Tom Verrilli, chief product officer at Whatnot, says hiring too many PMs infantilizes engineers and designers who can make good decisions themselves.
Whatnot assigns PMs to problems rather than teams and reshuffles them every six months in planning, with about 21 or 22 PMs in total.
Verrilli argues the top-down shift is about leaders staying in the weeds, noting he and four or five PM managers spend most of their time on IC work.
Whatnot's product chief says most companies hire too many PMs
Whatnot runs product with few PMs and treats the role as a trade learned by doing. Verrilli argues that over-hiring PMs strips engineers and designers of decisions they can make themselves. Senior leaders who stay in the work can replace layers of alignment, and AI adds leverage to individual contributors.
Takeaways
Standard pod ratios over-hire product managers. Scaling companies default to one designer, one PM and one engineering manager per six engineers. Tom Verrilli says this puts too many PMs in place and infantilizes engineers and designers who are perfectly capable of making good decisions.
PMs follow problems, not fixed teams. Whatnot has about 21 or 22 PMs in three loose groups covering buyer, seller, and trust and risk. The CEO, Verrilli and senior leads reassign them to core projects in planning cycles every six months.
Top-down leadership works when leaders know ground truth. Verrilli sees a cultural swing back to top-down leadership, helped but not caused by AI. Half his own time, and over 90 percent for the four or five people managing PMs, goes to IC work.
The hiring bar is extremely selective. Whatnot received 31,832 PM applications over two years and hired one. Verrilli favors macro and micro thinking over talk of alignment and stakeholder management, and is wary of big-company candidates who only polished existing products.
Think through every risk, then move. Verrilli asks what happens if an experiment comes back green or red. Teams need not solve every risk, but thinking through all of them means they solve more than expected. The internal moniker is no, then go.
Averages hide the users who depend on features. Growth can hide all sins at scale when a company works from averages. A feature used by 3 percent of people may be the core use case for a smaller group, so cutting it can break their product.
Agentic commerce will not replace live shopping. E-commerce has never exceeded 20 percent of US retail spend, and US retail is worth $7.5tn, mostly in store. Sellers shape live auctions, and engineers once cut a seven-second timer to three seconds in real time.
What comes next for product leadership
Verrilli predicts the industry will bifurcate between hands-on and managerial product leaders.
He wants Whatnot's product bench to be senior people who build as individual contributors, like the former CTOs of Workday, Instagram, Box and Super.com now engineering at Anthropic.
Paying a few senior people VP-level money is easier than paying many junior layers.
Tap a name to go there · tap a line to see why they are linked
On this episode 2
Show 1
He expects more free space for anyone, whatever their role, to make improvements they are sure they can make.