The foundation of AI-Native Marketing Systems

An AI-native marketing system is only as good as what it knows. At Magi, it is based on a knowledge model or marketing ontology. The Marketing Ontology is Magi's structured model to learn about your brand, your market, and your product, and the relationships between them. Every agent, every campaign and every output reasons from this foundational layer.

Why do AI-Native Marketing Systems feel disconnected?

Today, context on Market, Brand and Product are held by different teams, and the connections between them live in people’s heads that gates every piece of work. A marketing ontology helps encode those connections into a single model that every agent and every team member reasons from.

Your AI-native marketing system without an ontology is just a collection of tools.

Why now?

Every production AI application is built on context engineering: the art of filling the context window with just the right information for the next step. For a marketing AI agent, that context is the AI-native marketing system — and the marketing ontology is what that system is built on. Before, marketers reconciled Market, Brand, and Product in their head once per brief. With AI generating 10x the output, that reconciliation has to happen at every generation step. Only a structured ontology, running inside an AI-native marketing system, holds up under that load.

What does Magi’s AI-Native Marketing System comprise of?

Magi's AI-Native Marketing System is built on a marketing ontology — three connected domains: Market, Brand, and Product.

Brand holds voice, visual identity, language rules, and the messaging principles that make a company sound like itself.

BrandOS deep-dive →

Market holds category, ICP, personas, competitors, and influencers — each level inheriting context from the one above.

Product holds positioning, messaging by segment, hero claims, features, and benefits.

The relationships between all three are encoded in the ontology and accessible to every team member and every Magi agent.

What outcomes does our ontology deliver

Marketing execution stays unified and consistent, and holds up at scale across complex organizations of AI agents and human teams.

With Magi, every piece of work already carries your brand’s point of view.

What you stand for. Who you sell to. What you sell.

No one has to brief and prompt the AI from scratch every time.

And the same goes for your team. New hires, agencies, contractors, and freelancers all start from the same playbook. They understand how your marketing works from day one, instead of piecing it together through old Slack threads and scattered docs.

That means your team can move as fast as the AI. Work no longer gets stuck waiting for the one senior person who “knows how we do things” to weigh in.

What's live in Magi today

Magi's AI-Native Marketing System’s foundation is based on three parts: Market, Brand, and Product.

Market

Market defines who you sell to and who else is in the conversation.

Buyer hierarchy: Market → ICP → Persona

Most marketing tools treat the buyer as a flat tag, a persona dropdown, an ICP field. Magi models the buyer as a three-level structured hierarchy, because that's how marketers actually think about who they sell to. ICP is the type of company you sell to within a Market.

Each ICP has its own way of buying. Its own KPIs. Its own pain points. Its own idea of what “good” looks like. It also carries the details that shape how you market to it, like use cases, buying cycle, revenue size, employee count, industry, and country.

Take Revenue Intelligence Platforms as the Market. Inside that Market, you might have two ICPs: Mid-market B2B SaaS and Enterprise SaaS.

Same Market. Very different buyers.

A mid-market SaaS company may move fast, care about quick adoption, and need clear proof of ROI. An enterprise SaaS company may have a longer sales cycle, more stakeholders, and a heavier focus on security, procurement, and internal alignment.

Persona is the specific person inside that ICP. Each persona has their own role, goals, pain points, use cases, and KPIs. Within the Mid-market B2B SaaS ICP, you might sell to a VP of Sales and a Director of Revenue Operations.

They work at the same type of company. But they do not care about the same things. The VP of Sales may care about pipeline, rep performance, and forecast accuracy. The Director of Revenue Operations may care about data quality, process efficiency, reporting, and tool adoption.

Magi understands these layers. So when a marketer creates a brief for a Director of Revenue Operations at a mid-market B2B SaaS company, Magi does not treat that person in isolation.

It automatically pulls in the persona context, the ICP context, and the broader Revenue Intelligence Platforms Market context.

All in one generation step. No extra briefing needed.

How this is used:

When a marketer launches a campaign in Magi, they pick the specific Market → ICP → Persona they're targeting. Magi inherits the full context down the hierarchy, so every output in that campaign is grounded in the right Market, ICP, and Persona without the marketer having to re-explain the buyer.

Competitive landscape: Competitors and Influencers

Adding competitors and influencers into Magi’s marketing ontology turns the live market into structured signal that feeds back into Magi at regular cadence.

Competitors are added with name, description, and the sources Magi should monitor: their website, blog, and LinkedIn company page.

Influencers are added as individuals: analysts, podcast hosts, category thinkers, the people whose POV shapes the conversation in your industry. You add the specific sources to monitor for each one: their website, blog, and LinkedIn profile.

How this is used:

Magi runs multiple agents that do different jobs across the OS. The Magi Research Agent runs continuously against the added sources for competitors, and influencers, bringing up positioning shifts, new commentary, and market signals on a regular cadence. Those signals directly influence Magi's understanding of the market, so content engages with where the conversation actually is, and stays aligned with the campaign strategy and unified execution across the team.

Brand: Who you are and how you show up

Brand defines who you are and how you show up.

Brand holds the context that defines what a company stands for and how it expresses that across every interaction.

Brand Identity captures the name, story, archetype, mission, vision, core values, and key differentiators that make a company distinct.

Visual Identity captures colors, typography, logos, iconography, illustration style, and photography direction. These are the visual decisions that shape how the brand shows up across every output.

Voice & Tone lets you define named voice profiles per person (founder, marketer, salesperson), with a tone library mapping each voice to different contexts.

Messaging holds the positioning statement, taglines, and messaging pillars that anchor every campaign.

Language & Grammar captures the preferred terms, capitalization, and grammar rules your brand follows, applied automatically across every output.

Keywords & Prompts are the search terms and conversational queries your buyers use, organized by persona and topic, for SEO and AEO.

Content Guidelines define content hierarchy, CTA templates, content type prompts, and real content examples Magi can reference as the structural pattern for new outputs.

Social Media holds the platform-specific rules for LinkedIn and X, including character counts, format-level guidance, and hashtag conventions.

Legal & Compliance holds the claim guardrails that get enforced at the point of content creation, not after review.

For a deep walkthrough of how Brand works in Magi, see our BrandOS deep-dive.

How this is used:

Brand sits at the center of every output Magi generates. When a campaign is briefed, Magi pulls the right voice profile, applies the right visual identity, enforces the right language rules, and respects the right compliance guardrails across content, design, social, and email.

Product: What you sell and how it gets framed

Product defines what you sell and how it gets talked about.

Each product is added with fields that define how the product gets framed across every piece of marketing.

Description holds the product overview, the category claim ("the defining leader in X category"), and an explicit list of what the product does not do. The AI knows what to claim and, just as importantly, what never to claim.

Positioning holds the positioning narrative: the problems the product solves, the audiences it serves, and the hero messages that anchor each solution to each audience.

Messaging holds the messaging focus per audience: buyer, end user, partner, regulator, internal team. The same product can be talked about in the right register depending on who's reading.

Features capture what the product delivers, in structural detail.

Benefits capture what the product changes for the reader, framed per audience.

Together, these five fields give Magi the context to translate the product into the right narrative, in the right voice, for the right audience, in every output.

How this is used:

When a marketer launches a campaign, they pick the Product(s) the campaign is about. Magi pulls the product's positioning, messaging, features, and benefits, and uses the audience and register the campaign is configured for to frame every output.

Already on Magi? Open Design in your workspace to get started.

New to Magi? Book a demo to see it configured against your marketing function.