Exporting Your Brand as JSON-LD
A PDF brand guide is for humans. JSON-LD is for machines — and increasingly, the AI tools your team uses every day are machines. Markolé exports your brand in JSON-LD so any AI tool can read your brand identity natively, without having to interpret a document.
For a live connection between an agent and your brand, see Markolé MCP: Connecting AI Agents Directly to Your Brand. JSON-LD is the snapshot equivalent — a file you can drop into any tool, anywhere.
1. Why JSON-LD
When you hand a brand guide to an LLM as a PDF, it has to:
- Parse the document — what’s a heading, what’s a caption, what’s the actual rule?
- Interpret what each part means — is this the mission or the vision? Are these core values or aspirational ones?
- Prioritize what matters for the task at hand.
Every step is a place where meaning gets lost. JSON-LD removes all three steps: each piece of brand content is already labeled with what it is, so the AI doesn’t have to guess.
We built our export on JSON-LD with Schema.org as the base vocabulary, and we publish the schema openly at markole.com/schema/brand-dna.
2. What’s in the Export
Markolé offers three export scopes, depending on what you need:
- Brand Report — all 17 narratives (vision, mission, values, archetype, personality, tone of voice, positioning, promise, story, manifesto, pitch, keywords, rituals, target audience), 4 meta-narratives, and the full brand summary.
- Visual Identity — color palettes (light and dark), typography, logo URLs in all variants, visual mood, concept statement, imagery guidelines, layout principles, and categorized dos & don’ts.
- Brand Book — both combined.
Every scope produces a single self-contained JSON-LD file you can copy, share, and version.
3. How to Export
- Open the brand you want to export.
- Navigate to the Brand Document panel or to Visual Studio for visual-only exports.
- Look for the Export action — choose the scope you need (Brand Report, Visual Identity, or Brand Book).
- Markolé generates the JSON-LD file and offers it as a download.
The export reflects the current state of your brand. Re-export any time the brand evolves to keep external tools in sync.
4. How to Use the Export
Drop it into ChatGPT, Claude, or Gemini. Attach the file or paste its contents into the conversation. The AI immediately has your complete brand identity — no 500-word system prompt, no copy-pasting fragments.
Wire it into automated content pipelines. The JSON-LD becomes your brand configuration file. Every generated piece inherits the right tone, messaging, and visual references.
Hand it to agencies and freelancers. Instead of asking them to read a 40-page PDF, give them a file their AI tools can ingest directly.
Use it in multi-agent workflows. When multiple AI agents collaborate on content, they all reference the same structured source of truth.
5. JSON-LD vs. MCP — Which to Use When
Both bring your brand into AI tools, but they’re different:
- JSON-LD export is a snapshot. You download a file and bring it where you need. It does not change unless you re-export. Available to all users.
- MCP is a live connection. The agent reads (and, on Pro, writes) your brand in real time. It always sees the latest state and respects locks and team permissions.
In practice, many teams use both: MCP for tools that support it, JSON-LD for everything else.
6. The Open Schema
The schema behind the export is open and freely usable at markole.com/schema/brand-dna. You can:
- Read the full vocabulary documentation.
- Use the schema in your own systems, even without Markolé.
- Build tools that ingest brand identity from any source that uses the schema.
We think structured, machine-readable brand identity is going to become standard infrastructure as AI takes over more of the content layer. The schema is our contribution to that.
7. Tips
- Re-export after significant changes. A new mission statement or a refreshed palette is worth a fresh JSON-LD file.
- Version your exports. Date the filename or store the file in your team’s asset system so you can track when each version was used.
- Pair with MCP where you can. JSON-LD is great for tools that don’t speak MCP yet. As more tools adopt MCP, you’ll lean more on the live connection.