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

Aug 2, 2026 Episode Page ↗
Overview

Tom Verrilli, CPO at Whatnot, discusses the evolving product management role, advocating for fewer, more senior IC PMs and emphasizing core PM skills in an AI-driven world. He shares insights on building high-impact teams and navigating founder relationships.

At a Glance
10 Insights
1h 24m Duration
19 Topics
6 Concepts

Deep Dive Analysis

The Premise: We Regret Product Management Exists

The Evolution and Role of Product Management

Why Specialization Can Underdevelop Skills

When PM Specialization Works and Doesn't

Whatnot's PM Org Structure and Planning

Hiring for PMs: Trending Skills Up and Down

Developing Systems Thinking Skills

The Shift to Senior ICs Doing Core Work

Advice for PMs in a Changing Market

How AI Transforms Work at Whatnot

The Data Scientist Problem and Trending Roles

The Future of Product Teams and Durable PM Skills

Getting the Most from Hired People

Navigating the CPO-Founder Relationship

The "Play the Accordion" Mental Model

Agentic Commerce vs. Live Commerce

Lessons from Twitter

Failure Corner: The Danger of Averages

Lightning Round and Final Thoughts

Product Management as a Trade

Product management is described as a skill developed through practice and repetition ('a muscle built by reps'), rather than just a qualification. This implies that hands-on experience is crucial for PMs to excel.

Infantilization of Engineers/Designers

Over-hiring PMs can prevent engineers and designers from developing their own decision-making muscles. When a PM is always present to 'babysit,' other functions lose opportunities to make product decisions themselves.

No Then Go

A mental model for risk assessment where you proactively think through all potential negative outcomes and scale breaks before proceeding. The goal is to anticipate and preempt problems, allowing for faster, more confident action.

Product Theater

A term referring to PMs who excel at communicating frameworks and strategies to leadership, but lack deep technical or customer understanding, or the ability to make concrete decisions. This often involves focusing on alignment and politics over actual product work.

Play the Accordion

A mental model for product development that balances strategic vision with fast iteration. It involves constantly zooming out to define the overall goal, then zooming in to build the smallest thing to test that understanding, and repeating the cycle based on new learnings.

Agentic Commerce

A trend where AI agents autonomously buy goods and services for users, often for low-intent or programmatic purchases like light bulbs or air filters. This contrasts with live commerce, which focuses on social, high-touch interactions.

?
Why does Whatnot "regret that product management exists"?

Whatnot articulates this to remind themselves not to hire a PM just for the sake of it, but only when there's a specific need, and to empower engineers and designers to make product decisions themselves.

?
What skills are trending down for PMs in today's market?

Skills related to "driving alignment" and "stakeholder management" are trending down, as these often indicate a PM whose specialty is politics rather than technical or customer-oriented work.

?
How can PMs develop systems thinking?

PMs can develop systems thinking by mentally exercising scenarios, such as imagining what would happen with a thousand times more usage or unexpected knock-on effects, before starting to build. This helps anticipate and preempt potential problems.

?
What does "IC work" mean for a PM at Whatnot?

For a PM at Whatnot, IC work involves being hands-on with customer support tickets, pulling and understanding data, querying the codebase, writing product specifications, and running stand-up meetings. It's whatever is required to most effectively ship a product.

?
How has AI transformed data science for PMs?

AI tools like Hex Threads enable PMs to quickly pull nuanced data cohorts, understand individual user behavior, build sensitivity models, and spot regressions. This significantly reduces the need to consult data scientists for initial analysis.

?
Why is "hire great people and get out of their way" not always the best approach?

While hiring great people is crucial, a purely hands-off approach can lead to a lack of ground truth understanding from leadership. Leaders should periodically go deep into the details to verify and guide, making better macro decisions.

?
What is the "two dads problem" in a CPO-founder relationship?

The "two dads problem" refers to the confusion and inefficiency that arises when both a CPO and a product-minded founder are giving conflicting instructions or reviewing the same work, creating an unnecessary extra layer.

?
Why do averages lie in product metrics?

Averages can be misleading because they hide the impact on individual users or specific cohorts. A feature used by only a small percentage of the overall user base might be critical for 100% of that specific group, and deprecating it based on averages can destroy their core use case.

1. Question Experiment Outcomes

Before running an experiment, clearly define what actions you will take if the results are positive (green) or negative (red). This forces thorough strategic thinking and prevents inaction or indecision after data collection.

2. Pre-empt Scale & Failure

Mentally simulate potential problems like a thousand times more usage or unexpected knock-on effects before building. This ’no then go’ approach helps preempt issues and allows for faster, more confident decision-making.

3. Focus on IC Work

As a PM, prioritize individual contributor (IC) work like understanding customer problems, pulling data, querying the codebase, writing specs, and running standups. This keeps you connected to ground truth and enables quicker, more impactful decisions.

4. Avoid PM Politics

De-emphasize ‘driving alignment’ and ‘stakeholder management’ as primary PM skills. Instead, focus on technical understanding, customer orientation, and making concrete decisions to avoid becoming a ‘politics’ specialist.

5. Understand Ground Truth

For leaders, actively engage in the details of how systems work, rather than just delegating. This involves delving into support tickets, data labeling processes, and code logic to make informed macro decisions and strengthen the system.

6. Question Averages in Data

Be wary of relying solely on average metrics, as they can mask critical individual use cases. Always investigate deeper to understand if a small percentage of users depend entirely on a feature before deprecating it, to avoid blowing up core use cases.

7. Calibrate Founder Relationship

When working with product-minded founders, understand their working style for feedback and pushing back. Don’t compete for vision; instead, focus on translating their vision into reality and know when to step back if they are directly involved.

8. Seek Truth, Not Just Yes

When debating with leadership, prioritize seeking the right answer over simply getting a ‘yes.’ Start with curiosity to understand their context, bring data if available, and accept their decision if it’s just two opinions without data.

9. Develop Systems Thinking

Practice mentally playing out scenarios like unexpected scale or negative impacts to develop systems thinking. This helps anticipate problems and unblocks faster movement by preempting risks.

10. Embrace “I’ll Figure It Out”

Adopt a mindset of ‘I’ll figure it out’ for most challenges. This encourages problem-solving and acknowledges that with sufficient effort, most problems are solvable, reducing perceived complexity.

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 babysit them.

Tom Verrilli

We regret that product management exists.

Tom Verrilli

We took all of our A players and then promoted them out of doing things.

Tom Verrilli

Why wouldn't you want Messi playing for your team rather than trying to have the academy coming along all the time?

Tom Verrilli

Averages mean nothing to the individual.

Tom Verrilli

Most of the time you hear it's really complex. It isn't leadership's just weak.

Tom Verrilli

If you truly find product market fit, like if you managed to bottle lightning, doesn't matter how badly you screw up the organization of it. Like it's huge.

Tom Verrilli

I'll figure it out is probably the closest, you know, like it turns out most things are not as hard as people think like, we'll work it out. I'll figure it out.

Tom Verrilli

Whatnot Planning Cycle

Tom Verrilli
  1. Every six months, the CEO, CPO, and other senior leads define what needs to be true for the next six months (outcomes and critical projects).
  2. Go through the defined list and assign a DRI (Directly Responsible Individual) who is accountable for each item.
  3. Allocate PMs to specific problems or core projects based on the DRI assignments, rather than mapping them permanently to teams.
31,832 applications, 1 hired
PM applications to Whatnot Over two years
20-22 PMs
PMs at Whatnot Current number in the building
50%
CPO's IC work time Tom Verrilli's personal IC work
90%+
Manager's IC work time Managers on Whatnot's PM team
Never exceeded 20%
E-commerce share of US retail spend In America
~25%
E-commerce share of UK retail spend Based on a 75/25 split with physical shopping
$7.5 trillion
Retail industry size in US Annual industry size
Less than 1,000 people
Non-economic stream threshold (Twitch) For entertainment streams, based on CPMs
30-50 people
Whatnot stream audience In a typical stream
140 characters
Twitter character limit (original) Before it was lifted
9 heads
Twitter heads of product In two years during Tom Verrilli's tenure
500 (50%)
Batting average goal For success in career/building