This CPO regrets that product management exists | Tom Verrilli (CPO of Whatnot)

Episode Summary

Executive Summary: Tom Verilli argues that product management should be a targeted leverage function, not a default layer in every team. At Whatnot, PMs are fewer, more senior, and closer to the work, while engineers, designers, and leaders are expected to build product judgment. AI amplifies this shift by making data, code, and analysis more accessible, but the deeper change is cultural: fewer meetings, more truth-seeking, stronger systems thinking, and faster decision-making.

Main Topics: Regret Product Management Exists (Priority: 5/5): Tom explains Whatnot’s provocative stance: PMs should only exist where they add specific leverage, because over-hiring PMs can weaken the product muscles of engineers and designers. IC-Heavy Product Leadership (Priority: 5/5): He describes a model where senior PMs and even CPOs stay hands-on in research, data, specs, reviews, and decision-making instead of managing layers of org chart. What Makes a Great PM Now (Priority: 5/5): He says the best PMs combine macro and micro thinking, customer understanding, technical fluency, and the ability to make and defend decisions—not politics or stakeholder theater. AI as a Force Multiplier (Priority: 4/5): AI is changing the operating model by making analysis, coding, product understanding, and rapid experimentation cheaper and more accessible to ICs and leaders. Top-Down, Truth-Seeking Culture (Priority: 4/5): Tom argues that strong founder/CPO involvement is beneficial when it is grounded in real data and direct context, and that reviews should be about finding truth, not getting a yes. Product Discovery, Live Commerce, and Market Design (Priority: 3/5): Using Whatnot as an example, he discusses how live commerce economics differ from traditional e-commerce and why product decisions must account for seller and buyer tradeoffs. Career Advice and Product Hiring (Priority: 4/5): He advises aspiring PMs to build IC muscles in their current role, seek fewer but stronger responsibilities, and expect a bifurcation between managers and builders.

Key Arguments: PM is a trade, not a qualification: it should be deployed where specific need exists, not as a default ratio attached to every team. Hiring too many PMs can infantilize engineers and designers by preventing them from developing product judgment through reps. The best PMs today are those who can think broadly about systems and also get into the weeds quickly to validate assumptions and make decisions. Political PM behavior—alignment meetings, stakeholder management, and keeping leadership updated for its own sake—is declining in value. Senior product leaders should remain ICs for as long as possible because hands-on judgment creates better decisions, more leverage, and faster execution. AI reduces the cost of analysis, scoping, code understanding, and experimentation, enabling smaller teams and more autonomous ICs. Top-down leadership can outperform bottom-up org design when leadership truly understands ground truth and can use it to remove friction. The best product reviews are truth-seeking sessions where leaders and teams inspect the code, data, and customer reality together. Whatnot’s culture emphasizes moving PMs between problems rather than permanently mapping them to teams, which keeps skills sharp and prevents over-specialization. Average metrics are often misleading; PMs must look at specific user cohorts and edge cases because small percentages can still represent critical use cases.

Data Points: PM applicants at Whatnot in two years: 31,830 - Tom said this many people applied for PM roles at Whatnot over the last two years. PM hires at Whatnot in that period: 1 - He contrasted the huge applicant pool with very selective hiring. Current PM headcount: 21-22 PMs - He said Whatnot has a little over 20 PMs in the building. Planning cadence: every 6 months - Leadership defines company priorities and DRIs in semiannual planning cycles. Time senior PMs/managers spend on IC work: 90%+ - He said managers on the team spend most of their time doing individual-contributor product work. Tom’s own IC time: ~50% - He said he personally still spends about half his time doing IC work. Time to scope old data science work vs AI: 1-2 weeks - He compared prior Amazon-level data science turnaround to what AI tools can now do quickly. Twitter product leadership turnover: 9 heads of product in 2 years - He cited this as evidence of chaos and weak leadership during his Twitter tenure. Retail share via e-commerce in the U.S.: under 20% - He used this to argue that most commerce remains physical/low-intent despite e-commerce growth. Whatnot stream attendance threshold at Twitch: 1,000 viewers - He said Twitch economics treated streams below this as non-economic because of CPM economics. Example live auction timer change: 7 seconds to 3 seconds - He described a real-time seller request that engineers changed on the spot.

Pivotal Quotes: "We regret that product management exists." — Tom Verilli: Describing Whatnot’s product philosophy: PMs should exist only where they’re specifically needed. "Hiring so many PMs infantilizes the engineers and the designers who are perfectly capable of making good decisions but just never had to because there was always a PM to baby system them." — Tom Verilli: Explaining why excessive PM staffing can weaken product judgment across the org. "The point of the analogy of play the accordion is: you've got to constantly be zooming out and pushing back in, zooming out and pushing back in." — Tom Verilli: His model for product thinking: alternate between broad strategy and deep execution to avoid local optimization.

Implications: Teams should reward truth-seeking, systems thinking, and product judgment over titles and coordination theater. Expect fewer, stronger PMs; more IC-style leadership; and broader use of AI to compress analysis and speed decisions.