Be the one AI recommends.

You can, once. Ternly asks five assistants with web search on, keeps every answer, and shows who was named, in what place and from which pages, so you can see what moves when you change something.

On its own, not much. That's why Ternly tracks the share of answers that name you, across assistants and questions, and how it moves over time.

You can change what it finds. With web search on, the assistants read pages as they answer, so your site and the pages they cite shape the answer. What a model learned in training only changes with new versions.

ChatGPT, Gemini, Perplexity and Copilot on every plan, and Claude on Studio and Agency, each with web search on, plus Google's AI Overviews in rank tracking. The free scan when you sign up asks ChatGPT, Gemini and Perplexity. For every question, Ternly records whether you're named, in what place, who is named instead, which sources are cited, and in what tone.

Because that's where the AIs learn about you. They read the web: your homepage, your docs and guides, your comparison pages, and the pages that review or link to you. Ternly checks that AI crawlers can read your site and that its pages carry structured data and llms.txt. If you have an app, it also checks that your site sends visitors straight to your App Store and Google Play listings.

No. Ternly crawls your site and reads your public store listings on its own. Connect Search Console and GA4 (read-only) when you want search traffic joined in, and App Store Connect and Google Play Console if you have apps.

Only what you connect it for. Signing in with Google gives Ternly your email address. Connecting Google gives it read-only access to the Search Console and Google Analytics data of the sites you pick, which it uses to show you your own search and traffic reports and alerts. It never uses that data for ads, never sells it and never trains AI models on it, and it follows the Google API Services User Data Policy, including the Limited Use requirements. You can disconnect at any time and have it deleted. The details are in the privacy policy.

No. It crawls at three requests a second by default (you can change it), follows robots.txt, and always comes from one fixed IP address you can allow-list.

From Common Crawl's public web graph, plus the linking pages themselves for anchor text and nofollow. Common Crawl samples the web, so you'll see fewer links than Ahrefs or Semrush show, but every one is real.

Yes, with as many people as you like on any paid plan; Free is for one person. Your sites and apps live in a workspace. Invite people as admins, members or viewers; viewers can look but not change anything.

An audit stops at the pages you have left. Other runs don't start until the 1st of the month, or until you upgrade. On a paid plan you can also add AI answers with a top-up: 750 more a month for $19, or 3,000 for $59.