We came for the swag. Pendo delivered deeper thinking.
The use case for Product Marketing
Not what can AI do now, but what does quality even mean when anyone can build anything?
Todd Olson, CEO of Pendo, opened with the provocation that’s been sitting with me since: Do you want your tax calculations vibe-coded?
This landed because it’s not hypothetical. This is the world we’re living in now.
The bottleneck moved. But have we?
For years, the constraint was engineering. Bandwidth, roadmaps, tradeoffs. Those were real limits and they gave everyone cover. That cover is completely gone.
When software is easier to produce and faster to ship, the bottleneck shifts to judgment. Prioritization. Knowing your user well enough to say no is becoming the new scarce asset — and it’s exactly what product marketing is made of.
The teams that feel most exposed right now aren’t the ones moving slowly. They’re the ones that were hiding behind the constraints. Which also means the product didn’t get easier. It got even more exposed.
When competitors can spin up features overnight, “we have X capability” stops being a differentiator before the ink dries on your launch brief. If we’re just packaging whatever ships, we’re not positioning anymore — we’re narrating. And you bet your customers can feel the difference.
Taste is a business skill now
aste is a business skill now
Todd named it directly: the difference between great AI products and fast-built noise isn’t the tool. It’s what the team brings to it. Taste. Originality. The willingness to decide what doesn’t belong.
This one is a job description for great Product Marketer.
In a product-led motion, the product is the pitch. It runs ahead of every human on your team — no rep to contextualize it, no SE to smooth it over, no CSM to catch the fumble. A product marketer’s taste gets baked in before the user ever arrives. Which means the edit happens upstream, or it doesn’t happen at all. In a self-serve motion, bad first impressions don’t get second chances.
Sarah Guo put it better than I could: “Real taste hurts. It’s saying no to features that could triple your TAM. It’s spending a week perfecting an interaction that users will barely notice, consciously. It’s choosing the harder technical path because the UX is 10% better. If your taste doesn’t cost you something, it’s not taste. It’s a preference.”
Taste is the edit. It’s the brief that kills the message that isn’t ready. The use case that gets pulled because the proof isn’t there yet. The customer story that stays in the drawer until it’s actually earned.
In a world of infinite output, discernment is the strategy. That’s not a soft skill. That’s the job.
Trust doesn’t transfer. It must to be earned at every handoff.
TThis was the through-line of the whole keynote, and Chris Butler from GitHub sharpened it from a different angle.
AI has introduced new layers: product to model, model to output, output to user decision. Trust has to hold at each one. It doesn’t scale automatically because you shipped a feature or wrote a launch brief that said “intelligent” and “agentic” enough times.
In a product-led company, trust is the conversion mechanism. Most users don’t talk to sales first. They touch the product, form a judgment, and decide. The gap between what’s promised in the market and what’s actually experienced in the product is the gap we need to close. When those are misaligned, even slightly — trust erodes at every touchpoint: the ad, the onboarding, the first “aha” moment that never comes.
You don’t earn trust with a launch. You earn it with consistency across the whole motion.
Butler’s framing hit harder for me personally: we document outcomes, not decision-making. Which means when something breaks, no one can follow the trail back to why it was built that way. For PMM, that’s a positioning liability. If you can’t explain the why behind what you’re putting in market, you don’t have messaging — you have copy. And if you’re anything like me, you’re probably sitting on mounds of it with no decision trail underneath.
The fix isn’t better documentation, though that seems like the obvious answer. It’s how you document. Butler suggested building a decision trail through every launch, every repositioning, every battlecard with each one carrying a short answer to: why did we decide this, and what would have to change for us to decide differently?
The source of truth is still the whole game
Todd circled back to the mainframe — not out of nostalgia, but as foundation. Every agentic system, every AI layer, every automated workflow has to connect back to where truth actually lives. If the system of record is shaky or disconnected, everything built on top of it is just confident guessing.
For a product marketer, that question cuts two ways.
Do you actually know your customer well enough to stake a claim? Or are you positioning to a user that’s shifted while you weren’t watching? Do you know the product well enough to make a promise that holds up under pressure — not softened to stay safe, but actually defensible when a skeptical customer pushes back?
Confident guessing as a product marketer looks like messaging that sounds right but can’t survive a tough customer conversation. The drift between what the product actually does and what your messaging claims it does is its own version of this problem. And most teams haven’t closed it. Butler has ideas on that… but I don’t want to spoil it.*
Getting from copilot to autopilot — and what that demands of us
The trajectory Todd named is already visible: we’re moving from AI-assisted to AI-driven. Less here’s a suggestion, more this just happened. This is where we are now.
Chris Butler’s warning was the necessary counterweight. Consensus is arriving too fast. Groupthink sinks ships. The absence of dissent isn’t alignment, and it’s a signal that something’s being suppressed. His call to action was practical: build an adversarial voice into the process. Before any launch brief, battlecard, or repositioning goes final, route it through someone whose job is to say, “This doesn’t hold.” Even if that someone is a prompt.
For a product marketer, this is a governance question in disguise. When the market accelerates, teams start chasing reactive battlecards, reactive launches, reactive positioning. More output, less signal. Autopilot without guardrails looks like messaging that technically covers the feature but holds no line. No point of view. No this is what we stand for and what we don’t.
Speed without positioning control is just louder noise.
A question is worth leading with
What actually belongs in market right now?
Not everything that shipped. Not everything in the roadmap. What has earned the right to be said — because the product delivers it, the customer validates it, and you can defend it when the room gets skeptical?
That’s the brief. That’s my job to be done. It always was. Every launch decision is a trust decision. What we put in market shapes what customers expect, what sales promises, and what the product has to deliver. Misalign any of those and you’ve done the competitor’s work for them.
The difference now is that the cost of getting it wrong — and the speed at which wrong compounds just went up significantly.
The teams that win in PLG aren’t the ones who launch the most. They’re the ones who’ve built enough trust across product, sales, and customers to know what not to launch.
So what do you actually do with this?
It sounds like a lot. And if you’re anything like me, the to-do list is already longer than the day. So start with one thing. And no surprise … all of it comes back to the source of truth. Start there.
Chris Butler had the most practical answer in the room:
Put your key messaging documents next to the product requirements document. Product marketers should sit at the intersection of product truth and market claims as researchers, influencers, and messaging owners. That position only means something if the message is still accurate. Does what you’re saying in the market reflect what’s actually being built today? If those two documents have diverged, and they usually have, that’s your source of truth problem made visible. Honestly, when was the last time you reviewed the PRD?
Build in an adversarial voice into your next launch. Before the brief goes final, route it through someone whose job is to say this doesn’t hold. No budget for a red team? Run it through a prompt. Ask Claude to steelman the competitor’s response to your positioning. You’ll find the gaps faster than any internal review will.
Document the why, not just the what. The next time you ship a battlecard, a launch, or a repositioning, write three sentences on why you decided this and what would have to change for you to decide differently. It takes five minutes. It becomes the decision trail you’ll wish you had six months from now.
Audit one weekly sync for Kairos potential. Butler’s term for event-triggered over calendar-driven. Which standing meeting exists because it’s Wednesday, not because something actually diverged? Convert one to trigger a competitive shift, customer signal, or product change, and see what that does to the quality of the conversation.
Check your ICP date stamp. When did you last pressure-test your ideal customer profile against actual closed-won data? If your positioning is built on a customer that’s shifted, everything downstream is confident guessing.