Overview
Tom Verilli challenges the assumption that every engineering team needs a dedicated product manager. His argument is not that product management lacks value, but that ratio-based hiring created too many coordination roles, weakened the product judgment of engineers and designers, and rewarded organizational politics over customer understanding and decisive execution. Whatnot instead employs a small group of mostly senior PMs, reallocates them to priority problems during six-month planning cycles, and expects product leaders—including managers and the CPO—to remain deeply involved in individual-contributor work. AI makes this structure more viable by accelerating data analysis, codebase exploration, regression detection, and effort estimation, especially for experienced operators with sound judgment. Yet Verilli rejects blind trust in AI output, dashboards, or delegated teams: leaders must investigate how data is produced, observe customers directly, and move between system-level strategy and implementation detail. He calls this movement “playing the accordion”—continually zooming out to reconsider the system, then compressing back into the smallest useful execution step. The broader conclusion is conditional rather than universal: Whatnot’s model depends on a hands-on, truth-seeking culture, but the underlying shift away from product theater and toward genuine customer, business, and technical understanding is likely relevant across the industry.
Sections
Higher-Order Implications
Synthesis of the organizational implications behind Whatnot’s operating model.
- The argument is less about eliminating product management than redistributing product judgment. A smaller specialist PM function works only when engineers, designers, data scientists, and executives receive enough customer and business context to make product decisions themselves.
- AI compresses the coordination tax that previously justified larger management structures. When senior contributors can interrogate data and code directly, organizational leverage shifts from supervising information flow toward exercising informed judgment.
- Flattening the product organization increases—not decreases—the burden on leaders to verify reality. Fewer layers are effective only when senior people personally inspect the evidence that those layers previously summarized.
- Whatnot’s hiring selectivity reflects a distinction between experience inside product organizations and demonstrated agency over product outcomes. Prestigious tenure and polished communication do not prove that a candidate has repeatedly made and defended consequential decisions.
- The future product team may retain familiar specialist roles while loosening their territorial boundaries. The deeper change is likely to be more permission for any informed contributor to identify and fix valuable problems outside a permanently assigned pod.
Central Tensions
The principal disagreements and trade-offs explored in the interview.
- Whether each engineering team needs an assigned product manager or whether PMs should be deployed only where ambiguity and leverage justify specialization.
- Whether senior PMs create more value through people management or through direct individual-contributor work.
- Whether top-down product leadership is efficient direction or harmful micromanagement.
- Whether AI-driven self-service analysis empowers product teams or generates unreliable work that specialists must clean up.
- Whether agentic commerce will displace human shopping experiences.
Operating Model Comparisons
Explicit contrasts between conventional product organizations and the model described at Whatnot.
- Permanent pod staffing versus dynamic problem staffing.
- Alignment-oriented PMs versus builder-oriented PMs.
- Roadmap-first planning versus iteration-first experimentation.
- Management leverage versus senior-IC leverage.
- Agentic commerce versus live commerce.
Memorable Quotes
Statements that capture the interview’s central operating principles.
- We regret that product management exists.
- The only argument as far as I'm aware for why you would want product management to be a specialist function is really it's a trade, not a qualification.
- We took all of our A players and then promoted them out of doing things.
- You can't make good macro decisions without the micro.
- using an AI tool to find piece of data much like using an AI tool to write code doesn't absolve you of responsibility to make sure that that was good analysis good code.