So what is it then? Google says this is the new (free) service where you can add any type of information, and they will host it and make it searchable online. That’s a broad definition but I think the underlying concept is just as broad. Even when I read through the Google Blog’s use cases – “Google Base helps us reach students and parents and deliver more of the information they need when making important college decisions” or “We compile an enormous amount of valuable data on fine art and artists, and we want collectors, dealers, students and art aficionados to have easy access to this information” – I have a hard time getting a feeling for just what it’s all about. So let’s give it a try then.
On the Google Base homepage, after being logged in with your Google Account, you will start out with choosing an item type. You can create items manually one by one, or bulk-upload them via CSV, RSS or Atom (for larger files or lots of files, Google suggests to use their FTP upload instead of the web upload). Available default types are:
Added to those you can also enter your own completely customized data type. But I’ll start with a default type and I choose “News and articles”. (That seems closest to what bloggers are producing – content.) Now I’m on the details screen and need to enter my data.
I’d like to enter yesterday’s interview with Wendy Cheng and start typing the title. Other information bits are author, news source, publish date, some free-style attributes I can define on my own, as well as labels (meta keywords and keyphrases) and descriptions. For the publish date, a nice calendar date picker pops up. At the bottom of the form, I’ll enter the URL of the interview. Before submitting, Google tells me they’ll reserve the right to stop any item that doesn’t comply with their Terms of Service.
One nice feature is that the default contact email isn’t the real email. Instead of displaying “email@example.com” as contact, Google Base offers me to display “firstname.lastname@example.org” (I assume the email then will be simply forwarded to my real one). If you want to, you can also choose to not include any contact at all. I’m giving it my real email address and click “OK”. Then, I hit the “Preview” button.
The resulting preview screen lets me check my information as entered once more, and then offers me a “Publish” button. Just where this information will be published after hitting the button, I don’t know. I suppose Google reserves the right to now displays it anywhere at all – Google Base itself, Google web search, Google Answers, Gmail, Froogle, Google News, and what-not – similar to what they’re doing when you create a Google AdWords ad.
Now I’m on the “My items” overview screen. The interview I entered is flagged as “Processing - will publish soon” in red letters. Are they manually checking these items? I doubt that, as it may take too long, especially for those who bulk-uploaded their stuff. Besides, Google usually prefers to do things automated as opposed to manually.
In other fields, Google tells me that the item won’t expire. Also, it hands me this URL: google.com/base/a/philipp.lenssen, which redirects to a Google Base search for author:philipp.lenssen with no results yet.
Let me search for some items others have entered then. I enter “recipe”. The top hit is in Chinese for a “Bird Flu Recipe”. I suppose Google removes prank results when they see them. Not to spoil the fun, but I’ll click “Report bad item” on that one. Google now tells me “It may take up to four weeks for us to review this item and take action.” (Indeed, they’re not that good at the manual stuff.)
Other recipes available are more meaningful, and include thumbnails as well. I’ll click on “Passion fruit sangria recipe” and am taken to this page on Epicurean.com. Boldly on top of this result page, by the way, are “Ads by Goooooogle” (Google AdSense).
Another search, for “car”, shows how services like GetAuto.com already make use of the bulk upload functionality. So far, the search result – at least the first few pages – are showing nobody else but them. (Imagine you’d be on Google web search for a term like “car”, and you’ve got the SERPS exclusively to yourself!) Certainly, the longer Google Base will be around, the less likely it is for any single site to have the results all for themselves for important keywords.
I wonder what kind of ranking algorithm Google employs here. While PageRank may be working on the web, it’s based on linkjuice, and I don’t know if it’s meaningful (or possible) to analyze the links to Google Base items. Possibly, they’ll check the PageRank of result pages for an estimate, or they check how fresh the item is, or what click-through rates items get, or a combination of those and others. The specific search for cars, by the way, can be ordered by price, and then I’ll get results from other sites for that ranking (cheaper car accessories from StepUp.com, that is).
At this time, the service feels a little like a Froogle for everyone, to offer anything, to be shown anywhere. (The actual Froogle, Google’s shopping service, has strong limitations on the types of people it accepts as merchants, as well as the types of items – just to give it a try, I once submitted Authorama.com’s public domain books, but was rejected – and, as far as I know, it only displays Froogle items on Froogle.com.) Even more than Froogle, GBase is close to Google’s mission statement (“organize the world’s information and make it universally accessible and useful”).
Other than Froogle, Google Base reminds me of another initiative: the Semantic Web, evangelized by the creator of the web, Tim Berners-Lee and the W3C. In this initiative, site owners will share meta-information on their available data using RDF files on their own servers. As opposed to Google Base, it would be decentralized, offering different external spiders to make sense of it all. The Semantic Web is more than just that – “search for yellow car in Chicago, and precisely find it on any kind of webpage” – but that’s a part of it, and I have to say this part so far didn’t surface in web reality. (Unless you consider RSS, and we have reason to doubt the success of RSS has anything to do with RDF syntax.)
Not only does Google Base have a centralized approach to adding meta-data to items, unlike the Semantic Web you don’t even need to have your own web server if you prefer; “Google Base will host your content (for free) on a unique web address”, as the help page states. And who knows when they’ll add their own PayPal-like money transaction service and start charging commissions. Google Wallet could be the last piece of the puzzle here.
I think over time more Google Base use-cases will pop up; the concept is certainly interesting, and I’d like to hear what kind of information you entered, and possibly, where else other than Google Base it appeared. Can Google Base be used to enter blog posts? Can it be used to release a daily cartoon? Can it be used as eBay replacement?
>> More posts