case studies

Proof that agentic content drives growth

Frequently Asked Questions

What's inside a Slate teardown?

A Figma board of 19 to 21 sections on one category leader: its funding and traffic timeline, every channel it uses to reach buyers, how it shows up in AI search, and the gaps a challenger could use.
The findings are grouped into 8 or 9 plays, and each play ends with the move you can copy.

Which channels do the teardowns cover?

Every channel the company uses to reach buyers. Across Clay, Ramp and Rippling that includes founder and exec LinkedIn, creators and community, review sites and Reddit, best-of lists, glossaries and guide libraries, free tools, newsletters, brand films, and the pages AI engines quote.

How do you measure a company's visibility in AI search?

Slate runs 30 to 50 buyer prompts for each company across ChatGPT, Perplexity and Google AI Overviews with its AI Search Analytics, and records who gets mentioned and which sources get cited. Some teardowns add AI mention counts from Ahrefs Brand Radar.
Traffic and keyword figures come from Ahrefs and Semrush, and company facts from public announcements and reporting.

How current is the data?

The Clay teardown uses data from September 2026, and the Ramp and Rippling teardowns from October 2026. Each teardown states when its data was collected, and the funding and traffic figures reflect the company's position at that time.

How do I get a teardown?

Open the teardown you want and enter your name and work email in the form. Slate emails you the Figma board for free.
Each board holds 50 to 80+ screenshots of the company's pages, AI answers and posts, placed next to the analysis they support.

Can Slate run the same analysis on my brand?

Yes. The prompt tracking and citation data in each teardown come from Slate's AI Search Analytics, which you can point at your own brand and competitors.
On a demo, the Slate team runs an AI search audit of where your brand shows up today and where competitors get cited.