101:43
product-management
Gibson Biddle argues that "product sense" is not a formula but developed judgment, built by running every idea through a simple test — does it delight customers in a hard-to-copy, margin-enhancing way — and by force-ranking priorities (his Growth/Engagement/Monetization "GEM" model), illustrated throughout via real, still-unresolved Netflix decisions (advertising, games, live sports, social features, pricing, account sharing).
Product sense is developing good judgment through debate and experimentation, not knowing 'the answer' in advance — 'no one knows the answers to any of these questions.'
Core evaluative lens for any product idea: does it delight customers in ways that are hard to copy and margin-enhancing.
GEM model: Growth, Engagement, Monetization — a forced-priority ranking because even a 10,000-person company never has unlimited resources; Netflix currently ranks Growth first, Monetization second, Engagement third.
Two-question filter for new initiatives: 'what's the problem we're trying to solve?' and 'is it big enough to matter?'
'Walk' vs 'run' framing (Netflix board-call language): early cautious steps vs. later scaled bets, applied to advertising and games rollout.
'S-curve' framing: is a new bet the next major growth wave, or just an 'enhancer'? Biddle argues advertising was an enhancer, not the vision; games is hoped to be the real next S-curve, expected to take 5-8 years like original content did.
SMT model (Strategy, Metric, Tactic), repeated monthly until the org can 'play it back'; a rolling roadmap is a guide, not a fixed promise.
'Swim lanes': distinct organizational tracks (e.g., viewing experience, advertising, personalization), each led by someone meant to be 'the smartest person in the room' in that area.
The 'right answer' shifts with context: advertising was right for Netflix in 2007 (needed profit), wrong in 2008 (killed once personalization proved a better lever), and right again around the time of the talk (needed to reignite growth after Netflix's first-ever subscriber decline).
Frameworks and models exist to enable intelligent conversations and debates between people, not to generate a definitive answer — 'name it to tame it.'
Product leaders are fundamentally storytellers; the real value of a framework is that it helps you build a compelling story/case (e.g., for why advertising is a good idea).
Qualitative empathy work — asking how a product makes people feel, studying competitors' products, one-on-one conversations — is treated as equally important to quantitative A/B testing.
Writing and giving talks are tools Biddle uses to refine his own thinking through public feedback and challenge.
Company culture functions as an 'operating system' that lets people make good decisions independently without checking with each other.
Price increases and the account-sharing crackdown were executed through deliberate 'overcommunication' (advance email warnings at 6, 3, and 2 months) rather than through an internally-debated ethics conversation.
Interviewing for product sense: ask about a candidate's favorite and most-hated product, ask about creative weekend projects, then apply the delight/hard-to-copy/margin-enhancing lens to their answers.
Delight / hard-to-copy / margin-enhancing model — Biddle's core evaluative lens: a good product idea should delight customers in a way competitors can't easily replicate and that improves the company's margin. Apply: Run any proposed feature or initiative through the question 'does this delight customers in a hard-to-copy, margin-enhancing way?' before greenlighting it.
GEM model (Growth, Engagement, Monetization) — A three-force framework for force-ranking company priorities when resources are limited, used by Netflix to decide where to focus (currently Growth > Monetization > Engagement). Apply: When resources are constrained, explicitly rank Growth, Engagement, and Monetization against each other to decide which initiatives get funded first.
SWAG (stupid/scientific wild-ass guess) — A term for the necessarily uncertain, best-effort prediction a product leader must make about an initiative's impact, ideally upgraded from 'stupid' to 'scientific' with data. Apply: When asked to predict an outcome with no hard data, make an explicit, falsifiable guess (a SWAG) rather than deflecting, and refine it with whatever data becomes available.
Name it to tame it — The practice of coining shared vocabulary and named frameworks so a team can debate a topic productively instead of talking past each other. Apply: Give recurring product debates (e.g., growth vs. engagement tradeoffs) a named framework so the team can reference and argue about it efficiently.
SMT model (Strategy, Metric, Tactic) — A recurring template pairing each strategic investment area with a defined success metric and the tactical projects meant to move it. Apply: For each of a company's key strategic areas, write down the strategy, the metric that will prove it's working, and the current tactics, and repeat/review this monthly until the org can recite it.
Walk vs. Run — Netflix board-call language distinguishing early, basic-step initiatives ('walk') from later, more ambitious, scaled initiatives ('run') within a new business line like advertising. Apply: Classify new-initiative ideas as either near-term 'walk' steps (e.g., basic ad-impression instrumentation) or longer-term 'run' bets (e.g., shoppable ads), and sequence investment accordingly.
S-curve evaluation — A framework for judging whether a new bet (e.g., games) represents the company's next major growth wave, versus being a smaller 'enhancer' to the existing business (as advertising was judged to be). Apply: Ask of any major new initiative whether it's meant to be the next S-curve of growth or merely an enhancer, since that framing should set expectations and timeline (Biddle cites 5-8 years for a true S-curve bet to mature).
Swim lanes — An organizational technique of assigning a distinct focus area (e.g., viewing experience, advertising, personalization) to a dedicated leader who owns strategy, metrics, and tactics for that lane. Apply: Give each major strategic area its own named owner/leader who is expected to be the most knowledgeable person on that topic and who proposes their own quarterly plan for leadership to challenge.
Force-rank prioritization under scarcity — The observation that even a very large company never feels like it has unlimited resources, making explicit forced ranking of priorities (e.g., via GEM) valuable. Apply: When multiple initiatives compete for the same resources, force a rank order among them rather than trying to fund everything equally.
Strategy → Plan → Consumer science → Product sense pipeline — Biddle's general mental model for how companies move from high-level strategy and planning through experimentation ('consumer science') to the judgment calls ('product sense') that fill remaining gaps. Apply: When making a product decision, first check if strategy/plan already answers it, then see if an experiment (A/B test) can answer it, and only then fall back on judgment/product sense.
Two-question initiative filter — A quick evaluation filter Biddle applies to new initiatives: what problem is being solved, and is it big enough to matter. Apply: Before investing in a new initiative, explicitly answer 'what problem are we solving?' and 'is it big enough to matter?' to decide if it's worth pursuing.
Big A/B testing as default validation — Netflix's default practice of validating major product changes (e.g., ad banners, plan-wide ad tests, pricing) through large-scale A/B experiments before wide rollout. Apply: Before rolling out a risky or uncertain change broadly, run a large A/B test to measure real user impact (e.g., on retention) first — except when a competitor has already proven the model, in which case testing may be skipped.
Qualitative feelings-based interview technique — A technique for extracting richer qualitative insight by asking about a person's first experience with a product and following up with feeling-based questions ('how does it make you feel?'). Apply: When gathering qualitative product feedback, ask about someone's first use of a product and probe their emotional response rather than just satisfaction ratings.
Studying competitor products hack — A habit of deliberately spending time using and judging other companies' products to build product sense. Apply: Regularly use competitor and unrelated products firsthand and consciously evaluate what's good or bad about their design choices.
Talk-then-write / write-then-talk loop — Biddle's method of alternating between giving live talks/workshops and writing essays to refine and pressure-test his own ideas. Apply: Work out a new idea by presenting it live for low-stakes feedback, then commit it to writing (and heavily self-edit) to sharpen the argument.
Live 'build on the idea' brainstorming exercise — A facilitation technique where an audience member pitches a product idea and the group collectively builds on and critiques it in real time. Apply: In a workshop setting, have someone pitch a concrete idea, then have the group add to it, challenge it, and cite supporting/contradicting data before evaluating it.
Product-sense interview technique — An interviewing approach that starts with 'what's your favorite product and why,' triangulates into a product the candidate knows deeply, asks about a hated product too, and probes a recent creative project (e.g., 'what did you do last weekend'). Apply: When hiring for product roles, ask candidates about favorite/hated products and a recent creative undertaking, then apply the delight/hard-to-copy/margin-enhancing lens to judge the quality of their reasoning.
"Hacking a Product Leader Career" interview checklist — An essay Biddle wrote describing how he interviews product candidates at every seniority level, functioning almost as a checklist. Apply: Use a leveled, checklist-style set of interview questions and evaluation criteria tailored to the seniority of the product role being hired for.
Overcommunication cadence for unpopular changes — A learned Netflix practice of announcing price increases or policy changes (like account-sharing rules) well in advance, with reminders at roughly 6, 3, and 2 months before implementation. Apply: Before rolling out an unpopular change (e.g., a price increase), send advance notice several months out and follow up with reminder communications closer to the change date.
Customer advisory board / "two in a box" — A B2B product-sense technique pairing a product manager tightly with a key customer contact, relying on deep customer trust for qualitative input since B2B is harder to A/B test. Apply: For B2B products, pair a PM directly with a trusted customer (or use a customer advisory board) to gather qualitative signal on whether a feature will generalize beyond one account.
Faster-experimentation leadership evaluation — Biddle's management approach of pushing leaders to experiment and learn faster, and reassessing whether the right person is in a leadership role if a year or two pass without measurable results against metrics. Apply: Coach direct reports to increase their pace of experimentation, and treat a prolonged lack of measurable progress against agreed metrics as a signal to reevaluate that leader.
Wisdom-of-crowds prioritization poll — A live audience exercise where participants individually rank a set of initiatives (e.g., Netflix's 'walk'/'run' ad ideas) to surface a collective, aggregated judgment. Apply: When facing an ambiguous prioritization call, poll a knowledgeable group individually and aggregate their rankings as a cross-check against internal assumptions.
Advertising's return at Netflix is framed not as a new strategic direction but as a recurring lever whose correctness depends entirely on company context and timing, not on the merits of the idea itself.
The biggest risk Biddle attributes to live sports isn't the ad economics but loss of control: leagues can unilaterally hike licensing prices (cited example: 50% year-over-year), meaning 'you don't control your future.'
The room's own live prioritization exercise ranked personalization highest and original content only third, surprising Biddle because it diverges from Netflix's own public investor narrative that games are the next big thing.
Netflix skipped A/B testing before both its original 2005-ish ad launch and its 2022-era relaunch — the first time because 'they just did it,' the second time because Hulu had already proven ad-tier viability — suggesting pattern-matching off competitors can substitute for direct experimentation.
Reed Hastings reportedly killed Netflix's first ad business in 2008 by asking Biddle 'who's going to be the best in the world at advertising?' (Google) and 'do we need to be?' (no — 'we needed to be best at personalization') — offered as a reusable test for whether to enter an adjacent business.
The account-sharing crackdown is reframed as clarifying an old permissive 'family' sharing model into a stricter 'household' model, rather than as a wholly new restriction.
Netflix's current ad load (~3 minutes/hour vs. linear TV's 8-10 minutes/hour) is kept light largely because the ad tier doesn't yet have enough reach to fully satisfy advertisers, implying the load is likely to grow as the tier scales.
Barry McCarthy (Spotify CFO, ex-Netflix CFO) and Ted Sarandos (Netflix content chief, Spotify board member) are both cited as internally skeptical of podcast investment, which Biddle uses as evidence that podcasting is 'just a tough business' compared to games or advertising.
«nobody knew what streaming was, they thought streaming was streamlining and that sounded bad.»
— 06:20
«inventing the future is wicked hard.»
— 06:53
«why would you put up a lemonade stand in front of a gold mine»
— 11:21
«you don't control your future»
— 11:53
«it's the product stupid, just build a freaking great product.»
— 17:23
«product sense for me is about developing good judgment, that's really all it is.»
— 25:44
«I care a lot about company culture because it sort of creates a operating system for how people make great decisions without talking to each other.»
— 25:48
«That's the main insight of Netflix: all you can eat for one monthly subscription — not very fancy, not too tech-based.»
— 32:44
«Why write a book? Because then I can't edit anymore.»
— 39:48
«Netflix subscribers can have this service with anything they want to watch for seven bucks a month. Is that delightful?»
— 54:54
«No ads for you — no soup for you.»
— 57:18
«I said we shouldn't do advertising for 10 years because it was going to be too complex, but I realized I got it wrong.»
— 59:37
«Customer choice is more important than simplicity.»
— 60:42
«good fights make good marriages»
— 76:28
«as a product leader you are storytellers and that's the real value»
— 77:25
«this idea of product sense is about having a good intuition for what projects are going to move the needle and which won't the answer is never clear and it's really really hard»
— 78:24
«I tested advertising ages ago everybody thought it was going to be horrible and it turned out to be great and most people didn't even notice it»
— 90:34
«communicate bad news really clearly and often»
— 96:43
«it's all about storytelling ... games is a Better Business and advertising is clearly a better business»
— 97:52
«I want each of these product leaders in their swim Lane to be the smartest person in the room in their area.»
— 99:33
«it's not it's like probably order magnitude like three minutes an hour which linear TV is probably at eight or 10 minutes an hour so it's not that bad.»
— 100:33
«all the advertisers want... they'd rather be advertising to all 250 million members at Netflix.»
— 100:50
Reception
No comments are available to gauge audience reception.
A framework-dense, practitioner-oriented talk that uses live, real Netflix decisions (ads, games, live sports, social, pricing, sharing) as running case studies rather than abstract theory, making it a rich source of named prioritization/evaluation frameworks, though most figures and anecdotes are Biddle's own approximations from memory rather than verified data.
101:43