Issue 3 - What are Synthetic Metrics in GEO?

Alright Rankonauts, let’s get ranking!

Last week we focused on traditional metrics like traffic and revenue, and how to see that in the tools you’re already using, like GA4.

You can read it here, if you haven’t already. And if you haven’t, why haven’t you? (sigh)

This week, we’re going to focus on the next category of metrics that I like to call synthetic metrics.

Unlike something like Google’s search result page (SERP), AI chats have a tremendous amount of variability in the way they cite sources, and what they choose to answer and what links they provide.

So how can you test how often you’re showing up?

Can you run thousands of slightly different queries and messages and see how many times you are mentioned?

That’s exactly what the latest generation of measurement tools do.

They try to simulate thousands of possible messages that a user may type, and collect the responses, analyze the mentions of brands and competitors, sources, and average position of your brand in responses.

Here’s a screenshot of hundreds of simulated queries in a tool called Am I on AI:

So what are the top 3 metrics that we need to focus on?

1. Brand Average Mentions (or Share of Voice)- out of 100 queries, how many times did your brand get mentioned in the response?

2. Brand Average Position - Is your brand being mentioned first, or after a number of competitors?

3. Linked Mentions - How many times is your website cited as a source to answer a query - this is what leads to the most traffic

Sometimes, these metrics are blended to come up with a rank among competitors.

Here’s an example screenshot of how a tool like AI SEO Tracker returns the Average rank across all engines.

🧰 4. Tools to Try

Here’s are the tools that I’ve tried over the past few weeks:

🧠 5. Prompt of the Week: What questions can people be asking about my brand?

One easy step to take to understand your brands visibility on ai search is to try to understand the queries people are putting in to find a product or service like yours.

This prompt will help you get that process started. Replace the placeholders with your brand and product information, and you’re off to the races!

You can click the button below, or copy and paste the prompt from the prompt box.

I want you to act as a prompt researcher for Generative Engine Optimization (GEO).

Your job is to help me build a high-quality dataset of realistic, customer-style prompts that I can use to test how my brand shows up in AI-generated search responses.

Here’s what you need to know about my business:

Brand name: [INSERT BRAND NAME]

What we offer: [Brief description of product or service — e.g. “non-toxic kitchenware for home cooks” or “paid social ad services for Shopify brands”]

Our ideal customer: [Brief persona — e.g. “new parents shopping for safe baby gear” or “growth-stage ecommerce founders looking to scale through automation”]

Your goal:
Generate a list of 20 diverse, buyer-intent prompts that a real customer might type into an AI chat tool when researching, evaluating, or buying a product or service like mine.

These should be:

Natural language

Conversational in tone

Reflective of different parts of the buyer journey

The kinds of questions you’d ask an AI assistant, not just Google

Prompt Categories to Include:
Try to include a variety of prompt types:

Comparative Prompts

“Best [product] for [specific use case]”

“What are the top brands for [category]?”

“Alternatives to [popular competitor]”

Problem-Solving Prompts

“How do I solve [common pain point]?”

“Is there a tool that helps with [specific job]?”

Feature-Focused Prompts

“What should I look for in a good [product type]?”

“Do I need [feature] when buying [product]?”

Buying Decision Prompts

“Which [product] should I choose for [situation]?”

“What are the most trusted [category] brands?”

Contextual or Persona-Based Prompts

“Best [product] for beginners”

“Tools that help [persona] do [job] faster”

Bonus Instructions:
Avoid SEO-optimized phrasing like “buy now” — write the way people actually ask questions.

Prioritize depth over breadth — it’s okay if 5+ prompts are very similar but just phrased differently.

Group the final list by intent type (Informational, Comparative, Transactional, etc.) if possible.

Make them suitable for synthetic GEO testing — meaning they should reveal what brands, tools, or websites AI engines surface in their answers.

If Brand, product or customer details are placeholders, ask me to provide those details before giving me the queries.

🆕 This Week’s Links You’ll Want to Read

1. Adobe Launches LLM Optimizer to Track AI Visibility
   Adobe introduced its LLM Optimizer at Cannes — a tool geared toward helping brands measure their visibility inside chatbots, web browsers, and AI-powered interfaces. It’s called LLM Optimizer but it might as well be “Does Daddy ChatGPT Love Me Yet?

2. SE Ranking now shows you how bad you’re losing to ChadBrand™️ in Claude’s citations. Competitor visibility in AI chats? Now measurable. Your ego? Now damageable.

3. SurferSEO Quietly Adds AI Tracker to Its Dashboard

   You’d think this would be a headline feature — but Surfer just soft dropped a tool that tracks how often you appear in generative answers.
   It’s like Google Search Console, but for vibes.

4. Google’s AI Overviews Cannibalizing Clicks, New KPIs Emerging

   Search Engine Land published a breakdown of 12 emerging metrics including “impressions in AI answers,” “citation rate,” and “answer surface share.” The old playbook is being rewritten — one bolded Gemini summary at a time.

5. Keyword.com has added a new feature to track how often your brand shows up in AI search results, including ChatGPT, Perplexity, Gemini, and Google’s AI Overviews. Their new AI Visibility tool lets you monitor mentions and citations across engines — think of it as rank tracking for the zero-click era. Now you can confirm, with data, that old man Chatgpt is ignoring you out of spite.

That’s it for this week.

If you’re experimenting with any of these tools, or building weird dashboards to track it all, send them my way — I want to see what’s working and what’s completely broken.

See you next week. We’ll dig into the next category of metrics - Objective Metrics.