Your own prompt sets
You define the questions your customers actually ask. RankScan suggests prompts, you decide which ones count.
Diese Seite ist auch auf Deutsch verfügbar.
RankScan is the AI visibility tool that queries ChatGPT, Perplexity, Gemini, Claude and Microsoft Copilot and shows whether your brand gets named, how early and whether the answer links to your website.
People ask an assistant first. What it answers decides who even makes it into consideration.
Are you named or is the AI talking about your competitors?
Your Google rankings are solid, yet you never show up in AI answers.
Competitors get named early while your brand appears at the edge, if at all.
The answers lean on sources about you that nobody on your team has checked.
Without measurement, nobody can say whether any of it is improving.
RankScan runs your own prompt sets against your chosen AI assistants on a schedule and reads every answer. You see whether your brand is named, in which position, in what context and which sources the reply cites.
The largest gap is in Perplexity, where Leitwerk IT leads.
Six readings that turn a block of prose into a number.
You define the questions your customers actually ask. RankScan suggests prompts, you decide which ones count.
You choose per property which assistants RankScan queries: 2 on Starter, 3 on Pro, 4 on Enterprise. You switch on Claude, Perplexity and Copilot yourself.
For each brand you see in how many answers it is named, named early and linked to its website. The history shows you next to the leader.
Which pages the answer cites, how often yours are among them and which links lead nowhere.
Not every mention weighs the same. RankScan counts separately whether an answer names your brand early and links to your website.
Different spellings of your brand get merged so your numbers do not split across variants.
Two examples of how RankScan turns AI answers into an action.
Your brand appears in 1 of 12 answers on the topic, a competitor in 7. RankScan suggests a page that covers the topic as a comparison. The action counts as fixed once you are named in more than 20 % of the answers.
The assistants use third-party pages on a topic. The nightly check does not find your brand on 5 of them. RankScan lists the pages and suggests a listing or a mention, depending on the kind of source.
You see the history of every metric and get a weekly report on how things develop.
The first scores are ready the morning after the first scan, after that RankScan measures every two days.
RankScan suggests country and language.
RankScan reads your website and suggests topics, keywords and prompts. You add your brand and competitors.
Mentions, competitor comparison and cited sources are ready the morning after the first scan.
AI visibility describes whether and how your brand shows up in the answers of AI assistants. It does not measure a position in a result list. It measures whether you are named, how early and whether the answer links to your website.
ChatGPT, Perplexity, Gemini, Claude and Microsoft Copilot. How many you can monitor at once depends on your plan.
Every two days. During the trial the scan runs daily.
RankScan sends your prompts through the providers’ official APIs, with web search switched on. The answers can differ from what you see in your own account, because history and settings play a part there. AI answers also vary from run to run. That is why RankScan calculates across several measurements and shows the history.
SEO measures positions in Google search. AI visibility measures mentions inside generated answers. The two belong together, and RankScan shows both side by side. For how to work towards mentions, see our article on Generative Engine Optimization.
You can write them yourself. RankScan also suggests prompts that match your topics and your website.
Yes. Competitors are counted in every answer so you can put your own share in context.
Start your free trial. No credit card, cancel any time.
Start free trial