Google Base has got us all guessing

The covers have finally been taken off Google Base and there seems to be little consensus in the blogosphere on its likely impact. Some quite strong negative reaction (see Techcrunch Google Base Launched. Yuck. ) but mostly a wave of speculation.

Michael Parekh hits the nail on the head regarding the implications for Google's Search business and also neatly summarizes the way in which most commentators on GBase have reflected their own particular fixations:

To put in mainstream terms, Google Base is a Lego set for users to submit and categorize any kind of content that's important to them.

This makes Google Base kind of the elephant being described by blind-folded folks:
1. "It's Online Classifieds" and will go after Craigslist.
2. "It's Online shopping" and will go after eBay and Amazon.
4. "It's an Online repository for photos, music and videos" and will go after Flickr, iTunes and others.
5. "It's a way to tag content" and will go after del.icio.us and others.
6. "It's a way to to put resumes online" and will go after Monster, Indeed and others.
7. "It's a way to do online photos, music, videos, etc." and will
go after Flickr, iTunes, and others.
8. "It's a way to back into online databases, potentially word processors and spreadsheets", and so go after Microsoft.

I'm not sure what happened to item 3, and 4 and 7 are the same, but its number 8 that I think is significant (but then that's the closest to my particular fixation.) Google Base looks to me like the beginnings of a general purpose content management platform. As Phil Windley and others point out, what is missing is a simple, flexible API. Charlene Li at Forrester puts it well and looks to the longer-term implications:

But as comments to my original blog post on Google Base point out, just having the data isn't enough - you've got to be able to DO something with the data and no, just being able to search the information isn't nearly enough. And this is where I think Google is on to something very big. At its core, Google Base is just one very big database of highly structured information. I can't believe Google will just let it sit there, and instead, will develop APIs on which developers can build applications, in much the same way it allows them to create mash-ups around Google Maps. So rather than have to figure out, build, and maintain lots of different applications, Google will allow developers to access the information, on the condition that the applications be "Base enabled".

Does this sound familiar? Microsoft's Windows Live and Office Live are built on a similar premise (albeit sans database -- at least for now) where Microsoft supplies the backend infrastructure and hosting, some tools and data, and a place where developers can market their applications to users.

And in her conclusion she points to the issue that has been bugging me the most - why is there no security model in Google Base?

One last thought on Google Base - right now, anything I post to Base is public, but I may want to keep something private, or accessible to a specific social network. At some point, Google is going to have to allow users to set up these permissions, which adds a layer of complexity to searching. If I'm doing a search for a particular recipe, and I have permission to look at my extended family members' Base content, Google would have to parse out that information in real time. Not an easy feat, at least on the surface.

So why build a content management system with a rich(ish) metadata model but no security model? It seems like a really strange decision because these have always been the two fundamental components of content management. Sure, building an ECM-style, role-based, hierarchical security model that works at an item and container level might have been a bridge too far, but basic private/public flagging and a simple "invite to share" mechanism must be in their kitbag.

I think this omission is very deliberate. Part of the decision may come from the general restriction of functionality in this first release. This in itself is pretty cunning - put out a tool with limited functionality and little explanation and see what the market says you should do next. Given the 60,000+ blog entries generated so far they should have plenty of ideas!

I also feel the lack of security is somehow connected to Google's slant towards personal rather than enterprise applications because without controlled sharing, GBase is pretty useless for companies. However, it is OK for sole-traders to post their wares on GBase and this is one of the use cases suggested by Google themselves (and the root of the eBay killer palaver). So this could simply be another reflection of Google's rather hypocritical "Robin Hood" positioning.

The other thought that occurs to me (and which I feel is more probable) is that the launch of GBase has been carefully calculated. Google's stated aim is to "organize the world's information". If they had added the security model, GBase could have been presented as a ubiquitous and infinitely scalable store for all the unstructured content (private, shared & public) generated by individuals and enterprises. Because it has a metadata model, the content can actually be "organized" rather than simply found.

But what would the public reaction have been? Would you trust Google to look after the entire contents of your hard disk or company shared file servers? I'm sure plenty would but paranoia about Google's ultimate potential power is growing. Maybe a softly-softly approach is just what Google needs at present.

So what does GBase actually do today? My first impressions are that the interface is pretty clunky (although given that Altien's entire focus is on making user-interaction with content management systems as efficient as possible I admit to strong bias). For my first use case I tried to add the latest Altien Document Manager Brochure. I created a new Item Type "Brochure" and started adding attributes: Company, Product, Version etc. I added some "Labels" (what's wrong with tags?) and the URL to the PDF on our website. I then tried to Publish it but my content was "Disapproved" because "Altien" is a misspelling. We had this same problem when we started with Adwords. Obviously those dictionary extensions are not in GBase yet. So I appealed for review (pointing out that Altien is our company name and trademark). More than 24 hours later it is still "disapproved".

Not to be deterred I thought I'd follow one of the suggested use cases and add a "Product" entry. Unfortunately the only products I have that are not Altien branded are my children so I entered the details for my two year old daughter and this time I passed the thought police first time. So far no one has reported this as a bad item nor have I had any offers. Others have pointed out that GBase will be chock full of spam in no time. Maybe then we'll see the security model implemented.