web2.0 notes day 1, morning (October 2005)

Notes from Morning sessions of Day 1 (Workshop day) of the 2005 web 2.0 conference

Some random notes

• registration was surprisingly backed up, ended up missing most of the first session I wanted to attend (VC2.0)
• Attended or planning to attend Identity 2.0, Web 2 Ad Models, AttentionTrust and Launch Pad.
• place is packed
• Nokia's Espionage ringtone is very annoying during a session. No, it wasn't my phone.
• if you're taking pictures with a digital camera, that shutter click sound is extremely annoying and can be turned off before you disrupt a session with it
• MS Powerpoint 1, presenters: 0.

Identity 2.0 session

• that identity is what you say about yourself
• no walled gardnes, data silos
• identity need for web services
• yaturp: yet another tedious user registration page
• then the laptop implodes
• SXIP: simple extensible identity protocol
• GUPI globally unique persona identifier
• GUPI is digitally signed
• sxip provides single signon
• protocol includes messages for exchanging property information
• reference Kim Cameron's Laws of Identity http://www.identityblog.com/
• SXIP 2.0 provides anonymous GUPI
• another thing to add to yasoboto
• and if you have a Nokia, turn off the Espionage ring tone while you're in a session
• wifi seems to be dropping every 30 seconds for 10s.
• so...how does SXIP and x509 interact?
• and comparison with OpenID
• SXIP 2.0 draft out, put out for standardization in 2006
• sxore -- reputation management for user generated content
• sxore -- http://www.sxore.com
• sxip community www.sxip.org
• should close door off by presenter
• re: standards, why aren't they , how are they interacting with various existing standards? In some ways they're not...many of the existing standards are huge, ungainly, and haven't been successful. They make use of some of the technology but at the implementation level, not at the user level.
• who certifies whom? that the user, homesite, and membersite make those decisions, SXIP doesn't certify any of the various sites (unclear about root sites).
• WRT SXIP vs Passport: that MS has ws-identity, which is not a replacement for passport, using some sort of transport mechanism.
• Notion of user centric identity

followup questions

• how do the member sites or root sites know that the person is who they say they are
• Infocard?
• question: under whoms TOS does your data exist? get stored?
• how do you audit and control who's accessing data?

§

Advertising 2.0

• Panel: Jeff Jarvis, Brain McCandless(?) CEO Aquantive, Dick Costolo (CEO Feedburner)
  Mark Pincus, Co founder tribe.net,
  Chas Edwards Federated Media,
  Matt Cutts, Google.

• session is packed, packed packed, approx 150 people
• advertising in the distributed world
• to get advertisers to "support us"

Measuring

• cookie anything
• reporting
• demos/behavour
• spam killers
• new measures
• trusted networks

Serving

• Open source call
• Sell-side advertising
• Ads in all media

Managing

• ROI

Notes

• more important in web 2.0 for ad networks to provide feedback to content providers and publishers other than here is the ads and the money you made
• we all need to focus more on experimenting
• need for audience determination
• google not good for precise targeting, blogads and adbrite not good for scalability
• that mathmatics of internet bring rationality to marketing, but advertisers and marketers are terrified of such rationalization
• cutts: ads can be essentially a service, like search. they can improve the user experience. If you view ads as a necessary evil that will color your perception of ads.

• tim hanlon, starcom: that what the online folks are talking about can inform the other media. Need better metrics, coming along. Starting to influence how other media forms are evaluated. Expectation of same time of metrics for other forms of media. that what is solved in online space may well solve problems in other media. that which gets measured gets bought.
• tim hanlon: engagement is all the term buzz on Madision Avenue.
  Not how many eyeballs but how deep or how related or how interested
  a consumer was in the ad, ad content, etc. MA has no clue how to measure engagement. Something far deeper is needed, CPC is where advertisers are going towards but there's something else between brand advertising and CPC to measure engagement.
• that internet about targeted reach as contrasted with television
• reach new audience and engagement
• critical mass is an issue
• profiling audiences a thing, who's linking where and where is this conversation being discussed
• cutts: allowing experimentation is big. opening the ad network is definitely a priority (for google?). Important to have freedom to tinker.
• jarvis: problems with RSS as content delivery mechanism for advertising
• costolo: that people who have subscribed to rss have expressed interest by subscribing to the feed.
• that feeds should have audience targeted advertising vs content targeted advertising in traditional pages.
• Fred Wilson: on Reed's law (derived from metcalfe's law) that the value of a network increases exponentially because each ...ok, I lost it. Something to the effect that while Metcalf said the value of the network increases as new nodes are added, Reed(sp?) says that each node can itself be a network, so the value increases ..what, quadratically?
• ross (mayfield?) that we have to stop measuring the impressions and start measuring who is impressed.
• jarvis: that advertisers should hand over control over targeting
• problem: what to do if the only differentiator is the brand? Is the brand the trust point?
• notion of peoplerank vs pageranks
• users need to get something out of providing all this information
• fred (nolan?wilson?) that some people find relevancy targeting creepy
• costolo: does google become a business platform like ebay or something else?
• Cutts: that competition is essential, if it kills off something like adsense then the "world" is a better place
• cuts: that goolge targets the interest, not the user
• need for an expiration of interest...that you may be interested in
  something ONE time and never want to see it again. To give the user
  the ability to expire interest.