How To Use Your AI Visibility Report

If you're reading this article from an AI Visibility Report link, you've just taken a significant step towards growing your company's presence in ChatGPT and other AI platforms. But if you looked at all that data and weren't sure where to start, this guide is for you.

Spyglasses' AI Visibility Report is designed to give you an answer to four questions:

1. Is what AI 'knows' about my brand accurate? Would someone asking about our brand in an AI chat get the same answer as they would by browsing our website?
2. Is our brand featured when users enter the types of prompts that indicate they're looking for our brand? This is also known as share of voice.
3. How often should I be monitoring this data, and how long does it take to see results?
4. If there are errors in how AI portrays the brand or we're not being recommended, how do I fix that?

This guide walks you through how to read the guide to answer all three of these questions.

Brand Consistency

When you run a Spyglasses AI Visibility Report we use a crawler based on the same technology that ChatGPT and other tools use when they browse your site in a chat. We fetch your site's homepage, and we look for up to five other key pages. We use a tool to extract just the critical information from these pages, and generate a brand snapshot using the data we find. The brand snapshot covers:

• Tagline (if you have one)
• Pricing model (for online businesses this might be a subscription but for a local restaurant it would be "menu"
• Category (e.g. "Digital Marketing Agency, "Korean Restaurant", "Automaker")
• Target Audience
• Key Features
• Problems Solved
• Differentiators

We also report any specific pricing if we find it, and we identify up to 20 competitors (With a free Spyglasses account you can manage these).

We take this brand snapshot and query four AI providers to ask them what they know about the brand. Since all major AI providers; ChatGPT, Claude, Perplexity, and Google Gemini, can browse the web as part of a chat, we query each of them twice; once with web browsing allowed and once without. This will tell you what the AI 'knows' about your brand in its training data vs. what it can learn in real time. If you are a newer company, don't be surprised if you are not in the AI training data yet! Mostly what you want to know is how accurately it communicates about your brand when it can browse the web to find information.

For each of these eight queries we do a comparison of the snapshot with what AI tells us. This generates a consistency score. Unlocked reports can see the specific details, but the score will always be visible.

For each AI, we average the score with and without web browsing to give you a platform average. We then average these to give you a brand consistency score. Higher scores mean your site is being described accurately in AI.

Discovery Queries and Share of Voice

Over 40% of people who use AI use it for product research, and that share is increasing rapidly. When they search for the service, solution, or products you provide, does AI mention you, or your competitors?

A Spyglasses AI Visibility Report can answer that question. Using your brand snapshot we generate five prompts that your prospects are likely to use when searching for your solution. If your company makes project management software, a Discovery Query might be "What software helps with managing project deadlines and team accountability?". For a Chinese restaurant in Santa Clara it might be "Top rated Chinese restaurants near Santa Clara, California, US". An AI Visibility report is guaranteed to have at least five queries from these categories:

• Best in Category ("What are the best [category] [service/solution/product] for [customer/use case]?"
• Use Case Exploration ("What [service/solution/product] helps with [use case]?")
• Budget Constrained ("Affordable [service/solution/product] under [budget range]")
• Comparison ("Compare different [service/solution/product] for [use case]")
• Segment Specific ("Best [business type] for [target customer segment"]
• Local Services ("[Business type] near [location]")
• Local Comparison ("Compare [business type] in [location]")
• Local Reviews ("Top rated" or "Best Reviewed"...)

We automatically generate these based on your brand snapshot, but with a free Spyglasses account you can customize these and run up to ten per report.

For each query we run it against our four AI targets. We track mentions of your brand and competitors in each, including where you are ranked compared to competitors:

We count mentions in each AI response to give you a share of voice for that query, and then we total these to give you an overall share of voice vs. competitors. A higher number of brand mentions and citations, as well as a higher share of voice, mean you're being featured more often in AI for these key prompts.

How often should I monitor this data?

Measuring your AI visibility too often isn't useful and can be a giant waste of money. You should treat this data as directional and use an AI Traffic Analytics tool as your measure of how effective your AI SEO efforts are. Run an AI Visibility Report when:

1. You make a significant change to your website content and want to know how your rankings have changed. Use Spyglasses Recrawl Frequency feature to see how often search engines pick up your content so you know how long after your update to expect to see improvements.
2. When AI companies update their underlying models. This happens about every six weeks. Here you'll be looking for indications that AI is describing you more accurately even without web browsing (or describing you at all).

By only running these reports when you need them you'll be able to track improvements to your site without adding another expensive subscription to your budget.

How do I improve my scores?

Improving your AI visibility is a four-stage process:

Fix any technical issues that would cause you to not be cited by AI

Make sure none of your content needs javascript enabled to render (AI crawlers do not use JavaScript), and your load times are fast. Also look at implementing short summaries near the top of your pages that are optimized for summarization, combining a headline in the form of a question with a short 25-word answer. Our AI Visibility Optimization Implementation Guide has a checklist you can follow.

Start optimizing for AI Discovery

Focus on these types of content:

• FAQ sections using natural language questions your customers actually ask
• Comparison content that positions you against alternatives
• How-to guides with step-by-step instructions
• Use case documentation for different customer segments

Follow the Implementation Guide to identify the Queries you want to optimize for.

Authority Building

One of the major differences between AI SEO and traditional SEO is the inclusion of off-domain authority. Google search links only take into account the content of the specific page, but AI 'knows' about your brand across everywhere it's mentioned. This stage focuses on building your authority off-site by focusing on:

• Industry publications: Pitch expert commentary on trending topics
• Community engagement: Answer questions in relevant Reddit, Stack Overflow, or industry forums
• Original research: Conduct surveys or studies that journalists can reference
• Podcast appearances: Share expertise on industry podcasts

See Our AI Implementation Guide for specific ways to get quick wins.

Platform-specific improvements

Each AI platform has a different user base and different strengths and weaknesses. They also have very different user counts, so if you aren't in a position to optimize for all of them, focus just on ChatGPT which has about 85% of the AI chat market.

Use the below details to identify the platforms that are relevant to your audience. Each platform links to a dedicated implementation guide.

• ChatGPT: Best for businesses looking to get in front of the broadest number of B2B and B2C buyers
• Google AI Overviews: Google is still the king of traditional search, and Gemini is baked into Google's Business products
• Claude: Best for developer-focused products and workflows
• Perplexity: Valuable for research-heavy industries
• Microsoft Copilot: For B2B companies using Microsoft 365