How to Build a Brand Kit That Makes Your Content Sound and Look Like You

Every marketing team using AI hits the same wall: the output is fast, but it doesn't sound like your brand. Your positioning gets flattened. Your tone drifts. And every tool in your stack starts from zero because none of them know your brand.

AirOps Brand Kit fixes that. It's a single, structured source of truth for your brand -- voice, audiences, content types, writing rules, and visual guidelines -- that any AI tool can access and stay aligned with. Set it up once, and every workflow, agent, and MCP-connected tool pulls from the same governed foundation.

This guide walks you through setting up your Brand Kit and configuring the key elements that make it actually useful -- not just a static doc that collects dust, but a living layer that shapes every piece of content your team creates. We'll use Ramp as our running example to make each step concrete.

By the end, you'll have a Brand Kit configured with content types, audiences, writing rules, and visual guidelines. Finally you'll see it generate a real piece of content and a dashboard that sounds and looks like your brand.

Step 1: Set Up Your Brand Kit in AirOps

Before you can orchestrate your Brand Kit from Claude or any MCP-connected tool, you need to create it in the AirOps app. This takes a few minutes and gives your AI a structured home for everything that defines your brand.

What to do:

1. Navigate to Brand Kits in your AirOps workspace.
2. If you don’t a Brand Kit, click the [+] to create one
3. Review your Foundations: brand name, domain, a short brand summary, your brand’s voice and tone, and the custom rules that govern how your brand shows up in the world. These are the guardrails AI stays within every time it generates content on your behalf.
4. Go deeper than the basics. Add product lines, content types, personas, regions, languages, and visual guidelines. The more specific your brand definition, the more specific -- and recognizable -- the output.
5. Publish your brand kit - to go from changes on a draft to the live version your workflows and MCP will access]

That's the scaffolding. The real value comes from what you layer on top: product lines, content types, audiences, regions, writing rules, and even visual guidelines. Each one gives AI more precise context, so it stops guessing and starts sounding like you.

‍

From here on out, we'll assume your Brand Kit is set up and you're working through the AirOps MCP connector, meaning you can do everything that follows from inside Claude (or any MCP-compatible tool). Every step below is a conversation, not a form.

‍

Step 2: Add a Content Type

Why content types matter

Not all content is created equal, and your AI shouldn't treat it that way. A product landing page has different goals, structure, and CTAs than a LinkedIn post or a technical blog. Content types tell your Brand Kit what kind of thing you're creating so the AI can apply the right template, tone, and format rules automatically.

Without content types, you're relying on prompt instructions every time. With them, the structure is baked in.

What we're building: Use-case Specific Pages

Ramp publishes expense management pages (e.g., "The complete guide to travel and expense management in 2026") that are high-intent, SEO-driven content. We'll set up a content type that captures the format, typical structure, and CTAs these pages use, so the playbook is documented for the wider team to use.

In Claude (via MCP):

Add a new content type to my "[Brand Name]" AirOps Brand Kit

• Name it '[Insert name]'
• Use markdown format for the template outline
• Use this URL as a sample: [your-sample-url.com/test]

The MCP connector will create the content type, auto-extract structure from the sample URL, and attach it to your Brand Kit. Ready for any workflow to reference.

Step 3: Define an Audience

Why audiences matter

Your brand doesn't talk to everyone the same way. The language that resonates with a CFO is different from what lands with a developer or an ops lead. Audience definitions let you scope your writing rules, tone adjustments, and CTAs to the people you're actually trying to reach. In that way, your AI doesn't default to generic "marketing voice."

When you define an audience, you're telling your Brand Kit: when we write for this person, here's what changes.

What we're building: CFO Audience

Ramp sells to finance teams, and CFOs are a primary buyer. We'll create an audience segment that captures what CFOs care about, cost control, visibility, compliance, and how Ramp's messaging should shift when targeting them.

‍

In Claude (via MCP):

Add a new audience to my "[Brand Name]" Brand Kit called '[Audience].'

These are [short description of audience].

Refer to my attached persona document.

Claude creates the audience, scopes it in your Brand Kit, and from now on you can reference your audience as a targeting parameter whenever you generate content.

Step 4: Create a Writing Rule

Why writing rules matter

Voice and tone get you 80% of the way there. Writing rules handle the last 20%. They are the specific, often opinionated guidelines that make your content unmistakably yours. Things like: "Never start a sentence with 'We're excited to announce.'" Or: "Always lead with the customer outcome, not the feature name."

These rules stack. You can scope them globally, or attach them to a specific audience, content type, or region. The AI applies the right rules for the right context automatically.

What we're building: A Ramp writing rule

Ramp's brand voice is direct and confident. We'll add a writing rule that reinforces this: no hedging language, no passive voice, and always lead with the financial impact.

In Claude (via MCP):

"Add the following writing rules to my Brand Kit: '

1. Always use Oxford commas
2. Avoid passive voice and hedging language like "might," "could," or "helps to."Apply these rules globally.'"

Claude adds the rule, assigns it global scope, and it immediately applies to any content generated against this Brand Kit.

Step 5: See It in Action - Generating On-Brand Content

Pulling it all together

You've set up your Brand Kit with a content type (Expense Management Pages), an audience (CFOs), and a writing rule (lead with financial outcomes, no hedging). Now let's see what happens when you actually use it.

We'll take an existing Ramp URL and turn it into a LinkedIn post -- targeting CFOs, using the brand voice and rules you've already configured. No re-explaining your brand. No pasting style guides. Just a prompt.

In Claude (via MCP):

"Using my Brand Kit, take this page, [your URL], and turn it into a LinkedIn post targeting our [Audience Name] audience. Use our brand voice and apply all relevant writing rules."

What happens behind the scenes:

Claude pulls your Brand Kit context via MCP -- voice, tone, the CFO audience definition, your writing rules, and the content format expectations. It reads the source URL and produces a LinkedIn post that:

• Speaks directly to CFOs in authoritative, data-driven language
• Leads with a financial outcome over product features
• Avoids hedging and passive voice
• Matches Ramp's direct, confident brand voice

No prompt engineering required. The Brand Kit did the work.

This is the payoff. Every content type, audience, writing rule, and visual guideline you add makes your Brand Kit sharper. The more you invest in it, the less time you spend correcting AI output -- and the more your content sounds and looks like your brand, everywhere.

Step 6: Adding Your Visual Guidelines

Why visual guidelines matter

Brand consistency isn't just about how you write -- it's about how you look. Color palettes, typography, logo variants: these are the visual signals that make your brand recognizable across every touchpoint. Without them, AI-generated reports, decks, and data visualizations default to generic styling that could have come from anywhere. Visual Guidelines give Brand Kit the same authority over your visual identity that writing rules give it over your voice -- store them once, and any MCP-connected tool pulls the right hex, logo, or type size automatically.

What we're building: Ramp's color palette

Ramp has a clean, distinct visual identity. We'll add their primary color palette to the Brand Kit so any AI-generated asset knows exactly which colors to use.

In Claude (via MCP):

"Update the [Company] Brand Kit, adding our color palette to the visual guidelines.

Source your information from the [URL] and [Attached document / Brandfetch / etc]

Claude creates a named palette inside the Visual Guidelines section, assigns each color its hex value and usage context, and it's immediately available to any tool generating Ramp assets via MCP -- data visualizations, decks, one-pagers, and more, all using the right colors without a separate design brief.

Step 7: Build a Branded AEO Dashboard

Bringing it all together

Your Brand Kit now governs how your content sounds and how it looks. This step puts both to work combining your visual guidelines with live AEO data to produce a competitive snapshot dashboard that's analysis-ready and on-brand from the first render.

What we're building: Ramp's competitive mention snapshot

Now we can turn AI search data in AirOps into something shareable, governed by our visual system, with just one prompt.

In Claude (via MCP):

"Using the [Company] brand kit and data, create a snapshot from today of how often the brand is mentioned vs competitors."

Claude pulls today's AEO data, applies Ramp's color palette from the Brand Kit, and produces a dashboard showing:

• Mention rate -- how often Ramp appears in AI-generated answers
• Share of voice -- Ramp's percentage of total brand mentions across the category
• First mention rate -- how often Ramp is cited first in AI responses
• Mention rate by brand -- a head-to-head comparison against key competitors

The Brand Kit does the visual heavy lifting -- the palette you added in Step 6 means every chart, label, and callout comes out in Ramp's colors, automatically. No design handoff. No reformatting.

Conclusion: Your Brand, Governed and Always On

Most teams treat brand guidelines as a reference doc. Something you check once and hope for the best. Brand Kit turns that into infrastructure -- a living, governed layer that every AI tool in your stack can access and respect.

Here's what you built in this guide:

• A Brand Kit with your core voice and identity
• A content type that captures format-specific rules (Use-case Pages)
• An audience definition that adjusts tone for a specific buyer (CFOs)
• A writing rule that enforces brand-specific standards (lead with financial outcomes)
• A visual guidelines entry that gives AI your color palette, so every asset it generates uses the right colors
• A branded AEO dashboard showing mention rate, share of voice, and first mention rate -- styled automatically from your Brand Kit
• A real piece of content generated from all of the above

And because Brand Kit connects via MCP, this isn't locked inside AirOps. It's wherever your team works: Claude, Cursor, ChatGPT, and your internal tools.

One source of truth for your voice, your rules, and your visual identity. Always fresh. Always governed.

Dive deeper into Brand Kit.

FAQs

1. Do I need to use MCP to set up my Brand Kit?

No. You can set up and manage your entire Brand Kit directly in the AirOps app. MCP just gives you a faster, conversational way to do it from inside tools like Claude -- which is especially useful for iterating on writing rules and content types without switching contexts.

2. Can I have multiple Brand Kits in one workspace?

Yes. If you manage multiple brands or sub-brands, each can have its own Brand Kit with distinct voice, audiences, content types, and writing rules. Workflows and MCP tools reference the specific Brand Kit they need.

3. What happens when I update my Brand Kit -- do all my workflows automatically use the new version?

Yes. Brand Kit is a centralized source of truth. When you publish an update, every workflow and MCP-connected tool that references your Brand Kit immediately uses the latest version. No manual syncing, no stale context.

4. How granular can writing rules get? Can I scope them to a specific audience and content type?

Very granular. Writing rules follow a priority hierarchy: Region > Audience > Content Type > Global. You can create a rule that only applies when generating blog posts for enterprise buyers in EMEA, for example. The AI applies the most specific matching rules automatically.

5. Can I revert a Brand Kit change if something goes wrong?

Yes. Every edit is versioned with full history. You can see exactly what changed, who changed it, and when -- and revert to any previous version with one click. This is especially important for enterprise teams where brand updates carry compliance implications.

6. What AI tools work with Brand Kit via MCP?

Any MCP-compatible tool. Today that includes Claude, Cursor, ChatGPT (via plugin), and any custom tool that supports the MCP protocol. Your Brand Kit context is portable -- it's not locked to AirOps.

7. How is this different from just pasting my brand guidelines into a Claude Project or custom GPT?

Three things. First, Brand Kit is structured -- content types, audiences, regions, and rules are discrete, queryable objects, not a wall of text the AI has to parse. Second, it's governed -- versioned, auditable, with approval workflows. Third, it's always fresh -- update once, and every tool in your stack gets the latest. A pasted doc is static, unstructured, and impossible to govern across tools.