Half of CEOs say their job stability depends on getting AI right this year. AI strategy has officially become "the CEO mandate". The directive is coming down: build, automate, ship.

So everyone is building. Scroll LinkedIn for 30 seconds and you'll see it: new workflows, new agents, new automations. Our team has been deep in it ourselves.

(In fact, we just ran a workshop showing four AI workflows we built on Klue data.)

But all that building and vibe-coding made me want to pressure-test what's actually happening underneath. So I started doing real CI work in Claude and other AI tools; building battlecard drafts, pulling Gong calls to analyze win-loss themes, the kind of work any PMM or CI leader would actually run.

And things broke...

The scary thing is that it wasn’t obvious! The outputs looked clean. But when I started clicking sources, checking what the system actually pulled, and asking where specific claims came from, cracks appeared fast.

A broken link cited with the same authority as a G2 review. Win-loss themes pulled from calls from 2023 when our positioning has completely shifted. The same competitive question asked two different ways returning completely different answers from different sources.

So, I talked to CI leaders, dug into it with our product team, and decided to share the six ways generic LLMs break when processing data for competitive intelligence. Take a look if you’re in the thick of builds as well.

A CSM tapped me a few weeks ago in the office asking about a new feature that had just gone GA with one of our integrations.

I had no idea it had shipped.

That specific kind of embarrassment, the one where you're the product marketer and someone else on the go-to-market team knows more about your product than you do, is a particular flavour of awful.

I felt like I wasn't doing my job.

And then I thought about it more. Even if I had caught that one, there were probably three other things I'd missed at a different point. There’s an implicit expectation that a PMM should have a live, always-current understanding of every feature, every integration update, every positioning shift (oh, and also be running campaigns, writing content, supporting sales) that is stressful and unrealistic.

It then made me think ‘okay, well what is a better way so I don’t get caught with pants my around my ankles?’. A live, always-on source of product knowledge so that when a CSM asks, the answer is findable based on the TONS of sprints and deadlines our product team are working on.

And then thought continued! What if we had that same logic apply to buyer insights, ICP evolution, messaging, competitor moves. The things that are always changing and always need to be current for marketing teams.

This is where the AI opportunity for PMMs is genuinely exciting. Connecting the relevant live data sources to build the sources of truth we've always had to manually maintain (or store in my short-term-memory-plagued brain). Then you can automate the stressful and mind-numbing parts, and spend your time on the work that actually requires you.

I’m sure I’ll get another jumpscare seeing a Slack ping from a GTM teammate. But, I also see light in the tunnel PMMs are sucked into.

Let’s tackle everyone’s favourite time-consuming task… battlecards! I broked down how you can build these in minutes (and even seconds if you’re speed) based on:

Objection handling grounded in how your best reps actually handle pushback. Winning talk tracks pulled from real deal outcomes. Pricing deep dives based on what buyers have said on calls. Win and loss stories recapping recent deals. And what buyers are genuinely saying about your competitors in their words, not yours.

The blog walks through how to build each of these, and how to get to something you'd actually trust handing to an AE.

(Yes that's me in the thumbnail. No I will not be taking questions about the pose.)

Snap renamed Spotlight to "Reals" for April Fools — a one-word jab at the decade-long history of Meta copying everything Snap ever built. Snap invented Stories. Instagram took it. TikTok ran with it. One word, no explanation needed.

Google closed its $32B acquisition of Wiz — the biggest acquisition in Google's history and a direct shot at Microsoft Azure and AWS in cloud security. Wiz was trusted by 50% of the Fortune 100 before the deal. Now it's Google's.

Zillow launched pre-market listings that let sellers preview their home before going public — and immediately turned it into a recruiting weapon for partner brokerages. The agents not on the list are already asking their brokers why they're on the outside looking in.

Brown-Forman and Pernod Ricard confirmed they're in talks to combine — Jack Daniel's meets Absolut and Malibu in one portfolio, a direct response to slumping alcohol sales and a market being reshaped by tariffs.

Stay caffeinated,

– Adam & Hen (don't be a stranger, connect with me on LinkedIn 👋)