Skip to content
AgentLensBeta

Start with one analysis

Demand your agentic visibility report.

Your domain, at least three questions your customers ask, and where to send the answer. Every field has an ⓘ with an explanation and examples.

The brand
The name customers use when they talk about you, not the legal entity. This is the name we look for in AI answers. For example: Garanti BBVA; Mavi. Aliases (a shortened name, an old name) can be added later in Brand Management.
The homepage of the site you own. AgentLens crawls about 25 pages from it, so it should be the domain customers land on, not a campaign microsite. For example: garantibbva.com.tr; https://www.mavi.com. We keep only the domain; paths, tracking parameters and anchors are dropped.
Selects the governed prompt set and the rival pool the report benchmarks against. For example: Banking & Finance → 47 banks, 168 prompts; Fashion & Apparel → 99 brands. Choose Other if nothing fits; your own prompts still run in full.
Up to three brands you want the answers compared against. Leave empty and we use the sector's default rival pool. For example: Akbank, Yapı Kredi, İş Bankası; LC Waikiki, Koton. Rivals the models mention on their own are surfaced anyway; these are the ones you want tracked from day one.
The questions
Full questions a customer would ask an assistant, in the language your customers use. Keywords and product codes do not work: assistants answer questions. For example: Which bank has the best travel credit card in Türkiye?; Recommend a Turkish denim brand with sustainable jeans.; Ankara'da ev almayı planlıyorum, düşük faizli konut kredisi için hangi bankayı önerirsin?. Three to ten prompts. Mix intents: a recommendation, a comparison and a research question give the most useful first report.

At least 3, up to 10. Full questions in your customers’ language, not keywords.

  1. 01
  2. 02
  3. 03
3 of 10
The surfaces
The assistants the report queries. ChatGPT, Claude, Gemini and Grok run through their vendor APIs; Google AI Mode is read as a live search surface. For example: All five (recommended for a first report); ChatGPT + Google AI Mode only. Only surfaces AgentLens can measure with a real adapter are listed. Perplexity is not one of them today.
You
Who we address the report to. For example: Ayşe Demir.
Where the report and the follow-up go. A company address lets us match the request to the brand faster. For example: ayse.demir@garantibbva.com.tr. We never share it and we do not add it to a newsletter.
Helps us pitch the report at the right level: commercial summary for marketing leads, method detail for SEO and engineering. For example: Head of Digital Marketing; SEO Lead; CTO.
Anything that changes how we should read the results: a campaign in flight, a domain migration, a market you care about. For example: We launched a new cards page last month.; Focus on the German-speaking market..
0/500
We reply within five business days.