Whatnot的CPO说AI时代PM最该练的是数据分析，还吐槽很多PM平时都在表演。他这8条管理建议挺反常识，值得看看。
Whatnot CPO Tom Drobny在Lenny's播客中分享，AI给产品经理的最大解锁在数据科学——2017年亚马逊L7数据科学家需一两周的分析，如今几分钟就能完成。他认为产品经理是练出来的手艺而非资格，并批评过去五年行业奖励的“产品剧场”正在被AI揭穿。Drobny要求每个产品评审必须回答“实验变绿怎么办、变红怎么办”，称系统思维是上升最快的PM技能。他还透露，Whatnot的PM经理有90%时间在做IC工作，反对把优秀PM提拔成脱离实战的管理者。
My biggest takeaways from @tdrobbo (CPO at @Whatnot): 1. AI’s biggest unlock for PMs now is in data...
My biggest takeaways from @tdrobbo (CPO at @Whatnot ): 1. AI’s biggest unlock for PMs now is in data science (not prototyping). Cohort retention pulls, analyzing user logs, and sensitivity and regression models on a @_hex_tech thread—work that took an Amazon L7 data scientist a week or two in 2017 can be done in minutes. Tom spends 10x more time in data than at any point in his career and less time talking to data scientists than ever. 2. Tom “regrets that product management exists,” but this isn't an insult to PMs. Tom argues that PM is a trade, not a qualification—a muscle built by reps. Hiring too many PMs as a default takes reps away from engineers and designers, infantilizing people perfectly capable of good decisions who never had to make them because “there was always a PM to babysit them.” 3. AI is exposing rampant “product theater.” Stakeholder management, alignment, storytelling—real work, but not product work, and they’re what the industry has rewarded PMs for the past five years. But theater only survives where real work is expensive to verify, and it isn't anymore. Every Whatnot hire is given a problem and a dataset and defends a POV out loud — and you watch how quickly the thinking decays in people who can present but can't think clearly. 4. The PM skill trending up fastest: systems thinking. Tom asks this question in every product review: “What do we do if the experiment is green? What do we do if it’s red?” If the answer is “I don’t know,” you haven't done the mental work necessary to run this experiment. 5. Stop promoting your A-players out of the work they’re great at. The common trap: take an A-player PM, make them a director, tell them to stop being hands-on, watch their instincts atrophy while junior PMs yo-yo the same PRD through review. At Whatnot, the four or five PM managers each spend at least 90% of their time on IC work; Tom spends about half his time as CPO as an IC. 6. “It’s really complex” is usually code for weak leadership. Twitter ran working groups on lifting the 140-character limit for Tom’s entire tenure. Everyone knew it was inevitable—Japanese users tweeted six times as often because kanji packs more meaning per character—but nobody would make the call. It shipped almost two years after he left, and nothing bad happened. When something seems complicated, ask whether it’s actually complicated or whether leadership just won’t decide. 7. Map PMs to problems, not teams. Every six months, Tom, the CEO, and senior leads define the company strategy and key pillars, then assign DRIs by skill set. Some engineering teams go a year without a PM. The result: a PM who’s worked on buyer, seller, and trust problems instead of spending three years on notifications. More reps, more muscle groups. 8. Averages mean nothing to the individual. This is the thread running through Tom’s worst decisions. A feature only 3% of users touch looks like a deprecation candidate—until you go a layer deeper and find that for some users it’s 100% of their use case, and killing it is like “a Westfield mall turning off the power in the lead-up to Christmas.” Always ask what distribution hides under the average, and when the anecdote conflicts with the data, trust the anecdote. 9. True PMF is something you feel before you see it in the data. Twitter had nine heads of product in the two years Tom was there and survived anyway, because the product had a level of PMF that absorbed extraordinary mismanagement. Elon changed the name, brand, team, and culture; the network effects held. The career lesson: work somewhere with true PMF early so you learn what bottled lightning actually feels like. 10. End turf wars by giving one person both sides of the tradeoff. At Twitch, discovery and ads were perpetually at war for impressions. One of Tom’s first moves was to put ads inside discovery under a single PM accountable for total GMV from the feed, organic and paid alike. The tradeoff gets made naturally, cutting months of back-and-forth and the politics that turn decisions from company-first to career-first. Lenny Rachitsky @lennysan The CPO who regrets that product management exists Tom Verrilli ( @tdrobbo ) is Chief Product Officer at @Whatnot and a long-time product leader at @Twitch and @Twitter . We discuss: 🔸 How AI is exposing rampant “product theater” 🔸 The PM skill trending up fastest (hint: systems thinking) 🔸 Why AI’s biggest unlock for PMs is now in data science, not prototyping 🔸 Why you need to stop promoting your A-players out of the work they’re great at 🔸 Why “hire great people and get out of their way” is wrong Full episode for your enjoyment 👇 Your browser does not support the video tag. 🔗 View on Twitter 🔗 View Quoted Tweet 💬 7 🔄 2 ❤️ 28 👀 7910 📊 11 ⚡