Enrolled Members Only
Unit 2

Messaging & Positioning in the AI Era

50 minutes 5 sections 🧪 Messaging Lab included

Learning Objectives

By the end of this unit, you will:

Section 1: The Two-Audience Problem

~10 min

Writing for Humans and Machines

Every piece of content you create now serves two audiences: humans and machines. This isn't a future state — it's the current reality, and most PMM teams are still writing exclusively for the first audience while ignoring the second.

When you craft a positioning statement, a human reader processes it through emotional resonance, brand affinity, and personal relevance. They might not remember your exact words, but they'll remember how you made them feel. Traditional messaging guidance optimizes for this: be memorable, be differentiated, create an emotional connection.

But an AI reading that same statement is doing something different. It's evaluating semantic content. It's cross-referencing claims against other sources. It's assessing specificity, looking for verifiable details, and deciding whether to cite you in a synthesized response or pass over you for a competitor who made a more concrete claim.

Fails Both Audiences

"Empowering Digital Transformation"

Works for Both

"Teams complete quarterly close 40% faster, with audit findings reduced by 65% in year one."

The vague claim might resonate emotionally with an executive tired of legacy systems. But it gives an AI nothing to work with. No specificity. No verifiable claim. No differentiation from the thousand other vendors who say the same thing.

The Core Challenge

These two optimization targets don't always conflict — but they often do. The 10x PMM solves the two-audience problem by creating messaging that works for both: emotionally resonant for humans, informationally dense for machines.

The Dual-Audience Framework
Figure 2.1: The Dual-Audience Framework — positioning must work for both human buyers and AI agents

Section 2: The Specificity Imperative

~10 min

Vague Claims Vanish

The single biggest shift in messaging strategy for the AI era is the move from vague to specific. Compare these two value propositions:

Vague

"Our platform helps teams work faster and smarter."

Specific

"Teams using our platform complete quarterly close 40% faster, with audit findings reduced by 65% in the first year."

The vague claim is technically true of every software product ever built. It provides no differentiation. When an AI agent is asked to compare accounting platforms, it has nothing to cite.

The specific claim gives the AI something to surface. It includes numbers. It implies verifiable outcomes. It can be cross-referenced. And crucially, it can be used in a comparison: "Platform A cites 40% faster quarterly close; Platform B does not provide comparable metrics."

The Specificity Ladder

Think of specificity as a ladder with five rungs. Most B2B messaging lives at level 2 or 3. The companies winning the AI visibility game operate at level 4 and 5:

1
Abstract

Zero specificity — completely invisible to AI

"We make things better."
2
Categorical

Category-level — AI surfaces but doesn't differentiate

"We improve productivity."
3
Directional

Audience-specific — AI can filter to relevant queries

"We improve productivity for sales teams."
4
Quantitative

Measurable — AI can compare against competitors

"We improve sales productivity by 30%."
5
Evidenced

Verifiable — AI trusts and cites confidently

"We improve sales productivity by 30%, based on a 2025 study of 200 deployments."

The New PMM Skill

This requires a cultural shift in how PMMs work with stakeholders. When a product manager says "our new feature improves performance," your job is to push: "By how much? In what conditions? Can we measure it? Can we publish it?" That negotiation becomes a core PMM competency.

The Specificity Framework
Figure 2.2: The Specificity Framework — moving from vague claims to evidenced positioning

Section 3: Claims That Get Cited

~10 min

The Citation Hierarchy

Not all claims are created equal in the eyes of AI. There's a clear hierarchy that determines what gets cited versus what gets ignored:

Claim TypeAI TreatmentExample
Third-party validatedHighest trust, frequent citation"Named a Leader in Gartner Magic Quadrant"
Customer-attributedHigh trust, cited with attribution"Acme Corp reduced costs 40%"
Published researchMedium-high trust"Based on analysis of 500 deployments"
Quantified internalMedium trust, cited carefully"Average customer sees 30% improvement"
Unquantified claimsLow trust, rarely cited"Industry-leading performance"
SuperlativesIgnored entirely"Best-in-class solution"
The question to ask of every claim: "Would I cite this if I were writing an analyst report?" If not, neither will AI.

Building a Citation Portfolio

Treat your claims like a portfolio. You need a mix:

The narrative is for humans. The proof points are for AI. You need both.

Section 4: The Dual-Optimized Framework

~12 min

Building Messaging That Works Twice

Here's a practical framework for creating messaging that serves both audiences:

Three-Layer Positioning Stack
Figure 2.3: The Three-Layer Positioning Stack — Capability, Evidence, and Architecture layers

1. Start with the Claim

What specific, verifiable outcome can you promise? Not what you do — what happens when customers use you. Push for numbers.

2. Add the Evidence

Where does this claim come from? Customer name? Sample size? Time period? The more specific, the more citable.

3. Wrap in Story

Now add the human element. Who cares about this outcome? Why does it matter? What transformation does it enable?

4. Test Both Paths

Ask Claude or ChatGPT to compare you to competitors. Does your claim surface? Is it cited accurately? Iterate until it does.

Before & After Example

Before: "Our AI-powered platform helps finance teams work more efficiently and reduce errors."

After: "Finance teams using our platform close quarterly books 40% faster. In a 2025 study of 200 customers, average audit findings dropped 65% in year one — with CFOs reporting they finally have time for strategic work instead of spreadsheet wrangling."

Why it works: The "After" version has specific metrics for AI to cite AND emotional resonance ("finally have time for strategic work") for humans to connect with.

Section 5: Testing Your Messaging

~8 min

The AI Visibility Test

Here's a simple test you should run on all key messaging:

  1. Ask an AI to compare: "Compare [Your Product] to [Competitor] for [Use Case]"
  2. Check the output: Are your claims cited? Accurately? With the specificity you intended?
  3. Identify gaps: What did the competitor say that you didn't? What claims got attributed to you that you don't own?
  4. Iterate: Strengthen claims that didn't surface. Add evidence to claims that surfaced weakly.

Warning: The Training Data Gap

AI models are trained on historical data. Your latest messaging might not be in the training set. This means you need to ensure your claims appear in multiple authoritative sources — your website, analyst reports, press coverage, customer reviews — not just internal docs.

Key Takeaways

← Unit 1: Foundations Take Assessment →