The Stength of Weak cooperation:
A Case Study on Flickr
Christophe Prieur(a+b), Dominique Cardon(b),
Jean-Samuel Beuscart(b), Nicolas Pissard(b),
Pascal Pons(b)

(a)
LIAFA, UniversitØ Paris-Diderot
Case 7014
75 205 Paris Cedex 13

prieur@liafa.jussieu.fr
(b)
SENSE, Orange Labs
38 rue du gØnØral Leclerc
92 131 Issy Moulineaux
00 33 1 45 29 57 74
domi.cardon@orange-ftgroup.com

ABSTRACT
Web 2.0 works with the principle of weak cooperation, where a
huge amount of individual contributions build solid and
structured sources of data. In this paper, we detail the main
properties of this weak cooperation by illustrating them on the
photo publication website Flickr, showing the variety of uses
producing a rich content and the various procedures devised by
Flickr users themselves to select quality. We underlined the
interaction between small and heavy users as a specific form of
collective production in large social networks communities. We
also give the main statistics on the (5M-users, 150M-photos)
data basis we worked on for this study, collected from Flickr
website using the public API.
Keywords
web2.0, social media, flickr, folksonomies, self-organization,
social networks.
1. INTRODUCTION
Without trying (once again) to define what lies (and what does
not) behind the label Web 2.0 , one can at least d eal with the
articulation of individual self-production practices and
cooperation between Internet users, resulting in the collective
construction, on the WWW, of big, structured sources of
information made of a huge amount of individual contributions.
The development of the good-old Web had always be en driven
by a community ideal, and it had been built up mainly through
organized cooperation between voluntary participants. In this
context, the cooperation between members has often been
described as strong: mutual socialization and defined roles give
members a feeling of belonging to the community and a joint,
shared aim [1]. The successful growth of Web 2.0 services
(driven by Wikipedia, blogs, Flickr, etc.) has led to the
definition of a much weaker cooperation between Internet users,
detailed in [1].
As a result of the spread of self-production tools (image, video,
blog platforms, wiki, etc.), Web 2.0 services enable cooperation
between Internet users as a side effect of their individual
publication activities. The strength of weak cooperation’1 lies in

1
The expression is of course coined in reference to
Granovetter s S trength of Weak Ties [3].
the fact that it is not necessary for individuals to have a
cooperative plan of action or an altruistic concern beforehand.
They discover cooperative opportunities simply by making their
individual productions public. Public space is seen as an
opportunity for one s visibility, leading to relation making and
eventually actual cooperation with different levels of
involvement. And this cooperation can work in a very large
scale precisely because it is non-demanding. This weak
cooperation in a numeric space also allows cooperation between
small and heavy users which could be problematic in real life.
As a website for photo publication providing tools that enable
coordination, Flickr is often showed as a typical example of the
Web 2.0 [4]. The aim of this paper is to detail the concept of
weak cooperation on this example, showing the great variety of
uses, from plain stockpiling of photos to complex combinations
of all the functionalities, and how these functionalities serve
both individualistic purposes such as building one s notoriety
and altruist ones since they lead to a highly structured base of
photos with many user-generated procedures to select quality
from quantity.
We first describe in Section 2 the functionalities of the website
and the database we used for our study, giving basic figures on
the uses of the website. Section 3 deals with individual aspects
of these uses such as the variety of individual practices and the
necessity of playing the game to get acknowledged . This last
point leads to Section 4 where collective issues are addressed,
studying the user-created groups, mixing a both thematic and
social functionality whose role in the weak cooperation is
crucial since it enables users to invent their own procedures of
selection.
2. FLICKR, SYMBOL OF WEB 2.0
Although Flickr is among the original officially Web 2.0
websites2, its founders had not anticipated that it would become
a photo publication tool. Stewart Butterfield and Caterina Fake
(see [4]) initially intended it as a multi-player game, then as a
platform with chatrooms where people would share objects
materialized as pictures. But uploading of personal pictures took
more and more importance in the service launched by Ludicorp

2
in the exemplified definition seminally given by Tim O Reilly
[3].

in February 2004. The functionalities evolved to suppress
chatrooms and provide personal pages to users. After a few
months of growing success, Flickr was acquired by Yahoo! for
reportedly $30 million.
The ability of the creators of Flickr to follow the actual uses of
their service was a key to its original positioning and thus to its
success. It came at just the right time, combining the boom of
the sales of digital cameras, the growth of social networks
services and the success of blogs, for which Flickr soon
provided posting tools.
Some studies have already been done on Flickr. The history of
the site and its emblematic importance in the web 2.0 paradigm
has been introduced by Cox [4] and Van House [7], while
Marlow et al. [8] present Flickr as an example of Folksonomy
systems. But few studies are based on extractions of Flickr
database. To our knowledge, the only large statistical analysis of
Flick data has been done by Kumar et al. [9] at Yahoo. They
present a series of measurements of the evolution of the different
components of Flickr s relational structure. In this seminal work
they have demonstrated that Flickr (and Yahoo! 360) is
composed of a growing giant connected component (59.7% of
users at the end of the studied period) that represents the large
group of people who are connected to each other through paths
in the social network. Beside this giant component, they
describe a middle region of less connected users, and then
isolated singletons. While this structure is characteristic of large
networks, they show that the proportion of the middle region is
constant over the time, taking 1/3 of the users. In another
context, Lerman and Jones [10] extracted small samples of data
from Flickr in order to show the role of contacts for browsing on
the site. The most important part of the studies on Flickr deal
with the analysis of the evolution of photographic practices [7].
In an examination of digital photographers photow ork
activities [11], Miller and Edwards [12] have show ed that for
some people, Flickr supports a different set of photography
practices, socializing styles and perspective on privacy than
traditional photo amateurs. Our study comforts this idea that in
transforming amateur practices in a public activity, Flickr has
proposed a new paradigm for amateurs in which reputation and
visibility can be built by the intensity of the communicative
involvement with Flickr functionalities. Since Chalfen [13] early
book about the Kodak Culture of amateur photograp hy, the
rise of Internet-based photo-sharing has strongly affected
domestic practices of photography. In Kodak Culture, a small
group of persons (friends and family) share oral stories around
images with others. In the new culture of image c alled
"Snaprs" (a reference to the missing "e" in Flickr) by the authors
photos are used to tell stories with images, rather than about
images as with the home mode [see also 14, 20]. In this new
context, photo is not a story shared with closed relatives, but a
large-scaled conversation shared with people that participants
don t know in real life. Our study shows that Kodak and Snaprs
cultures coexist on Flickr platform, but that Snapr users lead the
community.
2.1 Main functionalities
Photos are the center of Flickr s activity. Users can index them
with tags (freely chosen keywords), post them to thematic user-
created groups, and put comments to them. Only the owner of a
photo can post it to a group, while any user can tag and
comment other users photos. Users can also mark as favorites
other users s photos.
Users have to register to a group to be able to post photos to it
and users can mark other users as contacts.
Basic membership is free but has some limitations with respect
to a paying so-called pro account (only the last 200 uploaded
photos of the user are displayed, the user can only create three
sets, and the per-month upload bandwidth limit is lower3.)
2.2 Harvesting the data
During Summer 2006, we have used the Flickr public API4 to
extract all public data concerning the five functionalities listed
above (tags, groups, comments, favorites, contacts). For users,
only the identifiers have been stored (no personal information)
and for photos, only identifiers and titles (of course not the
photo itself).
The extraction was done (in Java) by iterating on each user id u,
to get all contacts and (public) groups of u, and by iterating on
each (public) photo p of u, to get all comments, tags and
favorites of p. Another iteration was then done on each group g
to get the list of photos posted in g.
2.3 Basic figures
By definition, private photos are private, thus un reachable by
the API. However we can give an upper bound for their quantity
since the ids of Flickr photos are numbered by upload order5.
For instance, we have in our photo base the ids 222851183 and
222851185 but not the 222851184. The latter is thus either
private or has been deleted. By this mean, one can claim that
private photos are not more than 33%6 (since the ids that we
have in our base cover 67% of the range). In the rest of the
paper, only public photos will be considered unless specifically
mentioned.
Figure 1 shows the distribution of photos, which is of course
highly heterogeneous (although technically not in power law),
20% of the users owning more than 82% of the photos. One
counts 156 840 996 (public) photos for 4 788 438 users,
which makes an average of 33 for all users or 87 for users
having at least one photo. “Pro” accounts have naturally much
more photos: they own 59.5% of photos while they represent
3.7% of the users.

3
Since our data extraction, these rules have changed and pro
accounts don t have upload limit any more.
4
http://www.flickr.com/services/api/
5
Let us just mention for the anecdote the first public photo,
numbered 74, http://www.flickr.com/photos/bees/74/,
uploaded on December 15, 2003 and named big_test.
6
In an interview given in April 2005, Stewart Butterfield even
gave an 82% for public photos (see [12]). Note that our 67%
is rather constant in time (actually it goes between 65% and
70%), which does not contradict the 82% since we don t have
a way to know the amount of uploaded-then-deleted photos.