February 09, 2010

Steve Crossan commented (on Facebook, of course) that he likes the new FB design.

I couldn't agree more. This is the first time I think Facebook have got it absolutely right. Of course, I guess this is the result of them buying in Friendfeed. But I'm intrigued ... was it the technology they needed from FF? Or the understanding of feed-based user-experience?

Whatever it was, if FF was a material contribution to the new FB, the synthesis was brilliantly executed (in less than six months!). And the result is a triumph. It's disturbingly compelling. Facebook have again shown themselves to be very clever at learning from others, at adapting to changing fashions, and taking their large base of existing users with them. (In this, they're reminiscent of Microsoft in their heyday.)

If I was running Twitter, at this point, I'd start to seriously worry. They aren't going to grow their network at Facebook's expense. Whereas the opposite is highly plausible. So what do they do now?

On first glance, Google's Buzz, reminds me of Microsoft in all the wrong ways. It looks like a "me too" clone of Twitter / Facefeed that exists for no other reason than that Google are frightened that there's a space they don't dominate. And now they want to muscle in on it.

A couple of things may change my mind :

1) Dion Hinchcliffe says Buzz has loftier goals. (Then again, doesn't Bing?)

2) At this point, the fact that it's hung off of Gmail is merely not dumb, as opposed to actively smart. Nevertheless, maybe this is the beginning of the wavifying of Gmail. In which case, that's an interesting evolution to watch. Gmail is a very nice upgrade of the standard email client. It could potentially turn into "the next Outlook" if Google do the right things with it. [1]

OTOH, lose the f***ing brand! "Buzz" is truly horrid; like a tired celebrity gossip page in one of those free newspapers you find on the tube. Except worse.[2]

But back to Dion's article. If the main claim of Buzz is that it brings better algorithms to the social web, then I think we need to be highly sceptical.

Firstly, the attraction of the social web, may not really be its data-processing efficacy. Yes, we all go round saying that it *is* useful. But it's also, necessarily phatic. Strip out that phatic, community forming flava, reduce it to factoid sharing, and your social network may become as charmless as dmoz.org. For many, FB will always be about little lost vampires turning up in your Mafia pizza emporium. And Twitter would never have found its way into my heart without Chinposin' Fridays.

Secondly, while Google are pretty smart at algorithms, returns diminish rapidly in hyperlocal social space. PageRank is not a genius algorithm : it's a clever heuristic based on some statistical characteristics of large datasets. By definition, neither the hyperlocal nor your meaningful social-network are anything like large enough for simple statistical algorithms to deduce much of any significance from. To add some kind of real value to that, we're talking "A.I. Complete"

As a comparison, think of it this way : Google is allegedly an algorithm company. Gmail has made a much better email experience. But Google have never been stupid enough to pretend that they can prioritize and schedule your email. Are they really going to add much value to a slurry of 140 character tweets?[3]




[1] With the right calendaring / feed-reading / tweeting / waving) Google could write a decent downloadable client (built on Chrome technology) as a direct replacement for Outlook, with their eyes shut. Why do this? Because it would signal that Gmail is ready to fight that battle.

[2] Buzz is not going to be Facefeed. It's not going to be a fun, populist social network. And Google already have too many brands; they don't need more. And they particularly don't need more failed attempts to be "groovy" that make them sound as desperate as Microsoft.

[3] Your tweet-stream is good because you've already chosen who's got a high-enough signal to noise ratio to pay attention to. And if you're wrong, you tolerate it 'cos it's your mistake. No way will a Google algorithm make that decision for you better than you could. And no way would you trust it to.