In the advent of the new Google Personalized Search, the subject of privacy and portability of personal data increasingly becomes a matter of debate. Are consumers aware or should they be made aware of the extent of data logging and do websites and online marketers need to take responsibility to create more awareness for consumers?
When we use a service such as Facebook or Amazon, we are not tuned into whats happening to our data once we’re out to lunch. Do we care if they use this data somewhere else in the meantime?
MySpace’s Mike Barret, chief revenue officer, explains how they use user data to make sure that ads are targeted in a more efficient manner.
‘We test the click-through and refine the algorithm to make certain that the target audience, in this case fashion aficionados, are truly going to perform better than if you just ran a fashion ad on run-of-site’.
What this means is, that they are closely monitoring, logging and analyzing all user behaviour. Does this user show behavioural patterns? Which interests are shared by connected friends? Imagine yourself standing naked in front of a panel of researchers who monitor your every move and reaction and take notes on your behaviour.
I don’t think users understand how much information is left behind and logged and indeed if someone else profits from that information. Nowadays metrics are no longer limited to traditional statistics such as page views, visit duration or tracing browser history of an IP address. They are technically complex analysis tools and not your average net stats applications.
Greg Linden, founder of personalized news service Findory and author of Geeking with Greg, commenting on Google’s new personalized search explains: “Searchers do not have to do anything explicitly to use it; it is all implicit. The current Google Personalized Search is likely to be using the same Kaltix technology (acquired by Google in 2003), building a high-level profile of you, then biasing all of your search results based on your long-term behaviour.”
So if you are a frequent user of Google’s services, rest assured that they are happily mining you probably 90% of the time you spend online. There is an off-switch, but first you have to find it and that can be a challenge.
I would like to share a short analogy on Facebook written by Chris Saad, an esteemed evangelist of APML and Open ID: ‘These shopping centres (referring to Facebook) are not like real shopping centres. They let or invite you in, you will form friendships while you are there, but they won’t let you leave together. They remember every purchase you make, but they won’t give you a receipt. They sell you plenty of stuff, but what you buy looses its value as soon as you leave. These shopping centers want to lock the doors and trap you inside – they don’t want you to go home’.
What Chris is trying to tell us is that these ‘shopping centres’ create profit from your ‘AttentionData’. This is a specific term used by APML (Attention Profile Mark Up Language) evangelists to describe data that represents what you have paid attention to in the past. It is also known as your ‘AttentionStream’.
A recent report by market analysts Bear Sterns Inc., claims that sites such as Facebook, who grow extremely fat by your enthused production of user generated content (UGC), is not a fad. In fact YouTube, Wikipedia, Blogger, Digg, MySpace and Facebook currently make up about 12% of total internet traffic in the USA, this does not include x-rated user generated content sites obviously, as Bear Sterns likes to be conservative not only with their figures.
Privacy concerns regarding storing and using past behaviour are very real, and sophisticated personalization of content is an inevitable evolution. Personalized search doesn’t just retrieve basic url’s from your behaviour on line. The new systems are actually on a quest to find the real you and before long it will know you better than you know yourself. The basic internet metric and analysis system of page views, hits and visits is very nearly quite dead. Marketers especially know that as soon as they can build a record of everything you have paid attention to in the past they can start to build what Chris Saad calls your ‘AudientStream’ which is basically a channel of records of everything you might need or want to pay attention to in the future.
Subsequently Chris Saad opts another term, the ‘Attentstream’. This would be a stream of events (attributed to you for example) that signifies attention given to you by another. “Attentstream data could come from the tools which people use to pay attention to other matters, i.e. browsers, RSS readers, embedded players, flash player, adobe reader, the Second Life client. Because the tool itself does the reporting, it can report more subtle information which cannot be gathered on the servers of significant websites.”
Back to cross referencing, the Bear Sterns report states: “apparent to us, that as the supply of video content rises, value will shift from content producers to aggregators and packagers of content that can best aid users in finding content that fits their specific interests.”
For marketers this step up in the use of technology to build online behavioural patterns and create in depth individual profiles of consumers, is very useful in targeting consumers one-on-one, but technology such as this certainly is not infallible. As any marketer has perhaps learnt in past experiences with one-on-one marketing, targeting individual consumers can be very embarrassing if you get it wrong.
Basically people enthused about APML agree that Attention Data should be portable and users of websites should be able to choose if Attention Data is stored or used to profit others. Using an APML file, a sort of meta/meta file where all your data is stored, users should be able to allow or not allow access to this information over various sites and services on the web.
Currently there are few communities and services on the web which understand why open portability of attention data is important. Check out Me.dium.com in beta if you want a visual of having control over your own Attention Data and how you can control who has open access to it. Me.dium is developing a next-generation collaborative browsing technology that dynamically combines visualization and chat, it’s not there yet and its pretty basic but certainly an interesting case study on open portability of personal data.
In June 2007 Me.dium secured $15 million in investment for this APML project from Commonwealth Venture Partners.
Quoting an article on Read/WriteWeb Written by Alex Iskold / June 11, 2007:
‘So why do these VCs think that Me.dium has such big potential? It so happens that Me.dium has a chance to literally break outside the box that we described in our recent Evolution of Communication post.
So the $15M dollar question is what is Me.dium.com planning to do next? In a nutshell, the company is planning to spend the money on refining and scaling its sophisticated matching technology. The secret sauce here is in connecting people based on their browsing patterns. Doing so dynamically for hundreds of thousands of users is hard algorithmically and very intense computationally…..
….. Me.dium takes a strong stand on information - they say that the user owns all of it. To succeed, the company needs to continue to drum this beat and to act accordingly, which they have been doing so far.’
Particls.com is a unique news feed; it combines APML data with your choice of channels together with some manual input to build a personalized news tickers-service which runs on your desktop. Other examples of projects which support open portability and use APML applications are Engagd, Dandelife, Cluztr and iStalkr (using the Engagd API).
My estimation is that concepts for open portability of data, Open ID and APML, are distinctly subjects which will gain effect and relevancy as users become more aware to which degree their data is purged. Sites (let’s call them both beta projects for now) such as Me.dium and Particls, show and enable consumers to visualize data and give them tools to self-manage their own data across the web.
Obviously I would be interested in understanding how these sites will go on to educate their advertisers and when they are letting them in. It seems to me that marketers may not directly benefit from giving more control to users over their attention data. Personalized advertising is a no-brainer for any marketer in the current online ecosystem, and obviously advertisers benefit if users do allow access.
Even though it is early days, I assume evolution of open portability of Attention Data means we are moving towards a challenge for online advertisers. Not only will marketers need to be able to handle large amounts of complex data and embrace one-on-one marketing online with all its pitfalls, they will also need to find means to encourage users to allow access to the data if open portability becomes a fact of life.
In my opinion, any information or data that a user leaves behind in the virtual world, solely belongs to that person and we should be given the choice whether or not we allow others to mine or trade this data for their own gain.
Further reading
Study: Search engine privacy policies improving, published August 7th 2007 Computer World:
A Slap In The Facebook Follow-up