Forum Moderators: open
On first impressions it seems to fool google thinking your pages are being updated.
Surely they are being updated, though?
I tried this about a year ago on my front page. I had a grid of four elements, each one a photograph and some accompanying text with links to features inside my site. I set up a pool of about 30 elements and a javascript to shift the elements forward every minute. So over a period of minutes, if a reader kept returning to the homepage, they would see ABCD, followed by BCDE, CDEF etc.
I wouldn't classify this as 'fooling' anyone. It strikes me as a perfectly legitimate way to keep the page looking fresh and different every time a reader returns... (though the refresh rate was probably too frequent).
However I scrapped it, because obviously the javascript meant nothing to the robots, so to them it looked like there was very little content on the page.
I'd love to reintroduce it at some point though in a format comprehensible to robots as well as humans. How are you implementing the randomization, TheSubtleKnife? Are you using a server-side script?
Our sites sell widgets, each site has upto 120 different types of widgets. Each site shows 5 "Featured Widgets" on the home page (a short description and a photograph), these are picked at random from our widgets database.
Seems to work okay for us, and it looks like it helps Google crawl all of our widget detail pages.