Zapier's CEO Wade Foster clarifies that you don't want everyone to be "transformative." These individuals are system-builders, constantly re-engineering workflows. A team needs a majority of "adoptive" PMs who are highly skilled but focused on shipping products, not just tweaking their process.
AI acts as an amplifier. According to Zapier's CEO, employees with poor judgment who use AI become "slop cannons," generating high-volume, low-quality work that burdens others. Conversely, those with good judgment become "turbo brains," achieving massive leverage.
Wade Foster offers a simple benchmark for professionals: you should be more advanced in using AI for your specific discipline than your CEO is. If your CEO is spotting obvious AI-driven opportunities or workflow improvements that you're missing, it's a major red flag about your own fluency.
A text-based Product Requirements Document (PRD) is no longer sufficient. Zapier's CEO Wade Foster expects PMs to use AI coding agents to build a functional prototype. Seeing is believing, and a working model provides vastly more insight than a document, dramatically accelerating the feedback loop.
The primary user of SaaS is shifting from humans to AI agents. Zapier's CEO argues that successful platforms must evolve to be "headless," exposing their core capabilities via APIs for agents to build upon. Relying on a human-centric UI is a strategic dead end in the B2B space.
Transformative product management isn't about writing specs or managing backlogs. It's about designing an automated system—a "product factory"—of AI agents that handles the entire development pipeline from customer signal to code draft. The PM's role becomes auditing and improving this system when it fails.
A product leader at Zapier moved beyond individual "second brains" to create a centralized "company brain." This system aggregates company strategy, design systems, and project context for AI agents to access. This created a step-change in productivity, putting a "jetpack" on every product person's back.
For robust personal automation, pure agentic loops are inefficient and unreliable. Zapier's CEO structures his agents to use deterministic code for predictable tasks like fetching calendar events, which is cheaper and more reliable. AI reasoning is reserved only for specific steps like summarization, creating a powerful hybrid.
As AI generates more output, the risk of "AI slop"—low-quality, unverified work—increases. Zapier's internal mantra addresses this head-on. It frames AI as a tool for delegation, but emphasizes that the human user remains fully accountable for the final quality, judgment, and accuracy of the work product.
Before AI, validating a product idea by gathering customer evidence across various sources (Gong, Zendesk, Reddit) could take weeks and required buy-in. Now, a PM can use AI to sift through mountains of data and build a prototype in hours, allowing them to go much further before needing approval.
While a morning brief agent is a good starting point, Wade Foster finds his "Scribe" agent, which runs at the end of the day, far more valuable. It synthesizes outstanding emails, Slack messages, and meeting transcripts to identify action items and draft follow-ups, reducing a two-hour daily cleanup to just 15 minutes.
Humans often bring "pet items" to meetings, avoiding difficult topics. Wade Foster uses an AI agent to scan metrics, meeting transcripts, and internal comms to generate agendas. The AI emotionlessly flags critical issues, like a drop in conversion rates, ensuring the most important topics get addressed.