Brand Facts
Define the canonical facts about your brand — name, products, pricing, claims — so the platform can detect entity drift and accuracy errors in AI responses.
Overview
Brand Facts is the structured ground truth about your brand that the platform uses to detect when AI engines say something wrong about you. Without brand facts, accuracy checks are guesswork. With them, the platform can flag specific contradictions: "Perplexity said your starter plan is $19 — your brand facts say it's $29."
This is one of the most important pieces of configuration for getting accurate signal out of the platform.
What to capture
Brand facts are organized into a few categories. You don't need to fill in every field — capture what matters for the claims you'd want to defend.
| Category | Examples |
|---|---|
| Identity | Brand name, alternate names, parent company, founding year, headquarters |
| Products | Product names, key features, supported integrations, version numbers |
| Pricing | Plan names, monthly/annual prices, free tier limits, currency |
| Claims | Performance numbers, customer counts, certifications, awards |
| Comparisons | How you describe yourself relative to specific competitors |
| Disallowed | Claims you do not want associated with your brand (e.g., outdated taglines, decommissioned products) |
How brand facts are used
Once captured, brand facts feed into several parts of the platform:
- Entity Analysis — uses identity and product facts to compute your entity clarity score
- Hallucination Corrections — detects contradictions between AI responses and your facts, then flags them as candidates for correction
- Shopping Visibility — uses pricing facts to detect price drift in AI shopping responses
- Accuracy reports — accuracy KPIs are computed against the facts, not against a vague text similarity score
The more fields you fill in, the more signal these features can extract.
Setting up brand facts
- Open Settings → Brand Facts from the sidebar
- The page shows a form grouped by category
- Fill in the fields you care about — you don't need to do everything in one sitting
- Click Save
- The platform re-runs accuracy and entity analysis on recent snapshots so you see the impact immediately
Editing pricing facts
Pricing is the most dynamic category — it changes whenever you launch a new plan or run a promotion. The pricing section accepts:
- Plan name (e.g., "Starter", "Professional")
- Price (numeric, with currency)
- Billing cycle (monthly / annual / one-time)
- Notes (e.g., "Includes 5 seats")
When pricing changes, update the facts and re-run accuracy checks. The platform will flag any prior snapshot that quoted the old price as a candidate for Hallucination Correction — useful for sending the right signal to AI engines that the old data is stale.
Disallowed claims
Disallowed claims are facts that you specifically want the platform to flag if they appear. Common cases:
- An old tagline that's been retired
- A discontinued product or feature
- A market position you no longer claim (e.g., "leading X" when you've pivoted)
- A founder's name after they've left the company
If an AI response includes a disallowed claim, accuracy checks mark it as a contradiction and surface it in the Insights action queue.
Best practices
- Start with pricing and product names. These are the highest-leverage facts — they're frequently quoted by AI engines and frequently wrong.
- Update facts immediately when they change. A 24-hour delay between a pricing change and updating facts means a day of false-positive accuracy errors.
- Use the disallowed list aggressively for retired products. AI engines have long memories — explicitly flagging old products keeps you ahead of stale citations.
- Don't capture facts you can't defend. If a number isn't on your website, don't put it in brand facts — accuracy checks should reflect what AI engines could verify.
Related docs
- Entity Analysis — how facts feed entity clarity scoring
- Hallucination Corrections — submitting corrections when facts are misquoted
- Shopping Visibility — pricing accuracy in AI shopping responses