Quick overview

This skill measures whether AI assistants actually name your brand when someone asks them what to buy. It puts a realistic set of customer questions to ChatGPT, Gemini, Claude and Perplexity, records who gets named and who gets cited as a source, compares you with your competitors, and turns the whole thing into a report your client can read — branded as your agency, not as a tool. It answers the question every client is now asking: are we visible in AI search, and how do we compare?

Who it’s built for

For SEO and digital agencies being asked about AI search and needing evidence rather than opinion, and for in-house teams who have to report on it. It works as a new-business conversation starter and just as well as a recurring monthly line item once a client decides AI visibility is worth tracking.

What you get

• A report your client can read. Branded as your agency, no tool logos to explain, no raw data to interpret for them.
• Two numbers that never get confused. How often you’re mentioned and how often you’re actually cited as a source are different problems with different fixes, and they stay apart.
• Four assistants, measured identically. ChatGPT, Gemini, Claude and Perplexity get the same questions and the same scoring, so the comparison means something.
• The questions on the page. The full prompt list is printed in the report, so nothing looks cherry-picked and a client can see exactly what was asked.
• A list of where you’re invisible. Named question by question — which is the work list.
• The sources the assistants lean on, so you know where being published would actually pay, plus competitors side by side on the same measures.
• Something to compare next month against. A data file ships with the report, so month two is a trend rather than a fresh guess.

What you need to get started

• A DataForSEO account and your API credentials.
• Claude Code, with the DataForSEO connection added once.
• The AI Visibility Report skill installed.
• The brand and market you want measured, a handful of keywords, the competitors to compare against, and your agency name and logo for the cover.

Workflow breakdown

• You name the brand, its market, a few keywords and the competitors you want it measured against.
• From those keywords the skill writes a realistic question set — the mix a real customer asks: open questions, comparisons, direct recommendations and brand-specific ones.
• You see the questions and the size of the run before anything starts, and approve it.
• Every question goes to ChatGPT, Gemini, Claude and Perplexity, and each answer is read for whether the brand was named, how prominently, and whether its own site was cited.
• The same reading is applied to every competitor, so the comparison is like-for-like.
• You get the branded report — headline score, per-assistant breakdown, per-question view, the sources being cited, and a short list of moves — plus the data file behind it.