September 2004 Archives

The Enron email dataset is 619,000 emails and is available for download - a great set of real data for social network analysis, email classification,  text analysis, etc.

From Harvard Business School Working Knowledge: Four Ways to Pick a Winning Product

There are four benchmarks for predicting the success of your product or service, according to this view from Strategy and Innovation. Here's how a few well-known products tested out.

The article proposes that there are 2 categories that new products have to excel at to displace existing products - purchase motivation and purchase barriers. To win, a new product must:

1. Providing high purchase motivators
   A. It must be less expensive than existing products (lower price).
   B. It must provide better features than existing products (greater benefits).
   
   
2. Eliminating purchase barriers
   A. It must not have any switching or adoption costs (easy to use).
   B. It must be readily available (easy to buy).

Lots of questions and thoughts spring from this:

• Does the same logic/approach apply to blogging inside the firewall ?
• Are the cultural issues that many of us were talking about about at Blogwalk 4 covered within this approach ?
• How can we combine this approach with David Wilcox's excellent diagram ?

Desiree Gosby is blogging at Songlian. I met Desiree and her partner Omar Green [where's your blog Omar ??] at BlogWalk 4. Nice blog - great pictures from her Nokia 7610, and being a hardcore geek, she don't need no Flickr [like us poor mortals]. But she also has a cool Bluetooth Nokia pen that's allowing her to write stuff on paper, have the resulting image Bluetoothed over to her 7610 and from there to her weblog.

Had a good conversation with Omar, and they seem like good people.

I've been using audioscrobbler for a while now [and to a lesser degree its sister site last.fm]. Of course, as luck would have it, both sites are down as I speak [they've been having some hardware issues over the last couple of months].

Still, both sites are very much about connecting you with people with similar tastes, and thus discovering new music that might appeal to you. Audioscrobbler uses plug-ins for most popular media players to record what you're listening to - no further user action required means that a large body of data is accumulated easily, and with no extra user effort required. You can then see people who have similar tastes, etc etc.

If that's not enough, there's a nice network visualiser that works on the audioscrobbler data here, and needless to say, there's quite a community developed around the site. Now, if they could just keep it available, which is actually another big challenge for services such as this [and I'm including the likes of bloglines, technorati and del.icio.us in this category]. Their success depends upon large scale take-up, and that large scale take-up inevitably brings the service to its knees.

Allofmp3.com at first glance looks like another dodgy music download site, destined to be shut down and possibly leaving you open to legal attack from the music industry rottweillers. The fact that it's based in Russia adds not one jot to its credibility.

However, in this case, looks are deceptive. They're legitimate - they pay fees to the Russian music copyright society and under Russian law they are thus allowed to provide music for dowload.

The service itself is astonishing - you can choose format [mp3, ogg, aac, etc], quality [64k to something very high], and even download lossless versions of the music. And then you come to the price, which for an album encoded at 192K typically works out to about 50 pence, at which point you're downloading stuff like Sinatra's Greatest Hits, just in case there's ever an occasion that needs it.

Oh, and just to tie it back to the original theme, as you select music to download, the site tells you what albums people with the same taste as you downloaded. Given that the risk is so low [at 50 pence per album you can afford to download a few duds], I'm just a clicking downloading fool at the minute.

From Mopsos

Ultimately, the economics of knowledge has something to do with the economics of love.

Haven't been posting much recently, but when one sees a phrase like the above, how can one resist ?

And it's an important point - you can setup as much infrastructure as you want, incentivise the hell out of people to produce and consume knowledge, but unless the emotional component is there, then you've a long hard battle ahead - if you can't stand someone or don't respect their skills or knowledge, then you're not going to learn/connect/share.

This got me thinking about a phrase from software engineering guru - Steve McConnell  -"Don't flip the bozo bit". Flipping the bozo bit basically means deciding after a number of interactions that the person you're interacting with is a bozo, and no useful information and/or knowledge will come from this person - they've been written out. Does this emotional decision outweigh any market or infrastructure initiatives ? I would hazard a guess that apart from the most coldly rational of us [and any Vulcans out there], the answer would be yes.

Oh, and of course, there's BlogWalk 4 - "Blogging within the firewall". Very much looking forward to this - see you all there !