November 28, 2005
I was recently asked what I thought the purpose of Google Base was. Here were my thoughts:
Google had approached local search and iYPs such as Citysearch and Superpages for feeds/APIs to their listings…and they provided it. They had asked the same from news providers for Google News and many of them provided it as well. Not a long time ago, there were rumors that Google was approaching sites with job listings, classifieds, etc and asking them for feeds/APIs as well…but the rumor was that they turned them down.
Why/How is this possible? Well, location listings (as in local search sites) are mostly static…their information may change but the listing for the pizza joint down the street will still be there as long as that business is solvent and has not moved. Although those cases do happen, they are few and far between. Job listings and classifieds are temporal in nature. They are for a current time and place. How do you explain the fact that the news content is also temporal in nature yet the news providers want the content listed on Google? Well, the news is new every day. Old news gets archived and is still looked up for historical purposes. Job listings, for example, expire…once the job has been filled, you don’t want to display it (especially if you’re charging for it to be displayed).
What does the temporal nature of data have to do with wanting the listings appear in Google or not? Well, the local search sites and the news sites want to capture that traffic for people searching on Google using SEO methods. People go to Google and search for Spago and those sites want their URL to show up as close to the top as possible. However, due to the temporal nature of classifieds and job listings, people are better off searching directly at the source such as Monster, Dice, EBay, or Craigslist. Otherwise, the data Google will display may be out of date (it takes time for them to crawl every page of a site (they don’t often pick up every page of a site especially one with a large amount of pages…and when/if they get to that point, everything they had crawled earlier on is out of date!).
The job/classifieds/auctions listings sites do not rely on SEO traffic to these deep links (where the actual listings exist). They mostly rely on these search engine referrals to take them to category or home pages, mostly (and, in some cases, search results pages). The referrals to the deep links may not be relevant a week from now due to their temporal nature.
So, basically, if these sites don’t care to have google refer them traffic to the individual listings (keep in mind their business models…some are ad driven, some are transaction driven, etc), Google must create a strong destination site where end-users wanting maximum exposure would feel compelled to submit their data (and the sheer amount of searches/users on Google may be compelling enough).