On 10/26/07, Sage Ross ragesoss+wikipedia@gmail.com wrote:
But I also think that switching to Google search (with ads) would be much less likely to raise red flags with users, compared to article ads. After all, they are used to seeing ads in the context of search; it would all seems part and parcel of using faster and more relevant Google search (and/or MSN or Yahoo! or Ask). Most users sophisicated enough to worry about ads and revenue streams can also appreciate the functional difference between Wikipedia's current search and what an integrated commercial engine could do.
Some quick arguments against:
* Google doesn't exactly have the greatest privacy track record. Right now it's possible to do any kind of full-text search without fear of contributing to some eternal profile about yourself that will be (directly or indirectly) fed to advertisers. And surely Wikipedia searches are among the most interesting data for commercial tracking, trends analysis and profile building.
* While our search is far from great, right now we at least have the potential and the incentive to innovate. If the current fundraising drive is successful, we'll be hiring more developers soon -- and that will allow us to think in the direction of Semantic MediaWiki / Structured Data support, to build all kinds of interesting query tools. See dbpedia.org for a glimpse at things to come.
The likelihood that we'd be investing significant resources in any of these projects if we swap out the default search is very small, meaning that a very interesting slice of innovation & research in search and data mining would essentially be handed over to commercial interests. I think that'd be a shame.
* The same argument that applies to any advertising applies to search-based advertising: it could reduce the incentive to donate (both on a small and large scale), could complicate our relationship with other non-profits and private foundations, could reduce users' feeling of equity in the site, could endanger our tax-exempt status, and so on and so forth.
I find it interesting that this discussion is coming up in the context of a fundraising drive which, so far, seems quite successful, even though the execution is still leaving quite a bit to be desired.