The generator combines your details on profession, city, service, target group and business name into typical customer questions: open recommendations, specific occasions, comparisons and questions about your name. It runs entirely in your browser, stores nothing and uses no AI. You then ask the questions yourself in ChatGPT, Gemini or Perplexity.
The generator: your customer questions in seconds
Fill in the fields. Only the first two are required. The more details you add, the more varied the questions become. Your input stays in your browser and is not sent anywhere.
Your questions to test
Enter at least who your customers are looking for and where. The questions will then appear here.
Ideally, ask the questions in a new chat without earlier conversations and note the date and answer.
Why one question is not enough
AI assistants answer every question afresh. Small differences in wording, occasion or target group can change the providers named considerably. In our study of AI recommendations, the agencies recommended for two slightly differently worded questions overlapped by only about 4 percent. In 22 of 30 cases, there was not a single agency in common.
Anyone who only asks “Which dentist in Kiel would you recommend?” therefore sees only a small part of the picture. Only several questions from different angles show whether your business appears regularly or just by chance.
How to evaluate the answers
Ask the questions in ChatGPT, Gemini and Perplexity, ideally in a new chat without earlier conversations. For each answer, record:
- Is your business mentioned, and in which position?
- Which other providers are mentioned?
- Which sources does the assistant cite, for example directories, review platforms or websites?
- Are the details about your business correct?
Evaluate open recommendation questions and questions about your name separately. The first show whether you are recommended, the second how you are described. If certain sources keep appearing, the guides to business directories and review platforms are worth a look.
Limits of a manual test
A test in your own chat is a good start, but not a measurement. Your answers may be influenced by earlier conversations, your location and account settings. Other people may see different answers. AI answers also vary from day to day.
If you want to follow developments over weeks, you need consistent questions and a fixed process. Nalqerio asks selected questions regularly through the APIs of ChatGPT, Gemini and Perplexity and documents answers and sources. Personal chat histories are not measured, and a recommendation is not guaranteed. The measurement guide and the methodology have more.
Nalqerio offers a free 14-day trial if you have not used it before: one project, three active questions and one complete AI measurement round in total, without a card and without an automatic subscription.
Frequently asked questions
Does the generator itself use AI?
No. It inserts your details into proven question patterns. This happens entirely in your browser, without transmission and without storage.
How many questions should I test?
Five to ten questions in each assistant are enough for a first impression. The mix of open recommendations, specific occasions and questions about your name matters more than the number.
Why do I get different answers to the same question?
AI assistants generate every answer afresh and may use different sources depending on time, web search and context. That is why it pays to ask questions several times and record results with a date.
Sources and further reading
- Nalqerio study: AI recommendations for web agencies
- Nalqerio: measurement methodology
- Nalqerio: trial scope and pricing
Methods, product scope and sources reflect the publication date shown. Examples are not measurements of your website. Linked product pages provide current plan and provider details.
