Please Build Me Something Useful: A Letter To Web 2.0 Developers

I agree 100% with the article.<p>Many applications are not designed at all to be useful, because in the last months the market didn't rewarded <i></i>just<i></i> useful applications, but a lot of fashionable ones.<p>Sometimes to build something of really useful is much harder than to build something like twitter, but it's like there is no longer a connection between how well a tool will solve a long time problem and how much buzz, users, money it will get.<p>Add to this the fact that web 2.0 users tend to be the same set of users using all the services, and trying a lot of new ones: it's ridicolous the number of people that are just happy to add a new website to the list of websites they already use everyday.<p>Instead to look for web apps solving problems many users are looking for problems that the just-released-web-app can solve. Instead to focus on writing interesting articles for their blogs this users will spend all the day visiting analytics, feedburner, sending messages in twitter about new services, reading tons of RSS, ...<p>Fortunately there is a parallel web 2.0 market of valuable things that save our time and make our life better. Developers seeking for a real business in the last years should look at this.<p>It is also very important to try to improve over the useful things already released. To make a better flickr or a better delicious or a better reddit can be ways more interesting than to invent something of totally useless just to be new and original.

I understand his point, but I think he supports it poorly. Of course a product that has 1 review is not likely to benefit greatly from that single review. But when I buy a product, or book a hotel, I look at dozens of reviews from other people and I get a good sense of the pros and cons of what I might buy. I can then gauge those perspectives against what is important to me. And I find that overall they are almost all correct -- in aggregate.

This must be (by far) the most interesting and unusual piece of text I've ever seen on ycombinator.

It seems like the author would find traditional, client-based applications more useful than server-side Web applications. The whole point of the latter is to use everyone's data to make software more useful, and if he doesn't care about everyone else's input, there's no need to connect to the Web.

He's missing the point. User reviews are as much about the reviewer as they are about the object of the review. All the same it's important to remember that there are <i>many</i> people like this guy.

I hear that The Wisdom of Crowds has good reviews...