CitePack and Gumshoe: what each one is built to show.
On this page
How CitePack reaches 60 observations of one decision.
CitePack maps how AI engines choose in a defined buying decision: one offering, the buying category it’s bought in, and the market its buyers are in. The category and the buyer’s situation shape which companies an engine puts forward, so the decision is defined before anything is asked.
CitePack frames the questions. Each is drafted for your offering, its buying category and your buyers’ market, one for every real buying situation across the whole decision: a buyer just starting to look, one with a job to get done, one with a condition they won’t bend on, one weighing what matters most, and one moving away from what they use now. Each carries the buyer specifics that matter, like the work to be done, a must-have or what they use today. None of it is a fixed prompt from a library. You read every question, change any wording you like and approve all five. The final go is yours. Then every engine gets them exactly as written.
- 5buying situations, framed by CitePack
- × 4engines: ChatGPT, Claude, Gemini, Perplexity
- × 3clean repeats on each engine
- = 60observations of one decision, every answer kept
Each question goes to ChatGPT, Claude, Gemini and Perplexity, three times on each engine, and every repeat is clean. Five situations, four engines, three repeats: 60 observations of one defined decision. Every answer is kept word for word, with the searches the engine ran and the pages it read. So the pattern shows which names an engine keeps putting forward, which turn up once, and how each engine chose.
The Work Plan turns those answers into ranked work. Each item names the answers and pages behind it, who does it and what to check when you measure again.
Map each offering, category and market you sell into, and each map adds to the picture of where your company stands in its market.
The two methods.
| What is compared | CitePack | Gumshoe |
|---|---|---|
| Built to show | How each engine chose who to put forward in one defined decision: your offering, its buying category, your buyers’ market. | “Monitor your visibility & sentiment across ChatGPT, Gemini, Claude, Perplexity, and more AI models” |
| Where the questions come from | Framed by CitePack for your offering, category and market: one per real buying situation across the whole decision, with the buyer specifics that matter. You give the final go. | “Persona and topic combine into the kind of questions a real buyer would actually ask AI”, from personas Gumshoe builds by interviewing AI about your category. |
| How the pattern shows | Each question three times on each of four engines, every repeat clean: 60 observations of one decision, every answer kept word for word. | “each submitted once per model” |
| The result | No single score. Put forward, mentioned and cited are kept apart, and every finding opens to the answers and pages behind it. | “you get your overall visibility score”, with “±5 pp at 95 percent confidence” |
| What you act on | A Work Plan ranked from the answers, each item tied to the answers and pages behind it. | Visibility and sentiment monitoring, and a “Content Generator” that “generates structured content built to close those gaps” |
Where the questions come from.
Gumshoe starts from personas. In its words, it “interviews AI about your category to uncover who they see as your ideal customers, and those personas become the synthetic users we use to generate your AI visibility audits”. CitePack starts from the decision. It frames each question for your offering, category and market, one for each real buying situation your buyers face. Nothing is asked until you give the go.
What you get.
Gumshoe gives an overall visibility score for an audit. CitePack gives you a map of how each engine chose in your decision: who it put forward, what it said about each name, and which pages it read and found by search. Beside the map, the Work Plan ranks the work. Each item names the question, the answers and a verbatim search or quote behind it.
Who each fits.
Gumshoe fits a team that wants to monitor a visibility score across personas and models. CitePack fits a team that wants to see how the engines choose in each decision it sells into, and which work comes first.
Sources.
Every sentence about Gumshoe on this page is quoted or summarised from these pages, read on 22 October 2026.