Twitter: this is because of single-key shortcuts 4 spam, star, archive, delete. Really good. 12 hrs ago

Jan

10

The Three Degrees of Personalization

Editorials, Facebook, Google, Insights, Social, Yahoo 

Facebook’s Beacon brought some spice to the tech community, which had been longing for some juicy stuff to blog/discuss/pontificate about since the iPhone’s launch several months ago.

Beacon is (was?) part of Facebook’s new online ad system which shared a user’s actions on other (partner) websites with all his/her Facebook friends. (If you rent a movie on Blockbuster, your friends will be notified via Facebook’s mini-feed). Scary, eh?

So in addition to the flak that Facebook was already getting because of the havoc its “Applications” have wrought, it was criticized soundly for making Beacon the insidious nightmare that it was. I kind of lost track of that story after a few weeks (the tech blogosphere, like most of modern society, has an incredibly short attention span), but there were reports that Facebook had rectified most of what was wrong.

At around the same time, Google unveiled what it thought was a whole new chapter in social networking - now, your shared items in Google Reader would automatically show up in the Google Reader feed list of whoever was on your Google Talk list (and you could see their shared stuff). Privacy advocates made a tremendous amount of noise yet again.

However, Dave Winer at Scripting.com had a word for those who screamed blue murder when Beacon surfaced:

[t]here are thorny issues here, but we want these companies to give up control of our information, and we don’t want them to be overly scared of public opinion as they do it…[but] most important, I want them to give me control of my data.

Indeed. Almost all web-based applications we use today are personalized to some degree. So inured are we to this personalization; we don’t even notice it anymore. Until something like this pops up. Why do we love some applications and hate some? When does a feature cross the privacy line?

Here’s a look at some popular Internet applications/services:

Application/Service

User info collected?

Personalization requested?

Personally identifiable info shared by default?

Amazon’s recommendations

 Y N N

iGoogle

 Y Y N

Google’s personalized search

 Y N N

Orkut/Flickr

 Y Y Y

Google Reader’s shared items

 Y N Y

Google News

 Y Y N

Facebook Beacon

 Y N Y

iTunes Music Store

 Y N N

YouTube

 Y Y Y

Digg/Del.icio.us

 Y Y Y

Gmail

 Y Y N

As we see, the only two applications that have run into privacy troubles (Google Reader’s shared items and Facebook’s Beacon) are the ones where personalization was not requested, but information was shared anyway. Then there are the annoyances like Google’s personalized search, that displays India-tailored results for me because it knows my location (and there are several times - more often than not - when I don’t want this). There are other beauties of design like Amazon’s recommendations - that give me personal, useful information without needing to know anything at all about me.

Making sense of it all:Internet application services seem to divide themselves into three degrees of personalization:

1.) Personalization based on viewed contentAmazon’s recommendations (”people who bought this also bought…”), eBay, Digg, del.icio.us. Here, user information in either aggregate or anonymous form is enough to offer a customized experience.

2.) Personalization based on purely personal informationMost of Google’s services offer this degree of personalization - iGoogle, personalized search results, Google News, Gmail. In all of these cases, the service provider (or application vendor, depending on how you look at it) has access to a ton of information about the user (think Gmail), but relies on that information bank alone for the experience.

3.) Personalization based on personal and group-level informationHere’s where both controversies have arisen, and, unfortunately, here’s where most of tomorrow’s applications will likely be slotted. Orkut, Flickr, YouTube and other community-based applications.

Concluding:An application can move from one degree of personalization to another. I can visualize Amazon, for instance, forming a community where individual book purchased can be used as recommendations (PQR in your friend network bought…). Or Google News, for that matter.Whether or not you land in a privacy soup will depend on where you choose to draw the line.

Some time later: What role does owning a lot of web real estate play in this game of personalization?

Post your comment