This podcast is a live conversation I hosted for Skip Coach, a community of senior product and tech leaders navigating career transitions together. Members get access to live sessions like this one, with real-time Q&A, before the content is shared publicly. If you’re a product or tech leader looking to maximize your career, apply to join Skip Coach to access this content and other exclusive resources.
Listen on YouTube, Spotify, and Apple Podcasts
The Post-IPO PM Playbook Is Being Rewritten
There’s a lot of conversation right now about the PM function being rebuilt. Most of it comes from two ends of the spectrum: AI-first startups that are slim, flat, and often still finding product-market fit, or big tech companies with thousands of PMs operating at massive scale. But there’s a third category that doesn’t get enough attention: companies that are far older, successful, and have gone from startup to public company in the past few years. They not only need to rewire their teams to take advantage of AI. They also need to navigate being a public company. How is product management different here?
For this session, I sat down with three CPOs living exactly this challenge: Dheerja Kaur, CPO at Hims & Hers; Anneka Gupta, CPO at Rubrik; and Yuhki Yamashita, CPO at Figma. Between them, they oversee product at a healthcare company, a cybersecurity company, and a design platform. Three very different businesses, all recently public, all navigating the same tension: how do you operate with the urgency of a startup when you have hundreds of employees, established customers, and a public market watching?
The questions from the audience mapped to the same anxiety I hear constantly:
“How has the PM role shifted as your company grew from startup to IPO? How do you avoid slowed-down processes internally?”
“What drives AI adoption amongst product, engineering, and design the fastest? What incentives work better than others?”
“What product skills have become more important for ICs, especially at the entry-mid career level?”
What came through is that there’s a third playbook emerging, one that sits adjacent to the big tech model and the AI-first startup model. The insights below are specific to this stage of company, and several of them challenge the advice you’d get from either end of the spectrum.
Going Public Doesn’t Require Replacing Your PMs with “Business People”
“Has going public and the associated financial scrutiny impacted how your teams think about viability of products?”
The best PMs already had the business gene. Going public just activated it.
There’s a common assumption that IPO-stage companies need to swap out product-minded PMs for operators who think in P&Ls and quarterly targets. All three CPOs said the opposite happened.
At Robinhood, Dheerja watched the company move to a GM model shortly after going public. PMs took on P&L ownership, pricing decisions, and direct accountability to quarterly financials. It changed the flavor of the work, but the strong PMs rose to the occasion. As she put it, the sharp first-principles thinkers who knew the metrics didn’t need to be replaced. They needed a different objective function. Once they had it, they thought bigger, not differently.
Anneka built the same muscle at Rubrik through product QBRs. She gave PMs a template and started the ritual, knowing the first round wouldn’t be great. It wasn’t. But five years later, her team drives business outcomes proactively instead of reacting to customer requests. The transformation happened because she invested in the existing team rather than restaffing it.
At Hims & Hers, Dheerja is building toward the same model. Her conviction is straightforward: when you own products at a public company that are core to the business, you can’t operate with product metrics over here and business metrics over there. There needs to be direct accountability, direct ownership, and direct visibility into how your work moves the top line and the bottom line.
This is specific to this stage. At a startup, there’s no P&L to own. At big tech, GM models are already entrenched. The post-IPO window is where the transformation happens, and the surprising finding is that the existing team can do it if you change what you’re asking of them.
AI Removes the Engineering Bottleneck. So PM Workload Explodes.
“In this world of AI, where PMs are vibe-coding and engineers are finding more time to talk to customers, are we going to see the two roles merging?”
AI doesn’t make PM easier. It shifts the constraint from build capacity to decision-making capacity, and at scale, that means dramatically more work for everyone.
The popular narrative is that AI makes teams more productive and frees up time. Anneka described the opposite. At Rubrik, AI tooling has finally reached the point where it works on enterprise-scale codebases, something she says wasn’t possible six months ago, before the latest generation of models came out. Engineers can now diagnose and resolve bugs in days instead of weeks, ship small customer enhancements that used to take two-week cycles, and prototype solutions to problems that weren’t even on the roadmap. The result: “The amount of work that’s actually going to be there for the team in the coming year is gonna be drastically more because engineering is no longer the bottleneck.”
That’s a sentence worth sitting with. When engineering speeds up, PM workload doesn’t decrease. It compounds. There are more features to position, more output to review, more decisions to make, more surface area to protect. As Anneka’s product leaders take on more commercial responsibility, product strategy work pushes down to directors and ICs. Everyone’s scope expands.
Yuhki framed what this means for how everyone works: “I think everyone’s gonna become a manager in a sense. You’re a manager of agents, you’re reviewing work, making decisions based on work that’s coming back, and farming it out.” Dheerja was blunter: “There’s no such thing as somebody whose entire job is management. That’s just not a thing anymore.”
The management skillset isn’t disappearing. It’s diffusing across the org. Every PM at every level is now building, reviewing, and directing, whether that’s directing people or directing AI tools. The pure people-manager role is fading, replaced by player-coaches who stay close to the details while managing an expanding surface area of output.
This hits scale companies hardest. At a 20-person startup, the founder absorbs the extra throughput. At a public company with hundreds of employees, established processes, and quarterly earnings to hit, the decision-making bottleneck becomes structural. If your PM team isn’t ready for it, the speed gains from AI engineering will create chaos, not leverage.
Systems Thinking, Not Speed, Is What Separates PMs at Scale
“What product skills have become more important for ICs, especially at the entry-mid career level?”
At this stage of company, the hardest skill isn’t shipping fast. It’s understanding how everything fits together.
The AI-era conversation is dominated by velocity. Ship faster, iterate faster, prototype faster. If you listen to leaders at AI-first startups, speed is the primary virtue.
Yuhki pushed back. At Figma’s scale, the critical skill is understanding how changes fit the whole system without creating complexity for users who didn’t grow up with the product. “It’s really easy for teams to think they’re solving for a use case, but then accidentally complicate the entire system.” The systems thinkers, people who simplify as they add, who consider the mental model of a new user encountering the product for the first time, are the ones Figma needs most.
Anneka described the same instinct through the lens of hiring. Her idea for future PM interviews: live pair-coding sessions with Claude. Not to test coding ability, but to test whether candidates are systems thinkers who can break problems into pieces, sequence them, and ask the right questions. “As long as you ask the right question, AI can answer your question for you.” The skill isn’t having answers. It’s having the right questions and knowing how each piece connects to the whole.
This is the insight that doesn’t translate from the startup playbook. When a company with a greenfield product says “move fast,” the downside of complexity is low. When Figma or Rubrik moves fast without systems thinking, they create experiential and technical debt that compounds against an installed base of paying customers, with a public market watching the results every quarter.
Public Companies Are Compressing Planning, Not Killing It
“What has changed most in the way the product team operates since IPO?”
The planning-heavy model isn’t gone. It’s being squeezed into tighter cycles, and that compression is the harder transformation.
AI-first startups love to say “we don’t do roadmaps.” When you have 30 people and a product that’s a year old, that’s liberating. When you have quarterly earnings, board scrutiny, and enterprise contracts, abandoning planning isn’t an option. The interesting finding is that these public companies are making the same directional shift, but the degree of difficulty is dramatically higher.
Figma deprecated annual planning entirely. Yuhki described the transition from “performative AI,” cool demos that may or may not make a difference, to genuine reinvention of how teams work. His framing: if PMs can approach redesigning their own process the way they approach their product work, real productivity gets unlocked. “You’re constantly refactoring the way you work, and that’s just as impactful as the product work that you’re doing.”
Dheerja described what this looks like in practice at Hims & Hers: “You’re starting to see teams build as they make decisions, because the impetus on doing all the upfront thinking and the conversations and looking at figs just isn’t as necessary. Just starting to build and ship and quickly pivoting as you’re feeling a product in your hands.” She connected this to a broader framework: AI transformation at scale has three connected components. AI as core product (not just internal tooling), dramatically faster throughput of production code, and the collapse of decision-making speed because you can build a tangible prototype instead of debating a spec.
Anneka saw it from the enterprise side. Features that were “not even gonna get to talking about for a couple of quarters” are now getting prototyped and pulled forward.
None of them abandoned structure. They compressed it. The quarterly check-ins still exist. The board still expects predictability. The customer contracts still have commitments. But the time between those checkpoints is being filled with faster iteration, more prototyping, and less upfront alignment. That compression, maintaining accountability while dramatically increasing speed, is the transformation that defines this stage.
AI Adoption at Scale Requires the CPO to Go First
“What drives AI adoption amongst product, engineering, and design the fastest?”
The unlock isn’t mandates or training programs. It’s the moment a PM builds something and reconnects with the builder identity that drew them to product in the first place. But at this stage, someone has to go first, and it has to be the CPO.
At a startup, everyone just picks up new tools. At a public company with hundreds of employees and established processes, the adoption challenge is organizational, not technical. Everyone is busy. No one has bandwidth. Learning AI feels like a side hustle on top of an already overwhelming job.
All three CPOs pushed back on the idea that upskilling happens on evenings and weekends. Dheerja compared it to the perennial PM complaint of “I don’t have time for strategy.” Her response: you’re probably doing a bunch of things that are just not that important. Leaders have to be prescriptive: carve out real work time, assign specific projects to be done differently, and give people the space to learn by doing, not by taking a course.
Anneka’s approach was the most vivid. She wrote an accessible Claude Code guide for her PM team because the engineering documentation wasn’t PM-friendly. She gave everyone GitHub access. She personally troubleshoots when people get stuck. And she’s building an agent to help triage enhancement requests, between meetings, in natural language, iterating with Claude. Her take: “This is my job. This is my number one job right now.”
And then the part that makes adoption stick. Anneka observed that when you ask PMs to do more work, it’s usually drudgery. Update this ticket, handle this escalation. But building with AI tools is different: “This is like the most fun thing that I’ve done in years.” That joy is the unlock. The moment a PM builds something themselves and feels the satisfaction of creating rather than coordinating, that’s when adoption becomes self-sustaining.
Dheerja saw the same dynamic from a different angle. The coordination work that used to define PM at scale, getting marketing to build positioning, looping in design on iterations, waiting on other functions to do their part, is collapsing. PMs can now cross those boundaries themselves. But she added a crucial caveat: everyone has to hold hands on the transition. At bigger companies, if leadership doesn’t explicitly signal that blurred functional boundaries are okay, everyone operates with different expectations and the PMs hit roadblocks at every step.
The pattern across all three: AI adoption at scale isn’t about finding time. It’s about making it the priority. Visibly, imperfectly, and at the expense of other things. And the person who has to model that is the CPO.
The CPO’s Perspective
If there’s one theme connecting this conversation, it’s that the product management transformation at recently public companies is harder, and more interesting, than the one at AI-first startups. These leaders aren’t starting from scratch. They’re adapting a running system while the expectations around them accelerate.
If you’re a PM at a public company feeling the shift toward business accountability, first-principles product thinking and P&L ownership aren’t opposed. They’re complementary. The PMs who leaned into the GM model at these companies didn’t need to become different people. They needed a different objective function.
If you’re wondering why AI hasn’t made your job easier, recognize that it’s shifting the bottleneck, not removing it. Engineering speed creates more decisions, more surface area, and more work for everyone on the PM side. Prepare for your scope to expand.
If you’re focused on shipping faster, consider whether you’re also thinking about the system. At scale, speed without systems thinking creates debt that compounds against real customers and real revenue.
If your company is still doing annual planning, you’re not alone, but the compression is coming. Start treating your own process as a product to iterate on, not a fixed ritual to maintain.
If you’re a PM leader trying to drive AI adoption, stop asking your team to learn on evenings and weekends. Make it the core job. Go first. Build something. Be visibly imperfect. The joy of building is the adoption unlock, not mandates, not training programs, not OKRs.
All three CPOs closed with the same conviction: it has never been a more exciting time to be a PM. Building is the best part of the job, and you’re about to get to do a lot more of it. The coordination tax that defined product management at scale is finally being reduced. What replaces it is more building, more deciding, and more ownership.
The companies that figure this out first won’t just ship faster. They’ll attract and retain the PMs who want to do the most interesting work.
Have your own career question? Get personalized guidance at Nikhyl.AI. It’s where the questions keep coming, and where I’ll keep sharing what I’m learning.
