Enter a company’s website. Magister reads its public trail, dates what changed, and writes a teardown with the plays worth copying, each finding linked to its source.
The free teardown tool is not open yet. Magister's agent runs the same teardown in chat: sign up free and ask it how a company grew.
Any public company site. If Magister already has a teardown of it, you go straight to that page. A short abuse check runs first.
The sitemap and live homepage, six months of Internet Archive captures of the homepage and pricing page, public DNS records, and public Greenhouse, Lever and Ashby job boards. Anything it cannot find is shown as not found, never guessed.
Up to three years of estimated Google search visits, shown as an index where the first month is 100. It is modelled from third-party ranking data, not the company’s own analytics, and covers search only.
One pass over the collected evidence: what changed and when, what each channel shows, five plays to copy and what to skip. A finding that does not cite collected evidence is dropped before the page is saved.
Estimated search visits as an index over time, with the months growth picked up marked and what else changed around them.
Headline and pricing changes from the last six months, dated to the archive capture that first shows them, tools first seen on the homepage, and marketing roles by the month they were posted.
What the evidence shows for search, content, email, hiring and pricing, the tools found on the site and in DNS, and links to the company’s ads in the public ad libraries.
Each play names the evidence behind it and the effort it takes. Plays are Magister’s interpretation and are labelled Inferred.
Yes. No account and no credit card. Each visitor can start one new teardown every 30 days; teardowns that already exist are free to read for everyone. To run more, sign up free.
Public sources only: the company’s own site and sitemap, Internet Archive captures, public DNS records, public job boards and the public ad libraries. The growth curve is an estimate from third-party search ranking data, shown as an index rather than visit counts.
Observed findings were read directly from a public source, which is linked beside them. Inferred findings are Magister’s interpretation of that evidence, including every play, and may be wrong.
Anything that is not public: revenue, ad spend, conversion rates, email performance, and traffic from social, direct or referral sources. The archive is read six months back and its captures are monthly, so older changes are missing and a change may be older than its date. Tools loaded through a tag manager or after a consent banner do not show.
Each company has one teardown page. A page with enough evidence is published and can be found in search; one without is kept unlisted. A page is refreshed at most every 30 days. Magister does not write teardowns of adult or gambling sites.
Yes. Every teardown page lists the address to write to for a correction or a removal.
This check is one signal. The free audit grades your whole site for AI search and shows the plan Magister would run to fix what it finds.