engine'. She and her team built an app-install ad for the new mobile feed, iterated through five or six versions over months, was told multiple times it wasn't going to work, and took her old payments job back in parallel rather than shut it down.
→ A billion-dollar business within about 18 months, and direct response became the vast majority of the company's ad revenue.
In the new-hire PM class Liu ran, she told PMs plainly: most of you will go into a core product and be asked to grow a metric 3-5% every six months, and that's an excellent career. Those of you who want to build something new should know a very successful company has a ~50% hit rate on new products — half of you will have a different job in a year. Do you have the resilience for that?
→ Self-selection with eyes open, rather than PMs discovering the failure rate after their product is cancelled.
Common mistakes
Assuming big-company zero-to-one is easy because the resources exist
Liu: people think it's linear because you have all the resources of the company behind you. In reality every company optimises for the most important thing, and new products are seeds — you get very little.
Accepting weekly executive scrutiny on a v1
Innovation teams get checked on week-to-week for progress and strategy. That cadence is incompatible with the failure-heavy iteration that building something new actually is.
Taking the big swing too early
Very high risk for someone early in their career who hasn't yet learned the core mechanics. Liu explicitly tells juniors to build core-product skills in a stable environment first.
Is it for you?
Best for
Senior PMs and product leaders inside a large company who want to build a genuinely new business line and have the political capital to negotiate for less oversight.
Not ideal for
Early-career PMs who haven't yet built the core-product skills, and anyone who cannot absorb a year with nothing to show but lessons.
From the transcript
“I think sometimes large companies they say well this Innovation team and then they check in on them way too much”
“with every new product I'd rather kind of do it out of the L light do it with Lim minimal resources and have the freedom…”
“a very very successful company for a new set of products has a 50% hit rate so half of you are going to come back…”
“we actually tested fif like five or six versions of the ads product before we”
From the episode
Succeeding as an introvert, building zero-to-one, and why you should PM your career like you PM your product
Deb Liu (CEO of Ancestry, ex-Facebook, PayPal, eBay)