I'm running runeberg.org, which happens to be the external website with more links from the Swedish Wikipedia than any other. On a typical day, yesterday April 24, my site had 42,000 page views (robot crawlers not included) from 7800 different IP addresses, and 902 of them had Wikipedia as a referer, namely 732 *different* Wikipedia pages.
Here's one example, at 9:26 AM, one iPhone user got from http://sv.m.wikipedia.org/wiki/Smedjebackens_Valsverks_AB (an article about a steelwork) to http://runeberg.org/steelswe/0114.html (page 46 of the book "Iron and Steel in Sweden" from 1920). That book was scanned in 2004, the article created in 2005, and the link was added to Wikipedia in 2006.
Was this "worth it"? I certainly didn't make any money from it, and I didn't pay the people who put all those links in Wikipedia. I think it would be very hard to do this on a commercial basis. It takes time to add links to Wikipedia, not just the 30 seconds to edit the page, but perhaps 30 minutes to find the relevant article and the relevant webpage to link to. Can you speed up that process, without getting questions about link spamming?
That's more or less how I see it, too. But still, there are such users [1]. I'm estimating that about 50% of her links remain after cleanup. I don't want to be mean, but she will probably be blocked sooner or later for those activities (because most user still see this as linkspam). So it must be worth *something* to the newspaper, right?
Thank you for sharing those stats, Lars. Most interesting. We had the same thing happening at Swedish Wikipedia as Strainu describes here, from one of Swedens largets newspapers. Most of the edits were reverted, and the only lasting outcome was a lot of badwill for the newpaper among the community, so I think that was not a very well thought out strategy. /Leo Wallentin