The Product Podcast · Product School

Shopify VP of Product on How to Build AI-First Products | Glen Coates | E269

June 25, 2025·57 min·3 clips
Shopify's MCP server accidentally turned its developer API docs into a plain-English control panel for non-technical merchants.
Strategy docs take a little heat up front. The cold open sets a practical mood, moving from product leaders writing lofty docs to engineers wondering whether product managers are needed at all. Shopify then becomes the example. Glenn Coates says AI use is now expected across Shopify, not just among engineers. The useful employee looks different. He calls an AI-fluent worker an overclocked human, someone who can simply get a lot more done. From there, the conversation gets operational. Coates argues that product leaders still need enough detail in their heads to understand how the pieces connect. Scale puts a ceiling on that. He says teams somewhere around 100 to 250 engineers start running into the physical limits of one person's ability to understand a domain. That matters because coherence is where big products often break. His main critique is that components can sell well together while failing to work well together. Enterprise products become the warning case. Tools assembled through acquisition can look polished in a brochure, but underneath they carry different assumptions and make life harder for users. Shopify is not held up as perfect. Coates says one thing the company tries to do well is make related tools behave the way a merchant would expect. The ending favors clear calls over committee fog. He contrasts shipping with unresolved tradeoffs between merchants and developers, the kind that can slow a company into sadness. The host keeps it respectful and closes by casting Coates as a product leader using a startup mindset at much larger scale.

As heard by us

A sharp Shopify conversation on AI expectations, product ownership, and keeping scale from turning into committee drag.

Shopify's product culture comes through as unusually direct: AI is treated less like an engineering side project than a company-wide operating expectation.

Read the full review in PlayNext →

Why you'd press play

Want a grounded take on making AI part of product culture, not a slogan?

Read the full recommendation in PlayNext →
Listen to the show on

The audio for this episode is no longer available at the publisher. The show page links above still work.