Hit, Link, Like and Share. Organizing the social and the fabric of the web in a Like economy.
Carolin Gerlitz (Goldsmiths, University of London) and Anne Helmond (University of Amsterdam) Paper presented at the DM I mini-conference, 24-25 January 2011 at the University of Am sterdam .
In tro d u ctio n
Dierent types of social buttons1 h a v e d iu se d a c ro ss b lo g s, n e w s w e b site s, so c ia l m e d ia p la tfo rm s a n d o th e r typ es o f w eb sites. ese buttons allow users to share, bookm ark or recom m end the w ebpage or blogpost across dierent platform s such as Facebook, Tw itter, Digg, Reddit, Delicious, Stum bleupon, etc. e buttons often sh o w a co u n ter o f h o w m an y tim es th e p age/p o st h as b een sh ared o r reco m m en d ed : x lik es, x sh ares, x tw eets. ese likes, shares and tweets may be approached from a new media studies perspective as new types of hyperlinks and from an econom ic sociology perspective open up questions about the increasing interrelation between the social, technicity and value online. W ithin new media studies the hyperlink has previously been stu d ied as a fo rm o f cu rren cy o f th e w eb estab lish in g an eco n o m y o f lin k s (W alk er 2 0 0 2 & Jarvis 2 0 0 9 ) an d as an indicator of a discursive relationship (R ogers 2002).
e economy of links describes the link as a currency of the informational web in which search engines use hyperlinks to look at the relations betw een w ebsites in order to establish a ranking. e term in fo rm a tio n a l w e b is o fte n u se d to d e sc rib e th e w o rld w id e w e b a s a p u b lic a tio n m e d iu m fo r p u b lish in g content (R oss 2009) and is characterized by the linking of inform ation (W esh 2007).2 In th is w e b se a rc h engines act as m ain actors to be able to navigate through all the inform ation b y re c o m m e n d in g p a g e s b a se d o n authority m easures.
According to social networking site Facebook “the informational Web is being eclipsed by the social Web” (Claburn 2009). In contrast to the informational web where search engines focus on links between websites, the social web “is a set of relationships that link together people over the Web” and “the applications and innovations that can be built on top of these relationships” (H alpin & Tueld 2010) and is characterized
1 e term social buttons is used here to include: social bookmarking buttons, voting buttons from social news sites/content aggregation sites, sharing buttons and like buttons. is denition based on social activities on platform s excludes sharing through e-m ail. 2 e name informational web is often used as a synonym for Web 1.0.
1 by the linking of people (W esh 2007).3 W ithin the social w eb search engines and social m edia platform s look at the connections betw een people and their relations to other w eb users or w eb objects. Facebook popularized the term Social G raph “to describe how Facebook m aps out people's connections” (Z uckerberg 2009). A s Facebook considers its services inherently social and its plugins and buttons are called 'Social plugins' we su m m arize th e activities th ey gen erate as so -called “so cial activities.”
Where Google can be seen as the main agent of the informational web and the regulator of the link econom y, Facebook is currently seen as the em erging agent of the social w eb. Especially the com pany’s recent eorts to m ake the entire w eb experience m ore social m ark th e ad ven t o f a d ieren t typ e o f eco n o m y w h ich is based on social indexing of the web: the Like econom y. Key elem ents of this econom y are the social buttons, th e activities th ey g en erate a n d th e w a y th e y c o n n e c t F a c e b o o k w ith th e e n tire w e b .
According to Facebook, liking and sharing are valuable for users and the company because they enable to experience the w eb m ore socially. A sim ilar connection betw een the social and econom ic value has been developed by Adam Arvidsson (2009) with his idea of an ethical econom y in which value creation is based on collective negotiation and in which econom ic value creation is related to the quality of social bonds th at are g en erated . W ith in th is p ap er w e w an t to q u estio n th e cen trality o f so cial d yn am ics an d so cial relatio n s as key driver for platform engagem ent and the Like econom y. rough m erging a new m edia w ith an econom ic sociology perspective, w e w ill shift attention aw ay from the users and the social to the im pact of issu e s o n so c ia l a c tiv itie s, a s w e ll a s th e ir in te rre la tio n w ith te c h n ic ity a n d th e fa b ric o f th e w e b . B a se d o n a n extensive em pirical study of button presence and engagem ent w ithin a sam ple of 592 U R L s, w e ask how issues, technicity and the social create a productive assem blage of value creation in an em erging L ike econom y.
In w h at fo llow s, th is p ap er aim s to ad d ress th ese q u estio n s b y rst lo o k in g at th e h isto ry o f d ieren t typ es o f w eb eco n o m ies o ver tim e. H o w d o th ese ‘n ew ’ so cial activities cen tral w ith in th e so cial w eb relate to the hit and link econom y of the inform ational w eb? W hat creates engagem ent and how does this engagem ent organize the fabric of the web and sociality? And nally, what are the perspectives of a Like econom y?
3 Hence, the social web is a dierent way to address Web 2.0.
2 1. History, presence & use of social buttons
is section will look into the history of social buttons and their associated counters as a metric of social activities and as indicators of a particular w eb econom y. e buttons foster social activities w hich are then quantied in the button counter a n d c a n b e u se d a s w e b a n a ly tic s m e tric s. W e b a n a ly tic s is d e n e d a s “th e measurement, collection, analysis and reporting of Internet data for the purposes of understanding and optim izing W eb usage” (W eb Analytics Association). Distinct m etrics in web analytics can be seen as belonging to particular web periods and econom ies. is section contextualizes the em ergence of social buttons by addressing the shift from the inform ational web, characterized by the hit and link econom y, to the so cial w eb w ith its em ergin g L ik e eco n o m y. T ak in g a gen ealo gical ap p ro ach th ese w eb p erio d s an d th eir metrics are not mutually exclusive, but rather overlap, built upon, enrich and complicate each other.
e hit economy (web counters)
e history of the social buttons may be traced back to the mid 1990s when web counters were a common sigh t. ese w eb co u n ters d isp layed a n u m b er o f ‘h its’ rep resen tin g “a co m p u terized req u est fo r in fo rm atio n from th e site” or “a specic request from the user of a W eb brow ser to view the contents of the selected docum ent” (D 'Alessio 1997:498). e hit was used as an early engagem ent measure and becam e th e sta n d a rd for m easuring w ebsite trac in th e m id 1 9 9 0 s (id em ). e h it served as a m etric for w eb advertisin g in th e hit econ om y w h e re w e b site s w o u ld b u y th e ir w a y in to th e to p o f se a rc h e n g in e s in o rd e r to re c e iv e m o re h its:
“Preferred placem ent is a term em ployed by search engine com panies for boosting sites in query retu rn s. O rgan isatio n s p ay en gin e co m p an ies to h ave th eir sites p laced h igh er in search en gin e retu rn s, in o rd er to receive m o re h its. W h en th ey ad d u p , h its co u n t. In th e h it eco n o m y, organisations hope to gain banner advertising revenue and dem onstrable net presence. Hit counts sh o w p resen ce. ey in d icate m easu res o f site p o p u larity an d reliab ility” (R o gers 2 0 0 2 : 1 9 7 ). e web counter is a sign of the hit economy in displaying the number of hits as a very rudimentary4 in d ic a tio n o f e n g a g e m e n t w ith a w e b site .
e link economy (PageRank counters)
In th e late 1 9 9 0 s a n ew typ e o f search en gin e, G o o gle, sh ifted valu e d eterm in atio n o f w eb sites fro m ‘h its’ to links. Inspired by the academ ic citation index search engine G oogle introduced the link as a relevance and
4 e hit is a very rudimentary measure because “hits do not correspond in any direct fashion to the number of pages viewed or num ber of visitors to a site. For exam ple, if a viewer dow nloads a W eb page with three graphics, the W eb log le will show four hits: one for the W eb page and one for each of the three graphics.” (Ferrini & M ohr 2009:124)
3 authority m easure. G oogle founders Sergey Brin and Larry Page created a hyperlink analysis algorithm nam ed PageRank that calculates the ranking of a web page by looking at the links it receives. It established that not all links have equal value, as links from authoritative sources and links from sources receiving m any inlinks have a higher value (Gibson et al 1998). A high PageRank becam e a quality indicator of a website and many websites have displayed their PageRank on their website with a PageRank button. A few years after the introduction of G oogle’s PageR ank algorithm , W alker critically exam ines how the algorithm caused a great sh ift in th e w ay search en gin es ran k co n ten t an d m ak e it accessib le b y “u sin g lin k s a s th e p rim a ry m e th o d o f determ ining the value and thereby the deserved visibility of a website” (p. 72).
Shifting attention away from merely hitting to linking is a rst step to include social validation and relatio n al valu e to search en gin e algo rith m s. Y et, th e so cial valid atio n rem ain s an exp ert system , as th e valu e o f an inlink is determ ined by the degree of the inlinker's authority. e PageR ank algorithm established an econom y governed by search engines w ho regulate the value of each link (W alker 2 0 0 2 ). S u b se q u e n tly it le d to th e co m m o d icatio n o f lin k s as w eb o b jects th at can b e trad ed , so ld o r b o u g h t w ith in th e lin k eco n o m y. Eleven years after the introduction of the PageRank algorithm Je Jarvis describes how this link econom y is well established on the web with Google as the main economic agent at its center (2009: 28).
It w as th e b lo go sp h ere th at in tro d u ced a m etric w h ich started to in vo lve u ser en gagem en t rath er th an expert validation. Blogs created a new type of m etric that show s the num ber of blog subscribers as a m easure of engagem ent. e counters display how m any people are subscribed to receiving update notications through e-m ail or (R SS) feeds. e am ount of subscribers serves as a quality or engagem ent m easure of blogs. It feed s b ack in to th e lin k eco n o m y as an ad d itio n al, u ser gen erated facto r th at co n trib u tes to th e ran k in g o f a website or blog (Bihun et al).
e Like economy (social counters)
e social web further developed the user-focused metric and introduced it to the entire web. Within the so cial w eb w e can d istin gu ish an o th er typ e o f co u n ter: th e so cial b u tto n s w h ich d isp lay th e in teractivity w ith th e o b ject5. e b u tto n s a llo w fo r a n u m b e r o f p re -d e n e d u se r a c tiv itie s (e g . lik in g , sh a rin g , tw e e tin g ) w ith th e o b ject in relatio n to so cial m ed ia p latfo rm s. e rst social counters w ere found on social bookm arking sites lik e D elicio u s fo r sto rin g lin k s an d o n co n ten t aggregatio n w eb sites lik e D igg an d R ed d it w h ere u sers can vote stories up or dow n. Initially the ranking and displayed ranking count were internal to these platform s.
5 Any web object that has a URL ( Uniform Resource Locator) or more specically a URI (Uniform Resource Identier). An object can be a video, photo, w ebsite, w ebpage, blog, blogpost, etc.
4 You could only “D igg” a story and see the num ber of Diggs on the Digg website itself. e introduction of a button that could be placed on any w ebsite externalized the process of D igging and its count. Publishers placed these buttons on their websites to syndicate their stories across dierent platform s. Content aggregation sites lik e D igg m ark ed a sh ift in co n ten t reco m m en d atio n o n th e w eb . In stead o f w eb m asters lin k in g to interesting and relevant stories, regular w eb users w ere now linking and recom m ending stories through the act of sharing. In 2006 Facebook jum ped on the share-bandw agon because “Ever since this whole Internet thing got started, people have been sharing stu left and right.” In their rst im plem entation of share an item could be shared on Facebook by pasting a link into a eld on the M y Shares page (H ughes 2006a). Sharing could initially only be done from inside the platform . O nly a few days later they externalized sharing w ith the creation of a sim ple link w ith a Facebook icon that could be placed on any w ebsite to enable direct sharing (H ughes 2006b). Sharing could now be done from outside the platform and no longer required the m anual copy-pasting of a link into the platform . T hree years later Facebook introduced an ocial button with a counter to “enrich” the experience of sharing, to track the popularity of an item on the w eb and to invoke other social activities on the Facebook platform (Kinsey 2009):
Start conversations with your friends in just a few clicks whenever you see a Facebook Share button, and see their reactions through com m ents in your N ew s Feed. e Share button enables you to take content from across the W eb and share it w ith your friends on Facebook, w here it can be re-shared over and over so the best and m ost interesting item s get noticed by the people you care about (idem ). e share button evokes further social activities inside the platform such as commenting and liking. erefore, th e sh are co u n ter w as set u p as a co n tain er m etric to cap tu re th e n u m b er o f sh ares an d all fu rth er activities th ey in itiated su ch as th e co m m en ts o r in tern al lik es: “e box_count and button_count options displays a count of the total num ber of tim es the page w as shared on Facebook, how m any com m ents w ere added to the story shared on Facebook, and how m any tim es friends L iked the shared story” (Facebook D evelopers: Share).
Liking was introduced internally on Facebook as a quick and easy way to show your friends that you like the content they share. It w as put forw ard as a social activity that can be perform ed on a shared object within Facebook to replace short aective positive comments like “Awesome”6 a n d “ C o n g ra ts!” : “ e aggregation of the sentim ent "I like this" m akes room in the com m ents section for longer accolades. [...] W e th in k o f th e n ew "L ik e" featu re to b e th e stars, an d th e co m m en ts to b e th e review ” (P earlm an 2 0 0 9 ). e Like button was initially only available within Facebook and it allowed users to like almost all storied on their network's news feeds. It cam e with a counter that showed the total num ber of likes as well as the nam es of frien d s w h o click ed it. In A p ril 2 0 1 0 F a c e b o o k e x te rn a lize d th e a c tiv ity o f lik in g b y la u n c h in g a L ik e b u tto n
6 In a d etailed h isto ry o f th e L ik e b u tto n it is d escrib ed as a w ay o f ran k in g an d as a w ay to d isp lay an aective, p o sitive em o tio n , rst as a project codenam ed "Props" and later as the Aw esom e button with would becom e the Like button.(Bosworth 2010).
5 for the w hole internet at their F 8 developers conference. ro u g h th e L ik e b u tto n p lu g in w e b m a ste rs c o u ld add the Like button to their w ebsites and enabled the liking of any object anyw here on the w eb. e Like would appear on the user's newsfeed and the counter would be incremented. e counter shows the number of likes, com m ents and shares as the Like is a container m etric.
e buttons were introduced to enable sharing directly from within the content website, removing th e step s o f co p yin g th e U R L , o p en in g th e p latfo rm w eb site an d p astin g th e U R L th ere. W h at d ieren tiates th ese so cial b u tto n s fro m th e p revio u sly d escrib ed co u n ters is th at th ey are lin k ed w ith extern al p latfo rm s where the content is shared. ey allow for the cross-syndication and sharing of content across social media. Every platform has their own buttons, created by either the platform itself or by third-party services, which can lead to a (cluttered) array of buttons on w ebpages (see illustration 1)7.
Illu stration 1: S o cia l b u tto n s a cro ss v a rio u s w eb sites
7 In 2 0 0 6 th e rst “all-in -o n e” sh arin g service lau n ch ed , A d d is, w h ich d escrib es itself as “th e rst service to p ro vid e a gen eric gatew ay for collecting and distributing m any dierent types of content. Addis acts as a bridge betw een the web publisher, the web user, and the social media services” (Banks Valentine 2006). It soon became the #1 sharing service because it integrated all major platforms in one single button which removed the need for a cluttered array of buttons. By acting as an intermediator Addis is able to gather statistics on what is being shared, how many times and where. By implementing a Shareis button on th eir w eb site w eb m asters can access th ese statistics fo r th eir o w n w eb site to see w h ich item s are sh ared o ften in o rd er to o p tim ize th eir co n ten t fo r th eir visito rs.
6 e social counters displays the total number of people who have shared the page or post on the associated platform . As sharing can be done either from inside the platform or outside the platform on a website with a button, the counter show s the total of the internally and externally perform ed activity. e Facebook Share and Like buttons pose an exception as they are container m etrics and incorporate a w ider range of activities which shall be addressed next.
Use of social buttons: sharing and liking
is paper specically focuses on the social activities of the platforms Twitter and Facebook as they account for the m ajor part of sharin g trac (A ddis 2010). W hen looking at their technical conguration w e can dierentiate two types of social buttons: the share button and the Like button. W hen a user clicks on a share button (in this case a Facebook share or tweet button), they are usually confronted with a pop-up window th at d isp lays a d escrip tio n o f th e p o st an d a lin k to th e p o st. D ep en d in g o n th e p latfo rm u sers can ad d com m ents before sharing the post. A fter clicking share/post, a description, optional com m ents and link to the post are posted to the platform . e visibility of links shared on Tw itter and Facebook is dierent. On Tw itter, th e tw eet co n tain in g th e (sh o rten ed ) sh ared U R L is p o sted in th e u ser's tim elin e. L in k s sh ared o n T w itter are openly accessible and are visible without being logged into Tw itter.8 O n F a c e b o o k , th e sh a re d lin k is p o ste d o n th e u ser's W all w h ich is o n ly visib le w ith in F aceb o o k . O n e h as to b e lo g g ed in to F aceb o o k to see sh ared lin k s and their visibility m ay be further restricted to friends only or friends of friends, depending on the user's privacy settings. Sharing is enabled on the website itself through the overlay of a platform related pop-up,9 th e 'Share B ox,' so that the user never has to leave the w ebsite. If the share button contains a counter the num ber will be increased after sharing the post.Whereas sharing happens via pop-ups, liking is done on the page itself through a single click on the L ike button. O nce the button is clicked, the user receives feedback from the button10 w h ile th e lin k o f th e lik e d p a g e o r o b je c t is b e in g se n t to th e u se r's n e w s fe e d in th e b a c k g ro u n d . Liking can be considered a further enrichm ent of the previous sharing feature as it creates qualitative “I like th is” lin k s b etw een th e p ag es an d users and captures the users' aective reaction to a page.
8 Twitter oers the possibility to create a prole private, but relatively few users make use of this feature. 9 Som etim es sharing is not done in a pop-up but on a next page, after which the user will be brought back to the content page. 10 It eith er sh o w s “Y o u lik ed /reco m m en d ed x” an d /o r it sh o w s h o w m an y p eo p le h ave lik ed th e item . O n so m e p ag es it is also possible to add com m ents to the liked object in the prom pted “W hat's on your mind?” box.
7 With the introduction of a universal Like button Facebook started to deprecate its share button11 a n d collapsed the share and like counts “so th at th e co u n t rep resen ts to tal in teractio n s w ith th e U R L ” (Z ee 2 0 1 0 ). e like count also became a container metric for all Facebook activities displaying likes (including likes from outside the platform ) shares and com m ents. e like counter shows the total Facebook activities with the URL as a measure of engagement. e button is not only enabling a user-generated value (m ade visible in the counter) but also a platform linking m echanism . e buttons provide the glue betw een w ebsite and the social media platforms as will be discussed in the section on the fabric of the w eb. It can b e argu ed th at lik in g produces a particular form of linking which dier signicantly from traditional linking practices of webmasters and shall be discussed in the next section. rough the act of liking Facebook users are validating and linking content on the w eb, an act previously exclusive to w ebm asters and establishing w hat m ay be considered an em erging Like econom y.
Button presence
e presence of social buttons might be considered as an indicator of the importance webmasters assign to the so cial activities o f sh arin g an d lik in g. erefo re w e are in terested in th e p en etratio n o f sh are, L ik e an d tw eet buttons on the web. W ebservice BuiltW ith tracks technology usage on websites, including third party widgets su ch as L ik e, sh are an d tw eet b u tto n s.12 W hen exploring the presence of L ike and share buttons, the generic sharing button provided by A ddis h as th e h igh est b u tto n p resen ce, p resen t in 5 ,7 2 % o f th e to p 1 0 ,0 0 0 websites as of 12 January 201113. It is fo llo w e d b y F a c e b o o k L ik e w ith 4 ,9 2 % , F a c e b o o k S h a re r 1 ,9 % , T w itte r widget 1,76%, Shareis 1,47% and Twitter button 1,38%. In what follows we want to discuss these gures in relation to the results of our em pirical study. For that purpose w e w ill rst provide an overview of the scope and m ethodology of our em pirical research.
is paper is based on an empirical study that explores th e d istrib u tio n o f so cial b u tto n s a n d th e activities they generate in relation to social issues1415. e stu d y a sk s h o w so c ia l a c tiv itie s a re re la te d to th e h it
11 “W e don't recom m end the Share button for new developers. If you aren't already using the Share button, w e recom m end you migrate to the Like button and Open Graph protocol instead of Share for sharing pages from your website. e Like button is sim pler to user and is the recom m ended solution m oving forw ard.” (Facebook D evelopers) 12 http://trends.builtwith.com /widgets 13 based on Quantcast's top million ranking websites 14 e p ro je c t b u ild s o n to p tw o p ilo t stu d ie s w h ic h w e re c o n d u c te d d u rin g a n d sh o rtly a fte r th e D ig ita l M e th o d s S u m m e r S c h o o l 2010 which can be found here: http://wiki.digitalmethods.net/D m i/W ebC urrencies a n d h e re : http://wiki.digitalmethods.net/D m i/W ebC urrencies2 15 In a rst v e rsio n o f th e stu d y w e re trie v e d d a ta fo r a v a rie ty o f so c ia l m e d ia p la tfo rm s in c lu d in g D e lic io u s, D ig g , R e d d it, Hackernews and Friendfeed via the Backtype Tool. Due to very low number across these metrics we decided to focus on the most prom inent metrics only, those of Facebook and Tw itter and to explore the em erging Like econom y.
8 and link and w hat contributes to h ig h n u m b ers in so cial co u n ters. S p ecial atten tio n sh all b e p aid to th e im pact of social issues and the technicity of the w eb. erefore, w e have studied button presence and engagem ent in relation to six tim ely issues from a variety of categories, w hich generated considerate social media engagement during 201016:
1. "BP Oil Spill" (environm ent)
2. “G round Zero M osque” (politics)
3. "Rally to Restore Sanity” (politics/entertainm ent)
4. “Tea Party” (politics)
5. “C hilean M iners” (disaster)
6. "Lady Gaga" "M eat dress" (entertainm ent)
For each of these issues we retrieved the top 100 Google results by using the Google Scraper17 fro m th e D ig ita l Methods Initiative tools. W e d e c id e d to g e n e ra te o u r sa m p le o f w e b site s v ia th e in fo rm a tio n a l w e b a s th is w e b is m ore issue based than the social w eb, w hich is focused on personal netw orks. For each of these w ebsites the presence of a Like, share18 o r T w e e t b u tto n , a n d w h e th e r o r n o t it in c lu d e d a c o u n te r, w a s c h e c k e d m a n u a lly. We retrieved the number of Facebook Likes, Facebook Shares and Facebook com m ents for each UR L by using th e B ack T yp e S tats to o l19 a n d th e n u m b e r o f tw e e ts u sin g th e D ig ita l M e th o d s R e tw e e t R ip p e r to o l20. For a detailed study, we manually categorized the results per issue in regard to the media featured on the websites and their content. e data was generated between October and December 2010.
In a rst an alytical step w e lo o k ed in to th e o verall p resen ce o n th ese w eb sites as w ill b e d escrib ed n ext (illu stratio n 1 ), as w ell as in b u tto n p resen ce p er issu e an d b u tto n p resen ce w ith in p articu lar categories o f websites (which will be addressed in section four). We calculated the interrelation between button presence and value of activities in order to determ ine the im pact of button presence. Furtherm ore, w e explored w hich issu e s a re p a rtic u la rly so c ia l b y v isu a lizin g th e so c ia l a c tiv itie s p e r issu e (illu stra tio n 3 ). T o a n a ly ze th e se ndings in detail, we dierentiated social activities per issue in relation to media form ats and website content. 16 S e v e ra l o f th e se le c te d issu e s a re fe a tu re d in th e tre n d in g to p ic s o f F a c e b o o k a n d T w itte r fo r 2 0 1 0 : http://blog.facebook.com /blog.php?post=466369142130; http://blog.tw itter.com /2010/12/hindsight2010-top-trends-on- tw itter.h tm l 17 http://tools.issuecrawler.net/beta/scrapeG oogle/ (b a se d o n G o o g le .c o m ) 18 Facebook Share or generic share button that allow s sharing to Facebook 19 http://tools.digitalmethods.net/beta/backtype/ (based on backtype.com : “Enter a UR L to see its social impact”) 20 http://tools.issuecrawler.net/beta/twitter/nrRetweets.php (based on Topsy.com - real-time search & Tw itter Trackbacks)
9 Button presence
From our total UR L sam ple we rem oved UR Ls from sam e source in the sam e page conguration, as they will sh o w th e sam e b u tto n s, w h ich left u s w ith 4 2 0 U R L s. In th is sam p le th e p en etratio n o f so cial b u tto n s is as fo llo w s:
41% of all webpages have a Like button (of which 95% show a counter)
64% of all webpages have a share button (of which 8% show a counter)
68% of all webpages have a tweet button (of which 47% show a counter).
Illu stration 2: O v era ll b u tto n p resen ce (m ed iu m g rey) a n d co u n ter p resen ce (d a rk g rey) a cro ss a ll issu es
In o u r sam p le th e tw eet b u tto n is th e m o st p resen t, fo llow ed b y th e sh are an d L ik e b u tto n . A lm o st all L ik e buttons show a counter due to the button's default settings21. A b o u t h a lf o f th e tw e e t b u tto n s sh o w a counter22 w h ile o n ly a v e ry lo w a m o u n t o f sh a re b u tto n s sh o w a c o u n t b e c a u se o f th e g e n e ric sh a rin g b u tto n s lik e A d d is and Shareis that do not show a count. O nly the Facebook Share can display a count w hich is either the 'actual' share num ber or the totality of Facebook activities w hich m akes it a m essy counter. e num ber of shares and likes in the counters are often the sam e due to Facebook's eort to merge both counters. Despite Facebook's attempt to deprecate the share and merge it with the Like button, we found that the share is still m o re d o m in a n t th a n th e L ik e . is is caused by the popularity of the generic sharing service A ddis which uses the traditional Facebook sharing mechanism over the like by default.
When looking at the button presence within each issue (illustration 2), the majority of the issues sh o w a sim ilar d istrib u tio n o f L ik e, sh are an d tw eet b u tto n s as p resen ted in illu stratio n 1 . e o n ly excep tio n is th e T e a P a rty, w h ic h h a s sig n ican tly less button s an d even less coun ters. e q u e stio n a rise s if th e lo w
21 http://developers.facebook.com /docs/reference/plugins/like 22 e two most widely used Twitter buttons are the Tweetmeme button, which always shows a counter, and the ocial Twitter button which may be congured to display no counter.
10 button presence in this case functions as an indicator of a less social or less engaging space and shall be follow ed throughout the paper.
The social buttons distinguishes them selves from previous web buttons and counters due to their sp ecic congu ratio n . In th is n ew typ e o f co n guration the button serves as both a user-generated value and as a platform linking m echanism . e button is linked w ith an external (social m edia) platform w here a link to th e w eb site is p u t w h en th e b u tto n is click ed . e b u tto n serves as a typ e o f w eb g lu e b etw een th e p ag e th e button is located on and a social media platform . W e will look into how social activities relate to traditional linking practices, w hat type of link is being created through these buttons, and how the social activities of sh arin g an d lik in g relate to th e h it an d lin k eco n o m y.
Illu stration 3: B u tto n a n d co u n ter p resen ce p er issu e
11 2. Relation to link
e process of linking through the acts of sharing and liking is very dierent from the traditional linking practices of webm asters and bloggers. In the traditional model the webm aster of website A links to website B where the link is made visible on website A. In the act of sharing the link is not being generated by the webmaster of website A but by the visitor/user from website A. It is a user-generated link e n a b le d b y b u tto n s on website A. e second im portant distinction is that website A is not linking to B but that website A is being linked to on platform X . L in k s a re b e in g c h a n n e le d th ro u g h a n d in c o rp o ra te d o n e x te rn a l p la tfo rm s where they can be quantied (how many people share/like this) and qualied (who shares/likes what, where and w hen). is new w ay of linking is a form of light-linking, as it does not require the m anual labor of creating a link. O n top of that this link is initially invisible because it is already em bedded in the button.
Precongured links
e social buttons were introduced to make it easier to share content across the web without having to copy and paste a U R L and m ove betw een content and sharing platform . e link is em bedded in the code of the button23 w h ic h a llo w s fo r d ire c t lin k in g w ith o u t h a v in g to c o p y th e U R L . S o c ia l a c tiv itie s m a k e u se o f a lin k th at h as alread y b een m ad e b y th e service p ro vid in g th e b u tto n , o ften th e extern ally asso ciated p latfo rm itself. e link in the social button can be understood as a distinct type of hyperlink: a pre-congured link - or as su ggested b efo re, as lin k -ligh t. A s th e sh are an d tw eet req u ire m o re eo rt th an th e L ik e th ey create d ieren t associations and levels of engagem ent. e Like is creating a link in the background and m ay be seen a link th at in d icates an aective response and not so m uch an intentional relationship.
23 Examples from pre-congured links in the social buttons on the Hungton Post: FB Like
12 Revisiting the hit and link economy
e social buttons relate to both the hit and link economy through their user-generated values and linking mechanisms. ey build on the previous web economies, yet at the same time add a dierentiation. e Like relates to th e h it as th e b u tto n registers a 'h it' in th e fo rm o f a click after w h ich th e n u m b er in th e co u n ter is increm ented. H ow ever, not every single 'hit' is being counted, instead only intentional ae c tiv e re a c tio n s a re counted. e Like also relates to the link as liking a w ebsite autom atically creates a link betw een the user and th e site o n F aceb o o k w h ich is fed in to th e u ser's N ew s F eed . B u t th e L ik e also in tro d u ces a sig n ican t dierence to the link as it adds a social value to it. Facebook sees the Like as a very specic type of link and at their F 8 developers conference “Facebook announced L ikes as a form of "social links" -- better than a link because it's related to a specic user” (Cashm ore 2010). Liking can therefore be understood as a qualication of the lin k by addin g it to the user's prole, m aking it m ore personal and social (by fostering m ore Facebook activities). R eturning to the Like as a container m etric w hich also includes shares and com m ents, the Like is both more than a hit and more than a link.
As discussed in section one, the Like economy further changes whose links do matter. Whereas in the inform ational w eb links w ere created by w ebm asters and sorted according to G oogle’s PageR ank, in the social web links are created by users who at the same time add value to them through liking and comm enting. In the link econom y G oogle values links by using and expert or 'authoritative' qualication w hile Facebook validates links through the qualication in the social. W hat is at stake here is not only w ho creates links on the w eb but also how and by w hom these links are qualied.
Finally, the Like econom y changes the visibility of links. e link econom y is based on webm asters creating links that are publicly available on the w eb and craw led, indexed and ranked by several econom ic agents like G oogle, Bing and Yahoo! H ow ever, w hat dierentiates the Like econom y from the link econom y is th at th e lin k s created th ro u g h lik in g are n o t o p en ly availab le b u t in stead all ro u ted th ro u g h th e F aceb o o k platform . is means that the main econom ic agent in the Like econom y is Facebook which will be discussed in se c tio n fo u r. In w h a t fo llo w s, w e a im to d isc u ss th e se c h a n g e s in d e ta il b y e n g a g in g w ith o u r e m p iric a l ndings and by critically exam ining the so called social value of the social buttons.
13 3. Engaging issues and media form ats
From the very beginning Facebook has presented social dynam ics as key driver of activities su ch as lik in g , sh arin g an d co m m en tin g. It h as in tro d u ced th e so cial p lu gin s as a p o ssib ility to m ak e th e w eb exp erien ce in general m ore social as content can easily be shared with one’s netw ork (H a u g e n 2 0 1 0 ). e c o m p a n y stre sse s how the plugins enable users to create connections to pages they appreciate and to lter their web experience th rough th eir frien ds' preferen ces. e value of L ike buttons and the users w ho engage w ith them has been advertised as m aking use of existing social dynam ics, as the so called “Likers” are connected to 2.4 more people than the average Facebook user and click on 5.3 more external websites (F a c e b o o k a n d M e d ia 2 0 1 0 ). A sim ilar p ersp ective h as b een k ey to so cio lo gical stu d ies o f so cial m ed ia activities as in th e w o rk o f A d am Arvidsson (2009). Following a digital methods approach (Rogers 2009) this paper shifts attention away from stu d yin g th e lik ers, sh arers an d tw eeters an d th eir so cial d yn am ics, b u t p o ses q u estio n s ab o u t th e relatio n between the issue and social activities as well as their technicity. In what follow s we will explore which issue generates what kind of social activities and investigate, whether particular issues, web content or m edia are especially likeable or tw eetable.
As introduced in section one, each issue is characterized by a dierent distribution of social activities (illu stratio n 4 a n d 5 ). D esp ite d ivergin g resu lts, all issu es gen erate b etw een 6 to 3 0 tim es m o re F aceb o o k activities than tw eets. e signicant dierence betw een Facebook and tw eeting activities suggests Facebook's predom inance in producing social media engagem ent, a trend that resem bles Facebook’s lead in overall sharing activities on the w eb (A d d is 2 0 1 0 ). e h igh resu lts in F aceb o o k activities m igh t b e sim ilar acro ss issu es, but their internal com position of likes, shares and com m ents is not, as illustration 5 indicates. M ost sign ican tly, co m m en ts d o m in ate th e co m p o sitio n o f th e L ik e, a n d in g th at m igh t co m e as a su rp rise as most websites oer their own comment spaces and comments require both more eort and involvement. e most comment-intensive issue is Lady Gaga Meat Dress, followed by BP Oil Spill, Ground Zero Mosque and Tea Party.
14 Illu stration 4: A v era g e so cia l a ctiv ities p er issu e Illu stration 5: L ik e d istrib u tio n p er issu e
Media engagement
Within many issues, pages that feature audio-visual media content generate the highest number of social activities. 24 E sp e c ia lly p a g e s w ith liv e stre a m s, p h o to -e d ito ria ls25 o r v id e o s c re a te m o re a c tiv itie s c o m p a re d to in fo rm a tio n a l a rtic le s w ith o u t m e d ia c o n te n t. e issue G round Z ero M osque poses the only exception, as only articles with videos receive m ore likes and slightly m ore tweets com pared to regular articles, whereas articles w ith pictures generate less activities.26 e se d iv e rg in g re su lts m ig h t b e lin k e d to th e fa c t th a t m o st issues are actual even ts that ben et from visual docum entation, except of G round Z ero M osque w hich is a political and cultural controversy and therewith less dependent on visual docum entation than discussion, negotiation and the presentation of dierent argum ents.
Content engagement
e number of activities in relation to the content further suggests that controversial debates function as facilitators of social activities. In regard to L ady G aga M eat D ress, th e m ajority of activities occur on w ebsites which address or open up a controversial discussion of her styling choice, followed by websites featuring Gaga’s explanation of why she decided to wear a dress made of meat or feature an analysis of its potential
24 In th e sp a c e o f th e B P O ilsp ill, w e b site s fe a tu rin g p h o to s (1 9 5 5 1 L ik e s in a v e ra g e , 3 2 4 8 tw e e ts) o r v id e o liv e stre a m s (1 1 3 6 0 L ik e s, 1061 tweets) outnum ber general articles (1148 Likes, 194 tweets) or issue overviews (441 Likes, 103 tweets). e sam e applies to th e R ally to R esto re S an ity, w h ere w eb sites featu rin g p ictu res o r d o cu m en tin g th e visu al aesth etics o f th e even t ach ieve th e h ig h est so cial activities (2 2 4 5 6 L ik es, 1 0 8 2 tw eets), fo llo w ed b y life vid eo stream s (1 3 4 8 7 L ik es, 1 8 0 3 tw eets), even t p ages an d gen eral articles (1164 Likes, 67 tw eets). A lso in the case of the C hilean m iners, the category photo is m ost engaging (5571 Likes, 1319 tw eets), b u t h ere th e articles (1 5 5 7 L ik es, 1 0 8 tw eets) o u tn u m b er th e w eb sites w ith live vid eo stream s (6 0 3 lik es, 2 7 tw eets). In th e case of Lady G aga M eat D ress, the m ost activities have been generated by articles featuring several pictures (234 overall Likes and 184 tweets), as com pared to articles with videos (99 Likes and 62 tweets) or articles without media (37 Likes and 24 tweets). 25 For exam ple: e Big Picture - News Stories in Photographs from the Boston Globe. http://ww w.boston.com /bigpicture/ 26 e u n d e rly in g g u re s a re : A rtic le s w ith v id e o s 7 7 5 6 L ik e s, 1 3 8 tw e e ts, in fo rm a tio n a l a rtic le s 2 2 4 0 L ik e s, 9 6 tw e e ts, a rtic le s w ith photos 1881 Likes, 118 tweets, all num bers are averages per page.
15 meaning27. T a k in g th e h ig h n u m b e r o f F a c e b o o k c o m m e n ts in to a c c o u n t, th is su g g e sts th a t th e c o m m e n t sp ace is in d eed a n ego tiatio n sp ace an d th at issu es w h ich evo k e co n tro versial reactio n s su ch as th e M eat D ress, but also the Ground Zero M osque and the Tea Party do not generate many likes (as positive aective respon ses) but rather dieren tiated co m m en ts. is n d in g su p p o rts F aceb o o k’s claim th at th e L ik e is a sh o rtcu t to em p h atic, aective com m ents such as “A w esom e” or “G reat” and therefore leaves the com m ent sp ace fo r d ieren tiated en gagem en t, as ad d ressed in sectio n o n e (P earlm an 2 0 0 9 ). e case o f th e T ea P arty, which has already been discussed as potentially less social space due to its low distribution of social buttons, sees th e least activities gen erated o n w eb sites o f th e T ea P arty m em b er o rgan izatio n s28. N e w s a n d m e d ia contributions to the issue generate 10 tim es m ore Facebook activities than the m em ber organizations and critical contributions even 30 tim es m ore, w hich suggests that the low sociability of the issue is generated by the un-engaging space of Tea Party m em bers.
Issues, but also m edia form ats an d perspectives oered on the issues, can thus be understood as productive entities in creating social activities. Yet, sharing, liking and tweeting also contribute to the form ation and production of the issues them selves. W ebsites, social buttons, social m edia platform s as w ell as issues and social activities therefore form a productive assem blage in the sense of D eL anda (2002, 2006a) and Deleuze and Guattari (2004) in which each entity has an impact on each other. In the next section, we will discuss the role and organization of technicity and the fabric of the web within this assem blage.
27 W ebsites featuring controversial discussion: 3337 average L ikes, 169 tw eets. W ebsites featuring G aga’s explanation: 1180 L ikes, 69 tw eets. W ebsites featuring analysis and discussions of the potential m eanings: 1047 L ikes, 104 tw eets. e least en g ag in g w eb sites are general articles inform ing about the m eat dress incident as w ell as articles discussing style or food concerned issues 28 Tea party m em ber organizations: 402 Likes, 129 tweets. News and media articles: 4774 Likes, 241 tweets. Critical contributions: 12531 Likes, 743 tweets.
16 4. Technicity of sociality and the fabric of the web
As introduced above, social activities are predominantly fostered by buttons and counters, but the like, share and tw eet can also be generated independent of the button by just posting links or liking on the platform s directly. e follow ing section will explore social activities from a medium specic perspective focusing on th eir tech n icity an d th e o rg an izatio n o f th e fab ric o f th e w eb . P revio u s resu lts h ave sh o w n th at b u tto n presence within the analyzed issues is far higher than general trends on the web29. e re w ith th e q u e stio n em erges, to w hich extent the presence of a button contributes to a higher num ber of social activities – or not. ere is a general tendency that pages with buttons produce 100 % more activities than the ones without buttons. e only exception is the BP Oil Spill issue as it generates almost 90% more likes, shares and tweets on pages without buttons.30
As outlined in section one, the Tea Party space has signicantly less buttons than the other issues. When moving from overall button presence to button presence within specic categories of websites, the Tea Party mem ber organization websites stand out with a very low overall button presence of only 2.3% of pages with Like buttons and 7% pages with share and/or tweet buttons31. e n d in g s su g g e st th a t th e m e m b e r sp ace is d esign ed fo r d istrib u tin g in fo rm atio n an d n o t so m u ch n o t fo r sh arin g it acro ss th e w eb an d generating direct feedback, response and interactivity – and therefore generates m uch less activity than the other issues.
Facebook as the fabric of the web
Within the informational web, connections between websites are based on linking practices. Even though Google has deeply impacted linking behavior as its ranking algorithms gave rise to strategic linking practices, the search engine w as not involved in creating the interconnections betw een w ebsites, the so called fabric of the w eb. To explore the question how connections betw een w ebsites are organized in the social w eb, w e w ill discuss the fram ework of the Open Graph as a successor of the Social Graph in which the Like button was
29 On general button presence on the web: http://trends.builtwith.com/widgets/Facebook-Like 30 is general but not clear trend might have several reasons. First of all, especially news websites publish their articles both on their ocial website, but also on their Facebook page, thus m aking it possible to generate social activities on the platform directly on to p o f th e activities en ab led o n th e o cial w ebsite. Secon dly, particularly en gaged users can easily share an d tw eet w ebsites by posting them directly to the platform s, generating further re-tweets or re-shares as well as com m ents. Buttons can thus be considered as im portant, but not as required driver of social activities. 31 N e w s/m e d ia w e b site s h a v e a ra th e r h ig h d istrib u tio n o f b u tto n s (2 8 % h a v e lik e b u tto n s, 2 4 % h a v e sh a re b u tto n s a n d 5 2 % h a v e tw eet b u tto n s).
17 introduced as enabler for a m ore social experience of the w eb. W ith the Social G raph Facebook claim s that the com pany’s m ain assets are the connections betw een users that create social netw orks and ow s of inform ation. But the Social Graph restrained the network and the information ows to the space of Facebook. erefore the com pany decided to increasingly extend the graph to the entire w eb and enable cross syndication of inform ation, w eb experience and netw ork connections. A key step to include all w eb experiences into Facebook and connect them to a user’s social network was the introduction of the Like Button and the Open Graph in April 2010 (Zuckerberg 2010). As introduced in section one, the Open Graph allows for feeding web experience into the Facebook prole, as well as to experience the web ltered through the reco m m en d atio n s an d activities o f th e o w n n etw o rk . C ru cial elem en ts are a n u m b er o f so cial p lu gin s32 w h ic h allow for the cross-syndication of social activities. ese plugins include the Like button, a login button that enables users to connect to other w ebsites w ith their Facebook account, a recom m endation plugin for external websites which shows personalized recomm endations and highlights content that received the most social activities, the Facepile plugin that show s the prole pictures of friends or users that have liked the page or website, and nally the Live Stream for displaying user activities in relation to events or issues in real-time.
According to the Facebook CEO Zuckerberg, the company is “making it so all websites can work together to build a m ore com prehensive m ap of connections and create better, m ore social experiences for everyone” (Z uckerberg 2010). Sociality online is no longer conned to the space of Facebook, but de- centralized by extending Facebook’s key features to the entire w eb. W ith these feedback loops of inform ation, th e so cial b u tto n s n o t o n ly create in teractivity an d so ciality, b u t fu n ctio n as a w eb g lu e th at n o t o n ly organized but actually turns into the fabric of the web. (G e lle s 2 0 1 0 ). W h e re a s F a c e b o o k su g g e sts th a t th e Open Graph is interconnecting and personalizing the web, the argument developed here is that the web is both de- and re-centralized through social activities.
Decentralization
e decentralizing impacts of social buttons are manifold. e increasing integration of social buttons on websites renders the sites both more open and less xed. e buttons enable the distribution of content and com m ents across a w ide range of platform s and w ithin these platform s on m any proles, new s feeds or tim elin es. W ith in th is p ro cess, th e w eb sites are n o d iscrete en tities, b u t fu n ctio n as in itializers o f d iverse activities that happen across diverse (social m edia) spaces. But, at the sam e tim e, w ebsites are shaped by the
32 h ttp ://d e v e lo p e rs.fa c e b o o k .c o m /p lu g in s
18 so cial activities th ey gen erate, as th e so cial en gagem en t is d en in g w h at ap p ears in so cial b u tto n co u n ters, th e reco m m en d atio n p lu gin s, in th e live stream s o r in th e co m m en ts. e m o re so cial p lu gin s a w eb site in tegrates, th e m o re it o p en s u p to b e sh ap ed b y th e activities o f F aceb o o k o r T w itter u sers33. W h e re a s th e se a re ra th e r novel perspectives for the web, they are key characteristics of social media platform s (Boyd 2010), which have no original content and are shaped by the cross-syndication of content, activity and inform ation. In the fram ew ork of the O pen G raph, Facebook and the w eb enter a productive relationship in w hich both have a perform ative impact on each other.
Recentralization
But especially Facebook’s recent eorts to create the Open Graph indicates a simultaneous rewiring and recen tralizatio n o f th e w eb . W h ereas th e in fo rm atio n al w eb w as o rgan ized th ro u gh lin k s b etw een w eb sites, th e so cial w eb is ch aracterized b y lin k s p reco n gu red an d m ed iated b y vario u s p latfo rm s. In sectio n o n e, Facebook has already been discussed as em erging glue of the web. Besides, platform s also collect inform ation about user engagem ent w ith the w eb, especially Facebook is extending its data m ining practices rapidly w ith the O pen G raph. U ser engagem ent w ith Like or Share buttons or w ith links posted on Facebook w alls allow s the com pany to collect data that exceeds the inform ation each user is providing on their prole an d th u s contributes to a re-centralization of the fabric of the w eb and of the ow s of inform ation and aective association.
As a recent paper by Arnold Roosendaal (2010) shows, this process of re-centralisation is not even dependent on using the social buttons, as Facebook plugins and buttons function may as cookies. ey allow to trace brow sing behavior w hen a Facebook user opens a w ebsite that features a L ike button or includes Facebook Connect. Once the cookie is set up, it provides Facebook with every page the user visits until the cookie is deleted. N o m atter if Facebook users decide to use these buttons, their w eb behavior can be traced and connected to their prole34. R o o se n d a l fu rth e r sh o w s th a t th e c o o k ie s c a n a lso tra c e n o n -F a c e b o o k u se rs.
33 An example of this process is the news-blog Hungton Post (http://www.hungtonpost.com/) which has included all social plug- ins provided by Facebook. e blog provides article suggestions based num ber of Facebook L ikes rather than on hits, hence in c lu d e s th e m in to th e ir in te rn a l ra n k in g a lg o rith m s. E a c h a rtic le fe a tu re s a ll so c ia l b u tto n s in c lu d in g c o u n te rs in a h ig h ly prom inent form s including Facebook’s Facepile. Beneath articles, live stream s from Tw itter show the ongoing discussion of issues addressed in the article on the social m edia platform . C om m enting is available via either Facebook or Tw itter proles or individual com m ents can directly be shared on a w ide range of social platform s. Engaging w ith H ungton Post articles through social activities thus not only brings the social platform s, but also the H ungton Post platform into being, has an im pact on the position of articles in recom m endation banners and thus has an prodctive im pact on the page itself. 34 e only way to prevent Facebook from doing so is installing a plugin: e Antisocial plugin which “limits websites from em bedding Facebook content, thereby preventing Facebook from tracking your brow sing habits. It also bans 3rd party Facebook applications outright, thereby reducing the possibility of your inform ation from being leaked.” https://addons.mozilla.org/en- US/refox/addon/162098/ e o n ly e sc a p e fro m th e L ik e b u tto n is a v e ry w e b a n d te c h sa v v y so lu tio n .
19 Even though Facebook cannot connect this data to individual proles and directly use it for personally targ eted ad vertisin g , it en rich es th e d atab ase an d co n trib u tes to p ro cesses o f p attern calcu latio n . erew ith , potentially every web user becom es a Facebook user as their web behavior can now be traced across spaces.
Whereas Facebook suggests that the Open Graph enables personalized connectivity online and illu stra te s th is w ith a at netw ork m odel (Z u c k e rb e rg 2 0 1 0 ), th e c o m p a n y is a d v a n c in g to b e c o m e th e m o st prom inent social activity platform and therewith re-centralising the fabric of the web both spatiality and in term s o f in fo rm atio n co llectio n . e connection betw een m onitoring social activities and brow sing behavior su ggests th at w h at m igh t b e in th e m ak in g is a L ik e E co n o m y rath er th an ju st “b u ild in g th e so cial w eb to g eth er” (Z u ck erb erg 2 0 1 0 ). Facebook uses its Like button to create a fabric that connects the web to the platform , makes every web user a potential Facebook user and engages everyone in the em erging Like econom y.
5. e social
In a recen t F in an cial T im es in terview , F aceb o o k fo u n d er Z u ck erb erg co n sid ers th e so -called so cial as th e m o st prom ising organizing principle of the econom y: “If you look ve years out, every industry is going to be rethought in a social w ay”(G elles 2010). A sim ilar argum ent has been m ade by A dam A rvidsson w ho claim s th at eco n o m ic valu e is in creasin g ly co n n ected to th e q u ality o f so cial co n n ectio n s, th e so called p h ilia, th at com panies m anage to create betw een their consum ers or to their products (2009). In w hat follow s w e w ill critically engage w ith this idea of the social and discuss w hat kind of social liking, sharing and tw eeting create and how it is organized.
e informational social
Our main claim is that the sociality Facebook enables with its Open Graph is not only dening social relatio n sh ip s, b u t is co n cern ed w ith th e valid atio n o f in fo rm atio n th ro u gh p erso n al n etw o rk s. K ey elem en t o f th e O p en G rap h is th e lau n ch o f In stan t P erso n alizatio n 35, a c o lla b o ra tio n w ith se a rc h e n g in e s a n d inform ational sites such as B ing, TripA dvisor, Yelp or Scribd, that have started to include users' Facebook network preferences into their search results. e so-called social experience they prom ise is mainly an inform ational one, allow ing the user to use their friends’ recom m endation as a lter to explore the w eb and
35 h ttp ://w w w .fa c e b o o k .c o m /in sta n tp e rso n a liza tio n /
20 th u s q u alifyin g in fo rm atio n via u sers’ n etw o rk s, as arg u ed in sectio n o n e. e Facepile and recom m endation plugin follow the sam e principle as they are focused on show ing what one’s friends do, like, share and th erew ith tu rn fro m th e w isd o m o f th e cro w d to th e w isd o m o f th e frien d (C lab u rn 2 0 1 0 ). In th is co n text, Facebook is less concerned with enabling social relations but driven by the idea of an inform ational social, as Gelles suggest: “What Zuckerberg is talking about is a new way of organizing and navigating information” (2 0 1 0 ). A sim ilar q u alicatio n is h ap p en in g in relatio n to th e h it via th e L ik e b u tto n . W h ile th e h it w as merely counting the number of visitors without being able to tell anything about the visitor's attitude or aective reaction to a w ebsite, the Like button adds quality to this quantitative m etric w hile at the sam e tim e functioning like a hit counter through em bedded cookies.
Facebook, but also Tw itter and other social media platform s allow for channeling social dynam ics into technicity based and countable activities such as tw eeting, liking, sharing or com m enting. e tech n icity o f platform s and plugins makes it possible to transform intensive aective responses and social dynam ics which are in them selves rather dicult to m easure into button-based activities w hich allow for extensifying them , turning them into countable and com parable values in the sense of D eL anda (2006). W hereas users’ aective responses, their agreem ent, excitem ent or involvem ent have happened unnoticed and unm easured in everyday life before, the L ike button m akes it possible to turn the in ten sive reaction s in to exten sive activities an d in fo rm a tio n . is ubiquitous calculation of qualities has been understood as qualculation by N igel rift, a s “an in creasin g ten d en cy to fram e n u m b er as qu ality, in th e sen se th at calcu lation s are so n u m ero u s an d so pervasive that they show up as forces rather than discrete operations” (rift 2007, 100).
e traceable social
Facebook is unlikely to stop with the current possibilities of extensication of the social, as form er em ployee Matt Cohler explains: “Facebook has always thought that anything that is social in the world should be social online (…). Anything where people ask their friends to help them make decisions – whether it’s food or movies or travel – could be transformed online by social” (Gelles 2010). As a part of this drive to make the entire w eb social, Facebook is not only turning the social into inform ation, but increasingly collapsing the so cial w ith th e traceab le as th e still in ten sive, n o n -m easu rab le, n o n -visib le so cial is o f n o valu e fo r th e com pany. Both dynam ics im ply each other, as to m ake the social inform ational, it has to be possible to trace it, to turn it into com parable m etrics. B ut the social that is em erging here only allow s for particular ae c ts a n d activities – in the case of Facebook, the aects are lim ited to positive ones as Facebook has not signaled any
21 interest to set up anything like the dislike button. If the social becom es so closely connected to the inform ational and the traceable, w hat is considered social dynam ics on the w eb is increasingly den e d b y th e platform s which generate econom ic value through this social.
Personalization
is organization of the social is, as indicated before, closely linked to an increasing personalization of the web. On a broad level, sharing, liking and tweeting websites allows to connect web activity to existing Facebook or Tw itter proles. Via Facebook Connect, but also within Facebook itself, com m enting is not an entirely anonym ous activity, but potentially personal and thus accountable. W ith its eorts to create the O pen Graph, but also by functioning as a container login for multiple websites, Facebook has increasingly turned its proles into web-ID s which allow connecting multiple activities to one prole. W hereas other online ID projects such as OpenID or M icrosoft’s Passport could not achieve user acceptance, Facebook has indirectly tu rn ed in to o n e o f th e w eb ’s m o st cen tral ID s (G elles 2 0 1 0 )36. B u t F a c e b o o k ’s p e rso n a liza tio n is n o t o n ly based on individual user behavior – an approach currently follow ed by Google – it is taking the user’s network in to a c c o u n t.
Anonymity
While network-based personalization of social activities is a key element of the experience, anonymity also plays a crucial role, for instance in relation to button counters. Especially the Like button almost always com es with a counter as shown in section one. e counters produce a sheer, quantied number of general engagem ent, they are considered social, yet they are stripped dow n of the personal37. E v e n th o u g h n o t personalized, the high counter presence indicates that this metric of the mass, the general aective engagem ent, contributes to the so-called social experience of the w eb. H aving zeros in your social button counters suggests that a w ebsite has not engaged or aected its visitors. erew ith, the social buttons create a fabric of both m ore personalized and m ore anonym ized sociality, a m ixture of the recom m endation of the crow d and the recom m endation of the friend. Interestingly, Facebook only focuses on the personalized elem ent of this developm ent, stressing the im pact of personal recom m endation on w eb trac generation:
36 is d e v e lo p m e n t is e v e n fa c ilita te d b y th e in tro d u c tio n o f th e @ fa c e b o o k .c o m e m a il a d d re ss. 37 In th e case o f F aceb o o k , th e in d ivid u al u ser can n o t get access to all p ro les th at click ed th e L ik e b u tto n , in th e case o f T w itter, th is is possible via search tools such as T opsy http://topsy.com /.
22 “Publishers have also told us that people on their sites are m ore engaged and stay longer w hen their real identity and real friends are driving the experience through social plugins” (Z uckerberg 2010).
Scalable social form ations
e interrelation between the personalized and the anonymous suggests that there are dierent social form ations at stake w hen engaging w ith social buttons. ere is the an on ym ous m ass of all likers, sharers an d tw eeters. ere are so m e frien d s’ faces th at m ig h t ap p ear in th e F acep ile n ext to a b u tto n , o r a sh ared lin k in one’s tim eline. A tweet about a webpage that is being retweeted with dierent com m ents, the com m ent on a frien d’s w all w ho is respon din g to a shared lin k or the in form ation that several frien ds have liked a particular website. Social activities do not only create th e so c ia l, th e y c re a te a m u ltip lic ity o f d ie re n t so c ia l fo rm a tio n s.
In th e case o f F aceb o o k , th ese so cial fo rm atio n s are m ain ly d en ed b y th e u sers’ n etw o rk an d th eir p rivacy settin gs. B ein g ab le to see a u ser's so cial activities d ep en d s o n if th eir p rivacy settin gs allo w everyo n e, frien d s of friends, friends only, or selected groups of friends to see their posts or new s feed. W hen com m enting or likin g con ten t shared by frien ds, the visibility of that activity an d thus the social form ation they are exposed to are dependent on the friend’s privacy settings, as discussed in section one. Privacy regulations allow users to scale th e so ciality th ey p ro d u ce in th e sen se o f D eL an d a (2 0 0 6 b ), fro m carefu lly selected fo rm atio n s o f few friends to the entire population of Facebook. Yet, the social form ations not only change in num ber, but also in th eir q u alitative fo rm atio n . S o m e fo rm atio n s m ig h t b e m o re aected b y a u ser’s activities, su ch as th eir clo se friends w hile other form ations are m ore likely to engage w ith buttons as Facebook suggests (Z uckerberg 2010) 38. In o rd e r to b e b o th p e rso n a lize d a n d a n o n y m o u s, m u ltip le so c ia l fo rm a tio n s a re p ro d u c e d th ro u g h so c ia l activities.
38 Likers, so Zuckerberg claims (2010) are multipliers, particular users that have “2.4x the am ount of friends than that of a typical Facebook user”, “click on 5.3x more links to external sites” and are thus characterized as more active and more social. Facebook considers this as valuable, as it suggests external w ebsites that the people w ho w ill engage w ith the social buttons w ill share their so cial activities w ith larger so cial assem b lages an d th u s in crease th e im p act o f th eir so cial activity. H en ce, th e valu e o f a L ik e, a sh are an d a retw eet are co n n ected to th e size an d q u ality o f th e so cial assem b lage th ey en ter, th e b igger an d th e m o re in terested in so cial activities, th e m o re valu e d o es F aceb o o k assign to a L ik e.
23 6. Perspective of a Like econom y
In th e fo llow in g w e w ill d iscu ss w h at fo rm s o f valu e are p ro d u ced in th e em ergin g L ik e eco n o m y. e p ap er has introduced particular social buttons as an eort of platform s – especially of Facebook – to render web experience m ore social and thus to qualify the link and the hit by connecting it to existing proles and their personal networks. In this fram ework, Facebook em erges as a key agent, generating the most social activities and keeping them internal to the platform . rough an em pirical study of social button engagem ent in relatio n to six issu es w e d evelo p ed th e h yp o th esis th at th e so -called so cial activities are n o t o n ly d riven b y so cial d yn am ics, b u t are th e o u tco m e o f a p ro d u ctive assem b lage o f th e issu e, th e m ed ia co n ten t, th e so cial, tech n icity an d th e fab ric o f th e w eb . B ased o n th e cen trality an d im p act o f th e L ik e b u tto n , w e h ave su g g ested that there is an em erging L ike econom y w hich m ight not replace, but denitely recong u re th e h it an d th e lin k e c o n o m y.
Multiple forms of value creation
After exploring how the Like economy is organizing and organized by issues, sociality and through the fabric of the web, the question arises what form s of value are created within this productive assem blage. First of all, th e so cial b u tto n s allo w fo r tran sfo rm in g in ten sive so cial an d aective dynam ics into com parable m etrics and thus add a social an d person al qualicatio n to th e h it eco n o m y. S o cial, as th e activities are b ein g sh ared in so cial n etw o rk s an d p erso n al, as w eb activity is co n n ected to actu al p ro les rather than being anonym ous. According to Facebook, these processes enable the social indexing of the web as opposed to an expert indexing of the link econom y. Yet, these social indices have a lim ited visibility which focus on the personal network, as they appear on friends' w alls, in Facepiles or recom m endation plug-ins, but are not used for ranking algorithm s on social platform s them selves. Even w hen integrated by Instant Personalization partners, only inform ation from a user’s netw ork is taken into account, not the overall social activities generated by all Facebook mem bers. W hereas the inform ational web is taking the total aggregate of indexed sites into account, the social w eb only ranks in relation to a user's netw ork. H ence, the inform ational w eb is characterized by one xed ranking, whereas the social web has multiple, dynam ic rankings which are only visible to an individual user.
Besides generating personalized network value, social activities also contribute to the content validation of websites. Especially if websites include button counters, the display of the num ber of likes, shares or tweets received indicates how engaging and aecting the web content is for web users. e counter renders
24 th e so cial activity in to a cu rren cy o f a h ig h , sh eer n u m b er, n o lo n g er p erso n alized b u t still su g g estin g to b e a so cial en gagem en t m etric.
As it has been shown, the social value created by the buttons is connected to an informational value. e buttons allow for new modes of transactional user data collection, both in regard to the population of Facebook and all web users. rough cookies and button engagem ent, Facebook can extend its user database with web activities and content engagement outside the platform. is Facebook user data is enriched by the anonym ous data of w eb users w ithout a Facebook prole, an addition that enables even better patterning and prediction of interests. Besides Facebook, external social media research com panies also make use of the transactional user data (L ury and M oor 2009). E specially w ithin m arketing contexts, social m edia activities are carefully tracked, m onitored and analyzed, either by algorithm s or by hum an researchers w ith the help of to o ls39. H e n c e , ta k in g th e id e a fu rth e r th a t so c ia l m e d ia a c tiv itie s fu n c tio n a s th e c u rre n c y o f th e L ik e econom y, it is a currency of high num bers w hich is both social and inform ational. M oreover, this currency is also highly ubiquitous and technical, as the O pen Social G raph creates an infrastructure in w hich all w eb behavior contributes to Facebook data mining practices. No matter if a web users decides to engage with Facebook or not, the technicity of the social buttons makes web users participants in the Like econom y, constantly producing valuable user data and contributing to social indexing w ithout even know ing.
Social activities as exchange value
Social m edia activities should not only be considered as currencies in an abstract sense as they are already in use as direct econom ic exchange mechanism s. Especially in the creative industries, web users can buy content like books, m usic and video les through tw eets. e digital agency Innovative u n d e r fo r in sta n c e is se llin g an e-book on digital m arketing for a tw eet ('Pay w ith a tw eet')40 in o rd e r to p ro m o te so c ia l a c tiv itie s a s exchange m echanism s and has “sold” m ore than 150.000 books so far.41
e economic impact and exchange value of social media activities further becomes apparent by com panies’ eorts to increase their social m edia m etrics and online engagem ent through strategic planting of sto ries, ap p licatio n s an d cam p aign s th at are aim ed at d rivin g u p u ser en gagem en t. B u t u ser in vo lvem en t can also be sim ply bought. In the case of Facebook, com panies can buy dierent am ounts of likers via specialized
39 An overview of current marketing oriented social media research activities can be found on: http://m easurem entcam p.wikidot.com /. 40 h ttp ://w w w .o h m y g o d w h a th a p p e n e d .c o m /in d e x .h tm l 41 Sim ilar developm ents can be found on the locative service Foursquare, where frequent visitors of places, so called m ayors are oered discounts or receive free products, such as in the case of Starbucks.
25 services su ch as B u y Real F a c e b o o k F a n s, U so c ia l, F a n p a g e h o o k u p , S o c ia lk ic k o r F a n b u lle t42. e c u rre n t p ric e for 1000 likes starts at $57 43 b u t c a n g o fo r $ 1 1 9 44. C o m p a n ie s c a n c h o o se b e tw e e n ra n d o m o r so -c a lle d targ eted fan s, selected th ro u g h th eir lo catio n , in terests o r ag e. S im ilar d evelo p m en ts can b e m o n ito red o n so cial b o o k m ark in g sites, w h ere so cial m ed ia agen cies p ay p eo p le to p ro m o te issu es an d th u s in crease th e num ber of inlinks and hits to particular websites, hence connecting to the link and hit econom y (M ills 2006).
Orientation towards the future
To conclude we would like to argue that the em erging Like econom y is characterized by a double orientation tow ards th e future. First of all, th e gen erated user data is used for person alized advertisin g an d reco m m en d atio n - b ased o n th e assu m p tio n th at if o n e’s n etw o rk lik es o r sh ares co n ten t, o n eself is m o re lik ely to like as w ell. K now ing a netw ork’s preferences thus enables Facebook to generate patterns and to predict a user’s future interests and activities. Secondly, the idea of the Open Graph is build on the assum ption that the reco m m en d atio n w ith in a p erso n al n etw o rk , th e w isd o m o f th e frien d s, is far stro n ger th an a n o n -p erso n al reco m m en d atio n , th e w isd o m o f th e cro w d an d th at th ere is in h eren t valu e in th e so cial. W h at is m o st valuable in this context is that each engagem ent can potentially create m ore engagem ent, each Like of a Facebook user is meant to produce more likes of their Facebook friends, a shared UR L is meant to produce com m ents and likes, a com m ent is m eant to produce a response, a tw eet to produce a retw eet. e value of so cial m ed ia activities is b o th situ ated in th e p resen t an d in th e fu tu res, in th e actu al b u tto n co u n t an d in potential more counts. To put it in Nigel rift words: “value increasingly arises not from what is but from what is not yet but can potentially become, that is from the pull of the future, and from the new distributions of the sensible that can arise from that change” (rift 2007, 31). By building on the com bined dynam ics of th e h it, lin k , lik e an d sh are, th e em erg in g L ik e eco n o m y is creatin g a fab ric o f th e w eb th at cap italizes o n th e value of any potential social activities.
42 A n o v e rv ie w o f F a n a n d L ik e -se llin g se rv ic e s c a n b e fo u n d h e re : h ttp ://w w w .q u ic k o n lin e tip s.c o m /a rc h iv e s/2 0 1 0 /0 9 /b u y -fa c e b o o k - fan s-frien d s-lik es/ 43 h ttp ://w w w .b u y re a lfa c e b o o k fa n s.c o m / 44 h ttp ://b u y -fb fa n s.c o m /
26 Acknowledgments
anks to Erik Borra from the Digital Methods Initiative, University of Amsterdam for building tools for our research .
References
Addthis, 'Sharing Trends in 2010', Addis Blog, 29 December 2010
DeLanda, Manuel, A new philosophy of society. Assemblage eory and Social Complexity. L o n d o n : C o n tin u u m , 2006a.
27 DeLanda, Manuel, 'Deleuzian Social Ontology and Assemblage eory.' 2006b. In: Fuglsang, M. and Sorensen, B.M . Deleuze and the Social. E d in b u rg h : E d in b u rg h U n iv e rsity P re ss, 2 5 0 -2 6 7 . Deleuze, G. and Guattari, F. A thousand plateaus: capitalism and schizophrenia. L o n d o n : C o n tin u u m , 2 0 0 4 . Facebook and M edia, 'e Value of a Liker,' 29 Septem ber 2010,
28 Siegler, M g, ‘I ink Facebook Just Seized Control O f e Internet’, TechC runch, 21 April 2010
29