05.11.2007

Dion Hinchcliffe: What is Web 2.0 - successful web applications

[summary in my own words leaving out stuff]

Dion Hinchcliffe bloggt an 2 Stellen und hat eine IPTV-Show (die ich nicht finde, wo ist die?) und berät Firmen in 2.0-Fragen.
Die web20university.com gibt es auch.
(Wir haben gestren zusammen zu abend gegessen beim Geek Dinner, das war sehr instruktiv und) Gerade präsentiertv er bei bder web 2.0 Expo in Berlin zur Frage, wie man erfolgreiche Webapplikationen baut. Also: best practices.

Er will die Buzzwords weglassen und Butter bei die Fische tun.

Die Geschichte des Netzes:
- WWW: 15 Jahre alt
- neue Plattform: kommunikatinsorientiert durch Visualisierung
- 'boom and bust'-cycle: what works and what does not
- es besteht aus: Seiten und Hyperlinks
- "Does it work in the Browser?" -> Web (er sagt dann Flash gehöre nicht dazu, was ich nicht verstehe)
- "two way platform without an owner" (kein 'Lizenzsierung von 'Kanälen' wie bei TV)
- Web 1.0: wenig UGC, viele zentrale Publikationen, 45 millionen user global
- Web 2.0: 1 Milliarde User global. seit 2006: 51% User generated
- Partizipation statt Veröffentlichung
- Inhalte kommen vom 'Rand'
- belegte Geschäftsmodelle
- viele Interaktionsformen mit dem Web (Mobile etc.)

Was ist es?
- UGC
- AJAX
- Social Software
- Syndication + Services

Aber was liegt 'darunter'?

Name: "clearly apparent new trends in the way the web is being used" - "it is not a technology" - "it is somehting you can do" (even Ajax is not 'buyable'.

Websites 100% with user Content: Myspace (soaicl networking), Youtube (media sharing), Digg (news), Wikiupedia (knowledge), eBay (garage sales)

Webpage as software: Flickr, Google office, ...

Widgets and remixing in mashups, cretaing new Data-Combination.
-> HousingMaps.com, chicagoCrime.org

Blogosphere and sites like youtube see a massive peer production.

The 7 keys
- web as platform
- data as 'intel inside'
- end of software release cycle
- lightweight software and business mdels
- software n many devices
- rich user experience (ajax)
- collective intelligence

Design patterns
- long tail
- users add value
- network effects by defualt
- some rights reserved
- perpetual beta
- coperate, don't control
- architechture of participation (?: API?)

Network:
Der Wert des Netzwerks steigt quadratisch mit der Anzahl der Teilnehmenden.
(Metacalfe)
Reed: social networks are far more valuable than 'normal' Networks (by Factor 2).

Core Competences:
- online services w/ cost effective scalability
- unique hard to recreate info + control it
- trust users as co-develpoers and -creators
- harness collective intelligence
- target the long tail
- exploit leightweight business mdels ( self service, long tail, apps -> platforms, encourage unintended uses, control of had to recreate data

Why does it work?
- easy usage
- ... (missed it)


?: Wie erreiche ich einen Vorteil, wenn das Netz einen 'Plattform ohne Eigner ist?

Hinter allen erfolgreichen Produkten liegen große Datenbanken (und nicht notwendigerweise "tolle" Applikationen).
Warum sollte man die zweitbeste Datenquelle benutzen, wenn man die beste benutzen kann?
vgl. Amazon: Katalog angereichtert durch Kunden.

Opportunities:
- identity
- public calendering of events
- parental control data
... many more

Pro:
You get access to your customers and can leverage network effects.

Con:
You need to scale and scale and scale und yur reputation if fragile.

15 years of learning:
- simplicity now reigns technology
- = RSS, RoR, declarative programming, wikis, blogs, ...
(wird fortgesetzt)

Business model aspects:
- build open platforms instead of standalone apps
- spread your app beyond your site (widgets, APIs, , badgets, syndication, ...
- build on the shoulders of giants (Amazon etc.)
- do 'automated mass servicing' of low demand products (long tail)

Case: "Widgetbox": Billions of widgets are served to roughly 150.000 Sites.

Mobile:
- mobile experience is going to be better
- best experience: iPhone (except for Flash)
- mobile use expanding

Vertical or horizntal?
The Blogosphere :
a) real world capture (many devices),
b) blogplatforms, media sharing sites
c) aggregation sites (syndication, search: Google, Technorati)
d) syndication, rss readers (Bloglines, google reader, etc.)

The 'new' webpage:

- AJAX
- some are indistiguishable from desktop apps
- video and audio is NOT web technology
- Case: LZPIX in OpenLazlo


Challences of RIAs:
- page views are dead (because of Ajax, there *are* no new pages....)
- RIA-Content is not crawable -> no SEO (30% Traffic comes from search)
- link structure is lost (you can't link 'states')

Cases:
- XM Radio, satellite radio
-- put a channel on the web
-- turn it over to the users (wish a song, vote, play top 20)
-- huge win in visitors
-- in a very short time
-- users created a social network 150.000 users
- send the quersioners to the forum, it worked.

- GM: Cevrolet
-- Chevvy apprentice Campaign
-- Create a video commercial
-- top 10 win a car
-- "We don't censor unless it's not objectionale content"
-- 22.000 videos
-- GM responded t videos w/ answer videos - and they learned things (and there was no damage to the brand although there was criticism)

Software gets smarter the more people usage: implement an architecture of participation.
- ugc
-! database of intesions: get the 99% to generate value -> 'what they click on'!, search protocols, what are people emailing? what are they blogging? -> Analytics are too slow.

Gravitation:
- more content generates more users who generate mre content
- the 'draw effect' of the networks gets bigger

Success stories:
- sourceforge.org: open source
- Wikipedia: user contributions
- Digg: peer production
- MySpace: soail interaction
- Blogosphere: he two-way web in the large (the comments often form the bulk of info)

The videothing is at: e2tvshow.com

EOF


Facebook Kommentare