02 April 2008

Web 2.0: 10 minutes braindump

Idiot: a person who incessantly argues about unimportant issues. Such as: Is Web 2.0 a buzzword or a profound new vision? Well, establishing beyond doubt that Web 2.0 is indeed a buzzword would be just about as useful as establishing that it is a stroprantzer.

Let's say that we want to make Internet more useful. Let's assume that existing websites have various degrees of usefulness and that users favor the useful ones. Question: Can we find a good predictor for the usefulness (and hence popularity) of a site? My wife proposed: Useful sites offer services based on information gathered from the users. Proof by examples:

  1. YouTube offers
    • search for video-clips
    • video-clips
    • lists of similar video-clips
    based on
    • video-clips submitted by users
    • comments submitted by users
    • users' navigation patterns
  2. EBay offers
    • cheap products
    • buyers for stuff that your friends don't want
    based on
    • what users want to buy/sell
  3. Amazon offers
    • book recommendations
    • books
    based on
    • what users buy
    • second-hand books offered for sale by users
  4. IMDB offers
    • movie recommendations
    • movie ratings
    based on
    • user ratings and comments
  5. Google offers
    • search
    based on
    • stuff users put on the web

All these sites make it very easy for users to offer data: to tell YouTube what videos are similar you just browse the site, to put a web-page on Google you do… nothing.

Could someone inclined to make money do so based on these observations? No idea, but here's how I would start: What types of sites I know that do not follow the description above? Let's see:

  • home-pages of scientists
  • web-sites that present private accomodations in one area ("pensiuni" in Romania, B&B in Ireland and UK)
  • booking of airplane tickets
  • local renting (Ireland has Daft, but other place like Romania don't)
  • local movies (per city? country?)

Now let's say we want to make a web-site (2.0!) for movies playing in Bucharest (to replace the horrible thing available now). We should:

  • have a search box: people expect it
  • let users comment: we can use that to improve the search!
  • secretly use external data sources to boot-strap your web-site, but don't say so: write scripts to fetch data from existing web-sites, make deals with off-the-web information providers and so on (BTW, secretly means that you don't tell the users how you get the data, in case you had doubts about the meaning :P)
  • give recommendations and ratings
  • give comprehensive data about playing times
  • make it very very easy for users to provide information: movie ratings, comments, playing times (but, at least initially, do not rely only on them — see the point on using external data sources)

Above all, make it work flawlessly and fast (even test from dumb operating systems like Windows!) and have a very very simple user interface (a person that is >50 should find it easy!).

Ah! And, it may give you a head start if you don't mention to others what you are doing until you launch a beta. (If you try this method and make money then don't forget to send me some royalties.)


Ovidiu said...

Treaba cu filmele deja exista la altii de mult, si se foloseste la greu de vreo 10 ani probabil. Se cheama movies.tahoo.com. Bagi adresa, ai cinemaurile cu filmele, dai click pe film, ai note/comentarii de la popor, si separat de la critici, ai trailer, ai di tati.

rgrig said...

1. It hardly works flawlessly on all platforms: it insists on the latest version of Flash, which means that I see a big white blank page when I go there.

2. Does it have a local section for Bucharest? If not, then one can simply use imdb.

Please use English here.

Post a Comment

Note: (1) You need to have third-party cookies enabled in order to comment on Blogger. (2) Better to copy your comment before hitting publish/preview. Blogger sometimes eats comments on the first try, but the second works. Crazy Blogger.