Understand how AI sees your brand

This workflow helps you measure whether AI/web results mention your brand when users search for products in your category.

If you’re wondering “does ChatGPT/AI mention my brand?” or trying to improve your AI search visibility, this template gives you a repeatable way to test real, high-intent prompts and track results over time.

What it does

• Extracts SEO keywords from pages in your collection
• Converts keywords into high-intent prompts (optionally including the current year)
• Runs web research per prompt
• Checks whether a set of target brands is mentioned
• Writes a snapshot of the results into Google Sheets

How it works (in plain English)

1. Seed topics from your site: it reads your pages and extracts keywords you can realistically rank for.
2. Turn topics into buying-intent prompts: it generates prompts that resemble what a prospect would ask when comparing options.
3. Collect answers: it runs web research / browsing to create a response for each prompt.
4. Detect brand mentions: it checks whether your brand (and competitors) are mentioned in the response.
5. Store the snapshot: it appends a row to a Google Sheet so you can see trends over time.

When to use it

• You want to understand your brand visibility in AI-assisted discovery
• You want a repeatable way to benchmark mentions over time
• You want to compare your brand vs competitors across the same prompts

How to interpret the results

• Higher mention rate usually means your brand is present in the sources the model/web research pulls from (reviews, comparisons, directories, “best X” lists).
• Low or zero mentions can be a content/distribution gap: you may rank on Google for some terms but still not show up in the datasets or pages AI systems cite.
• Competitor-only mentions often indicates you need more third party coverage (partner pages, marketplaces, review sites) for those exact “buying intent” topics.

Setup

1. Put the pages you want to evaluate into a Needle Collection
2. Define a brands variable (your brand + competitors) used by the pairing step (example: ["Meisterwerk","Hero"])
3. Duplicate the Google Sheet (template) and use your own copy, then replace the Spreadsheet ID in the Google Sheets step
4. Connect Google Sheets and update the target spreadsheet + sheet (if needed)

Variables (required)

This template uses workflow variables so you can easily reuse it for any brand.

Tips

• Start with 10 to 25 prompts and expand once the results look stable
• Use consistent brand spelling (and add common variants)
• If you want to exclude your own brand from keyword generation, keep the exclusion rule in the keyword extraction prompt

FAQ

Does this measure “LLM SEO” / “AI SEO” directly?

It’s not a single universal score, but it’s a practical proxy: for the prompts you care about, it measures whether your brand appears in the generated answers (and tracks changes over time).

Why are results different day to day?

AI/web research can change with:

• updated web pages and rankings
• different sources being retrieved
• small variations in response generation

That’s why writing each run to a sheet is useful. You can look at trends, not just one run.

What brands should I track?

Track:

• your brand name (plus common variants)
• 3 to 10 direct competitors
• optional: category marketplaces/directories you want to appear on