Thursday, July 14, 2005
Market Gaps - Web 2.0
Link aggregators are incredibly powerful ways to navigate the www. What I can't understand is why so many current players are foregoing this huge opportunity. Feeds (distribution) and tags (revelation) are an important half of the picture, sure, but so is semi-smart aggregation, which can provide a quick snapshot of what the community values now.
I use blogdex, as I have done for a very long time - because, amazingly, no one's built a better mousetrap yet, despite many, many attempts to do so (read: there is a big market gap here).
Technorati used to have a nice link aggregator, but now it only counts traditional news sources. Why don't they include blogs and other micromedia again? Most likely, they don't want to cannibalize their enterprise offering.
I think Google is the same story. Think about how easy it would be to expose the huge amount of info that passes through Blogger every day. My kid sister could code a rough and dirty version in about a day. Why don't they do it? Because it ends up costing Google a nice chunk of marginal revenue if users learn to quickly navigate through link aggregators, instead of slowly through the larger www.
Bloglines has a relatively open link aggregator that counts traditional sources and micromedia...but it's really slow, and cumbersome. By slow, I mean that the links decay too slowly - I don't wanna see yesterday's news.
There are a number of other competitors, but I'll only mention one more - del.icio.us/popular. I don't use it so much because it's noisy - because del.icio.us's user base, which makes up the collaborative filter, is like an extremely biased sample, most of whose interests are totally irrelevant to me. I suspect this will become more useful as del.icio.us penetration becomes a bit more rounded.
Some general points. I would give users control over the decay function - how long it takes for links to not get counted any longer. That way, I can set the freshness of my own links. Second, I would get a little vertical. It shouldn't be that hard to extract contextual information to sort links into basic categories. Third, interface matters, a lot. A big part of the reason I use blogdex is that I have little time to trawl links, and I can get a lot of information very fast just by glancing - unlike Memigo, where I have to concentrate to extract relevant info.
Have you tried populicious?