The Copenhagen Project

Techmeme and Google News shares an interesting common characteristic: They are both entirely edited by algorithms that surveil a predefined list of other medias to determine what’s currently ”hot news”, and it uses this information to construct it’s front-page. The two sites are virtually functioning in the same way with one other important difference: Techmeme isn’t prioritized whereas Google News is, and Techmeme therefore ‘feels’ more blog-like.

Notice that out model holds these two sites as having the property of ”Editors” (Or as I called it originally: Centralized editing). It could maybe seem as if this should be ”voting” or ”decentralized editing” as the sites draws on data from a great number of sites – but the thought behind this property is not to tell on basis of which data the editing is decided, but where ”the power” in the editorial decisions resides. Bloggers that are included in the list of Techmeme’s sites aren’t consciously trying to influence the editing process of Techmeme, neither are the list of medias that Google News use. Users don’t have the power to tweak the algorithms behind the sites and therefore can’t influence what’s on the front-page in the same way users of e.g. Digg can do.

We might have to change the name of this property to make it more obvious what it’s all about – as e.g. voting isn’t the only way that users can participate directly in editing of a site. Structurally Techmeme and Google News are very much alike. What remains is the differences in the topical focus of the sites.

The only fundamental difference that’s left is the possibility for users to customize their Google News front-page – an aspect our model doesn’t yet take into account. We might add the properties ”Customizable content” vs. ”Fixed content” to the model.

Digg and del.icio.us looks and feels somewhat alike and are definitely very different from e.g. The New York Times. Let’s try to compare them, looking at the front-page of each service respectively.

They do indeed have a lot in common, but also differ in at least three crucial points.

First, whereas Digg is very much build and staged as a place for conversation and debate, del.icio.us has no build-in conversation functionality. While Digg is a community with interaction, del.icio.us is much more a product of it’s fundamental staging as a personal tool, where conversation isn’t possible.

Secondly, whereas the basic idea of Digg is to actively engage people in a process of editing a front page, the basic idea of del.icio.us is to enable it’s users to easily file their bookmarks. The del.icio.us-front-page is therefore rather different from Digg - it’s users aren’t consciously thinking in terms of editing. While Digg’s front page is constantly influenced by voters, the del.icio.us front-page is a map of user activity at it’s primary service. This is reflected in the type of material found on the front pages; digg.com’s objects are often characterized by being “news-worthy”, whereas del.icio.us’s objects usually are less timely, and a great deal of the objects are reference material.

Thirdly, Digg utilizes human editors (the users), whereas del.icio.us uses an algorithm to decide the content of it’s front page. It seems to be a rather simple algorithm, but it’s there. It’s impossible to say how it’s tweaked though.

So there you have it: From a media perspective Digg and del.icio.us differs in the areas of social interaction (conversation/no conversation), in the type of editing (editors/voting) and which type of editor that edits (human/algorithm).

Our model has been cleaned up a bit. We now have 8 structural categories that we think apply to all medias:

Organization: Non-profit / for-profit
Object: Link / Original content
Presentation: Prioritized / Flow
Conversation: Conversation / No conversation
Editing: Editors / Voting
Editor type: Human / Algorithmic
Frequency: Continously / Regularily
Object provider: User / Editor

To illustrate where we are heading with this model let’s take three different types of media: The New York Times (offline version), Global Voices Online and NewsTrust. Just by looking at the frontpages we can see that these three medias are quite different in structure, but how are they different? Our model gives us this answer:

NewsTrust and the two others are fundamentally different, because NewsTrust doesn’t create any content it self. What is created on the site is meta content like ratings and comments. The difference is so big that the only shared characteristic with The New York Times is the prioritization of news and that the editors are humans (in contrast to an algorithm).

It’s also worth noting that Global Voices, which basically is a group weblog, has quite a lot in common with The New York Times - yes, blogs and newspapers are different, but they aren’t that different. The model can be used for making a lot of interesting points, which I will start doing by analyzing different medias in the coming time.

As it is now, the model is capable of describing 512 different types of media. We have tried to create something that makes sense across a wide range of media, from traditional offline media, to online news sites, but also to services like Twitter and del.icio.us.

What we need now is some feedback from you. Does it make sense? Is there categories we should add? Can you see how this could become useful in idea generation? What do you think about the names of the media elements?

A medium with the characteristic of no conversation presents it’s item in a way that the user can only consume. This includes traditional newspapers, books, television, radio etc., but also news sites as nytimes.com, del.icio.us, Twitter and some blogs.

Online there’s a range of mediums which enables users to hold conversations about the objects that are central to the site. This includes most blogs, sites as Digg.com, Jaiku and online forums.

Analysis

The New York Times: No conversation
del.icio.us: No conversation
jezebel.com: Conversation
boingboing.com: Conversation
newstrust.org: Conversation
Monocle Magazine: No conversation
reddit.com: Conversation
digg.com: Conversation
technorati.com: No conversation
techmeme.com: No conversation
Google News: No conversation
daylife.org: No conversation
Newsblogger: No conversation
huffingtonpost.com: Conversation
talkingpointsmemo.com: Conversation
RSS reader: No conversation

A media can be edited by a central group of people, e.g. the editorial board of The New York Times or by the author behind a blog. The editing is restricted to a set group of people, or in the case of algorithms, on a set rule of filtering where no conscious outer influence is possible.

The opposite of centralised editing is what I think is best coined in the term “distributed editing”. By this I mean that the editing process has been opened up to outer conscious influence in some way or another. An example is digg.com, newstrust.net and others, where the editing is done by the community in a system of voting and rating.

Thoughts

This is not two definitions set in stone. digg.com uses algorithms to determine which items are promoted to the frontpage, but as the main determiner is the userbased votes I would label it a media that has distributed editing. techmeme.com on the other hand has centralized editing, as it determines it’s news on the items, that a chosen group of weblogs link to. The authors of these weblogs aren’t consciously trying to influence whether a story hits techmeme.com or not. Another example of centralised editing based on measuring is the del.icio.us-frontpage.

Analysis

The New York Times: Centralised editing
del.icio.us: Centralised editing
jezebel.com: Centralised editing
boingboing.com: Centralised
newstrust.org: Distributed editing
Monocle Magazine: Centralised editing
reddit.com: Distributed editing
digg.com: Distributed editing
technorati.com: Centralised editing
techmeme.com: Centralised editing
Google News: Centralised editing
daylife.org: Centralised editing
Newsblogger: Centralised editing
huffingtonpost.com: Centralised editing
talkingpointsmemo.com: Centralised editing
RSS reader: -