MMarketing Against The Grain
← All frameworks
Marketing

Four-Context Product Marketing Stack

Ground product messaging in customer, product, company, and market evidence

Difficulty
Easy
Time to result
~days to results
Steps
4
Confidence
96%

The Four-Context Product Marketing Stack assembles four complementary input layers before asking AI to produce messaging. Customer context supplies interviews, call transcripts, research, and other evidence about needs and language. Product context contributes positioning documents, one-pagers, decks, and technical research. Internal context supplies the company’s proven tone, successful posts, and communication patterns. Market context captures the external environment in which buyers compare solutions. The model then uses this combined corpus to generate or improve positioning, sales decks, one-pagers, and talk tracks. The mechanism works because each layer corrects a different failure mode: customer evidence prevents irrelevance, product evidence prevents inaccuracies, internal evidence preserves voice, and market evidence prevents isolated or outdated claims.

Origin

Extracted from Marketing Against The Grain, where Rachel Leist described the materials HubSpot combines to strengthen AI-assisted product marketing.

Core principles

  • 01Strong messaging combines multiple evidence sources
  • 02Customer language should anchor product claims
  • 03Internal voice and external market conditions both matter
  • 04Reusable context improves consistency across deliverables

How to run it

  1. 1

    Assemble customer context

    Collect customer interviews, research, decks, call transcripts, and other materials that reveal customer needs and language.

    Pro tip Include customers from different company sizes, regions, and perspectives.

    Watch out Do not treat a small or homogeneous sample as representative.

  2. 2

    Assemble product context

    Add current one-pagers, positioning documents, product research, sales decks, and other authoritative product materials.

    Pro tip Label outdated documents so the model does not mistake them for current truth.

    Watch out Conflicting product documents can produce inconsistent claims.

  3. 3

    Add internal and market context

    Supply tone-of-voice references, proven internal content, and relevant external market research.

    Pro tip Favor internal examples that have demonstrably performed well.

    Watch out Do not upload sensitive information without appropriate controls.

  4. 4

    Generate channel-specific assets

    Ask the model to turn the combined context into positioning, sales decks, one-pagers, talk tracks, or other required outputs.

    Pro tip Specify the audience, channel, and desired action for each output.

    Watch out Review generated claims and statistics before publication.

In the wild

Launching a new operations product

A product marketer loads customer-call transcripts, the current product brief, successful company posts, and recent category research into a controlled AI workspace. The model uses all four contexts to draft a positioning document and then adapts it into a sales one-pager and talk track.

The resulting assets share one customer-grounded message while remaining appropriate to each channel.

Common mistakes

Using product documents alone

Product materials explain what exists but may not reflect how customers describe their needs or evaluate alternatives.

Mixing stale and current context

Unmarked historical documents can cause the model to revive obsolete personas, claims, or positioning.

Is it for you?

Best for

Product marketing teams creating positioning documents, sales decks, one-pagers, and talk tracks with AI.

Not ideal for

Teams that lack reliable source material or cannot safely place internal information in an AI system.

From the transcript

So, first, all the customer contacts you have interviews, research, decks, and um transcripts with your customers, anything you can get your hands on that…

Rachel Leist · 02:00

And then of course, you want to pull in some internal context, like think about LinkedIn posts that I've done really well, or just tone…

Rachel Leist · 02:30

From the episode

The AI Stack That Makes Our Product Marketing 10x Faster