Brand Facts
The three-layer truth system that keeps AI output accurate and on-voice
What It Is
Brand Facts is a system for keeping every piece of AI-generated content factually accurate and tonally consistent with your brand. It has three components, each managing a different dimension of content quality:
Fact Categories — Verified factual statements about your business, organized by topic. These are the truths that SI, content creation, and every other AI operation must always respect. Categories might include Platform Capabilities, Target Audience, Pricing, Integrations, and Limitations.
Safe Phrases — Approved words, phrases, and expressions that are part of your brand vocabulary. Safe Phrases override any other content quality rules — they are whitelisted. If your tagline is "AI Powers Your Growth" and the system would normally flag "powers your growth" as hype, the Safe Phrase listing protects it.
Phrases to Avoid — Language your brand should not use, with content-type-specific strictness. Each phrase to avoid includes a reason (why it should be avoided) and a set of sliders that control how strictly the rule applies across different content types — because a phrase that is inappropriate in a proposal might be acceptable in ad copy.
Brand Facts are generated during implementation and can be managed in Brand > Brand Facts. You can add, edit, or remove items in all three sections at any time.
Why It Matters
AI language models are designed to be helpful, and that helpfulness can lead to two problems: overstating capabilities and drifting into generic marketing language.
Without Fact Categories, SI might claim your product integrates with a tool it does not support, or state a pricing tier that does not exist, or imply a capability that is on your roadmap but not yet available. Facts ground the AI in what is actually true about your business right now.
Without Safe Phrases, the system might filter out your own brand language. If your approved tagline includes a word that happens to match the hype-detection rules, Safe Phrases ensure it is protected.
Without Phrases to Avoid, the AI defaults to whatever language it thinks sounds good — which often means the same overused marketing cliches that every other AI-generated content sounds like. Phrases to Avoid ensure your brand sounds like *you*, not like a template.
Together, these three layers create a factual and linguistic boundary that every AI operation in Zyntro respects.
How It Works
Fact Categories
Fact Categories are organized by topic — Platform Capabilities, Target Audience, Pricing, Integrations, Limitations, and any custom categories relevant to your business. Each category contains individual fact statements that can be added, edited, or deleted.
Every fact is a plain-language statement like "Zyntro is a web-based SaaS platform. As of May 2025, Zyntro does not offer an iOS or Android mobile app." SI reads these before generating any communication and will not contradict them.
You can add new facts at any time using the Add New Fact button within any category. This is particularly important when you launch new features, change pricing, or need to correct a common misconception.
Safe Phrases
Safe Phrases are words and expressions that are explicitly approved for use across all content. They override any content quality filters that might otherwise flag them. Your brand tagline, product names, approved marketing phrases, and technical terminology belong here.
Examples: "AI Powers Your Growth," "AI-powered platform," "Integrated platform," "Automate your growth," "Command center for business growth," "AI Phone Assistant (Phona)."
You can add new phrases using the Whitelist a Phrase button. If you notice the AI avoiding a phrase you want it to use, adding it to Safe Phrases resolves the issue immediately.
Phrases to Avoid
Phrases to Avoid are the opposite — language your brand should not use. Each entry includes the phrase, a reason it should be avoided, and strictness sliders for each content type.
The reason explains *why* the phrase is off-brand. For example: "Revolutionize your business" has the reason "Overused hype. Zyntro prefers grounded, results-oriented language focused on practical benefits." "Game-changer" has the reason "Generic marketing cliche. Lacks specificity and Zyntro's knowledgeable tone."
The strictness sliders range from 0.0 (allowed to use) to 1.0 (never use), and they are set independently for each content type: Blog Article, Website Landing Page, Lead Magnet, Email Content, Webinar Script, E-Book, Ad Copy, Press Release, Proposal, Slide Deck, Postcard Graphic, and Social Media Post.
This means a phrase can be strictly forbidden in proposals (1.0) but tolerated in ad copy (0.4). "Revolutionize" might be completely banned in e-books and email (1.0) but allowed at low weight in social media posts (0.5) where punchier language is more common.
You can adjust sliders at any time and add new phrases using the Add Phrase to Avoid button.
Examples
A SaaS company prevents AI from overstating mobile capabilities
They add a fact to the Platform Capabilities category: "Zyntro is a web-based platform. A native mobile app is planned for Q4 2025 but is not yet available." SI will never tell a prospect that Zyntro has a mobile app, and if asked directly, it references the planned timeline without overpromising.
A coaching brand protects their signature phrase from content filters
The coach's signature phrase is "Launch your brand in a weekend." Without Safe Phrases, the content quality system might flag this as hype. By adding it as a Safe Phrase, the coach ensures the AI uses it confidently in emails, website content, and proposals.
An agency adjusts phrase strictness by content type
The phrase "crush the competition" is a 1.0 (never use) for proposals, email, and the website — it sounds too aggressive for those contexts. But it is set to 0.3 (low strictness) for social media ad copy, where bolder language performs better. The AI respects the context-specific rules and adjusts its language accordingly.
