Why would an account executive wait for a product marketing manager to create a competitive battlecard when they can go to Claude or ChatGPT and have it done in seconds?

I keep hearing this question. And it sounds reasonable. The old model was slow. Product marketing creates the materials, distributes them to sales, sales waits. Now the tools exist for anyone to generate their own. Why not let them?

Because product marketing is not about creating content. It has always been about leadership.

The decentralization trap

When you create a battlecard, you are not just creating a document comparing product A with product B. You are creating a conceptual framework that guides the sales team's messaging and positioning. You are teaching them how to talk about your product in a way that is consistent with your brand. You are teaching them how to represent your product opposite a competitor's.

If you let everyone in your company create their own battlecards, you end up with a messaging mess that is difficult to clean up and a sales process mess where there is no rigor or consistency across the organization.

An AE who asks ChatGPT "give me a battlecard comparing us to Competitor X" will get something that looks professional. It will have columns and bullet points and a clean structure. It will also be wrong in ways that are hard to spot. It will frame your product using generic language the LLM pulled from your website and your competitor's website. It will miss your actual differentiators. It will miss the objections your team has learned to handle through hundreds of sales calls. It will sound right and be wrong.

Multiply that by twenty AEs and you have twenty slightly different stories about your product in the market. Prospects notice. They always notice.

The old battlecard was too small

The traditional battlecard was constrained by format. One page, front and back. Maybe a PDF. You had to compress everything into a scannable layout an AE could glance at during a call.

That constraint made sense when the AE was the one reading the document in real time. You needed it to be short because a human had to find the right talking point in seconds.

But what if the AE is not reading the document directly? What if the AE pastes the document into an LLM and asks questions of it?

The format constraint disappears. The document can be as long and as detailed as it needs to be.

Recentralization, with AI

My latest experiment: multi-page battlecards instead of one-pagers. These are detailed reference documents, sometimes fifteen or twenty pages, with:

A tl;dr up top for human consumption that summarizes the key differentiators and messaging to ground any customer conversation
Feature-by-feature comparisons with context about why each difference matters, not just checkboxes
Objection handling with the exact language we have tested in real calls, organized by persona and deal stage
Competitor positioning analysis explaining how the competitor frames themselves and where that framing breaks down
Customer proof points matched to specific objections, with anonymized deal context
Honest assessments of our weaknesses and how to address them without dodging the question
Messaging guardrails that tell the AE how we want to talk about this competitor and what language to avoid

These documents are carefully constructed to support our positioning and messaging frameworks. They talk about our product the way we want it talked about. They frame the competition the way we want it framed. They truthfully assess our strengths and weaknesses.

The AE copies the entire document into their LLM chat window and asks specific questions. "The prospect just told me they are evaluating Competitor X because of feature Y. What should I say?" The LLM gives an answer grounded in the document. Not a hallucination pulled from scraped web content. An answer based on our actual positioning, our tested objection handling, our real competitive intelligence.

The AE can push back, too. "The prospect is a platform engineering lead, not a developer. Reframe that answer for their priorities." The LLM adapts, but it adapts within the guardrails of the document.

The AI reality

By now, we have all become conditioned to having every answer just one question in ChatGPT or Claude away.

Asking people to read and synthesize long documents (or perhaps even short documents) is a productivity tax. The information should be there when it's needed. But most of the time, let's face it, it's not. You need to answer a customer email quickly, and just want to know how to respond.

As marketers, we always have to know our customer. In sales enablement, our first order customer is the sales team.

What changes in practice

Three things happen when you move to this model.

First, the product marketing work gets better. When you know the document will be interrogated by an LLM on behalf of an AE, you write differently. You are more precise. You include more context. You anticipate more questions. The document becomes a knowledge base, not a cheat sheet.

Second, AEs stop freelancing their messaging. They still have full autonomy in how they run conversations. But the source material is centralized. Every AE working from the same document tells a consistent story, even when they ask different questions of it.

Third, you can update once and propagate everywhere. When the competitor ships a new feature, you update the reference document. Every AE using it gets the updated intelligence the next time they paste it into their LLM. No more chasing people to replace the old PDF.

The principle underneath

This is centralized strategy with decentralized execution. Product marketing owns the positioning, the messaging frameworks, the competitive intelligence, the quality bar. Those come from the center. The AEs use AI to interact with that centralized material in ways that are specific to their deals, their prospects, their conversations.

It's an experiment I'm running to balance the nature of the world today (infinite answers at our fingertips) with the needs of the business (consistency and rigor in messaging and go-to-market.)

Sales enablement with AI