The context web

In preparing some recent presentations I have been talking about three primary ways of experiencing the web which emerged successively and continue to work together. Here I will call them the site-web, the search-web, and the context-web (alternatives might be site-centric, network-centric, and user-centric).

Site-web. Our early experience of the web tended to focus on individual websites. Enumeration of websites was common, in lists, directories and guides.

Search-web. Attention soon shifted to the network of websites as a whole and search quickly emerged as central to our web experience. Google rose to prominence based on the insight - expressed in its pagerank algorithm - that not all websites are equal.

Search is now our primary way of finding resources and navigating the web. This was underlined, I think, by the introduction of the single box in the Chrome browser for both URL entry and search. A while ago, I was looking for something with my son. He was amused that I was typing in a longish URL - search is how he goes to everything, even where he knows the URL.

Context-web. It seems to me that we are now moving to what I call for my purposes here the context-web. Search remains important but is no longer enough. We expect services not only to know about resources on the web, but also to know about us. We are seeing servces contextualised by their knowledge of people using those services and their relationships. Think about how Google is incorporating location- and social-based results in their searches. If I search for cameras, I will be shown mapped results from near Dublin Ohio and I will be shown what people in my 'social circle' are saying about cameras (my 'social circle' is what Google knows about people in my social networks - Twitter and so on). As the recent controversy about Buzz showed, Google knows quite a bit about me through my use of is services (I regularly use Search, Reader and Gmail, and dip into other services).

Facebook and Twitter also know something about me. As does Microsoft, where I have a Zune profile where I can 'friend' my son who has an Xbox profile (and indeed I can see his Xbox friends).

Much of the current competition we see between the big web/media companies relates to the management of our context. Google, Microsoft, Twitter, Facebook and Apple would like to be able to mobilize our context to provide more valuable services.

Comments: 4

Mar 10, 2010
ericrumsey

I think part of the appeal of using Twitter for finding links is that it gives context to links -- The more RT's there are, the more context. Twitter Search doesn't get a lot of press attention, but I think it's popular because it combines the "Search-web" and the "Context-web." How much more interesting it is to search a hot topic in Twitter search, which shows how many people have tweeted a subject (and who they are) than it is to search in Google News and get an impersonal list of article titles.

Mar 10, 2010
Lorcan Dempsey

Interesting point Eric, and it is presumably one reason Google is integrating Twitter results into the main search.

Mar 11, 2010
ericrumsey

So far, what's in Google is a pretty pale imitation of Twitter search -- Seems pretty random, unhelpful.

Mar 31, 2010
Oleg K.

What sticks out about the context-web as you wrote about it here is how it is built upon the age-old premise that you can tell a lot about a person by who their friends are. In terms of getting the most out of the web this may or may not be a good thing; it all depends on whether one's "friends" relate to the query. Of course, including other variables like location and past searches may help guide relevance, still, I wonder what the limits of social search is?

Leave a Comment