Version classiqueVersion mobile
OpenEdition Books

What is digital identity?

Olivier Ertzscheid

Tools, markets and strategies

Texte intégral

Who wants this data that covers (almost) everything about us?

1The question of personal data is at the very heart of digital-identity issues. This data can be:

  • technical: my IP address, or cookies stored by my browsers
  • declarative: for example, my Facebook profile and information I deliberately provide
  • navigational: when a search engine or site collates my searches and information about my different types of online purchases
  • behavioural: in other words, based on my browsing history when this option is activated, which often is the case “by default”.
  • 1 Territoriality means that laws are applicable and valid only within a national territory.

2French law contains a number of data safeguards. These limit the length of time that personal data can be stored, prohibit such data being shared with third parties and afford individuals the right to have their data removed. However, in everyday life it is both tricky and time-consuming to make the major players implement these principles. The territoriality of law1 is the multinationals’ favourite Trojan horse when it comes to circumventing national legislation.

  • 2 This is a common corporate expression and is invoked by market actors each time their activities a (...)
  • 3 Olivier Ertzscheid, “Bing. Je fais confiance à Microsoft …”,, 28 May 2009, http:// (...)

3“Anonymity reduces the relevance of search results.2” That, in any case, is the rationale search engines put forward to both defend the increasingly personalised search results they display and stress why we need to let browsers save our browsing history. To fully understand this – partly justified – argument, let’s turn to the search engine “magic square3”, i.e. the four main types of searches that are most common in our interconnected world:

  • SHOPPING: searches to do with online purchases and product comparison
  • HEALTH: searches to do with health or finding a medicine, treatment or dosage
  • TRAVEL: searches to do with personal holidays or business trips
  • LOCAL: so-called “local” searches to find a restaurant or pharmacy near to where the user lives.

4Heightened personalisation and an increasing number of geolocalisation techniques do indeed appear necessary for each of these macro-behaviours, even before the user types in their search term. This seems to be to everyone’s advantage, at least to some extent: the user doesn’t have to type in their address each time they want to find a nearby pizzeria, and search engines can use the information required to personalise results as a way of optimising the increasingly targeted, contextual and personalised adverts they display.

Log files and cookies: a spot of cookery

5123.45.67.89 - 25/Mar/2007 10 :15 :32 - http ://www. ?q=flowers - Firefox ;Windows NT 5.1 dac465aa86edd2dc

6You have just read a log file. Each time a user connects to a page, the server that hosts the page adds a line of code to the log file.

7The above log file shows, in order:

  • the user’s IP address (every computer connected to the Internet has a unique IP address, which is roughly the equivalent of our own social security number)
  • the date and time
  • the site and search request: in this example, a search for “flowers” on Google
  • the browser and operating system used, followed by a list of settings giving details on various processing procedures applied to the request, for example.
  • 4 Garett Rogers, “Google now expires cookies after 2 years, not 31”, ZdNet, 16 July 2007, http://www (...)

8In addition to these log files, many sites require the use of small text files known as cookies in order to function. Cookies, which are used and saved on the user’s computer, contain information on the latest activities performed on the site. While these cookies are enormously useful, some sites use them in pernicious and invasive ways. For example, between 2003 and 2005, Google introduced a cookie that would not expire until … 2038!4

  • 5 Christophe Auffray, “Microsoft s’engage à anonymiser les données de Bing après 6 mois”, ZdNet, 19  (...)

9Through the meetings of Europe’s national commissions on information technology and liberties (Commissions nationales de l’informatique et des libertés, CNILs), public authorities are pushing hard to oblige search engines and leading Web actors to use this identification data in a reasonable way and for a limited length of time. They have thus far succeeded in getting Microsoft to commit to implementing their recommendations, which involve deleting cookies after six months5.

10Google, however, has maintained its policy of deleting cookies after 18 months. Furthermore, while the anonymisation of IP addresses is now a reality, this anonymity applies only to the last two numbers in the address, thereby allowing a great deal of latitude in terms of how the personal data they transmit is reused.

11Above and beyond the technical and legal issues, the definition of a user’s identity on the network is also radically altered. This has been noted by Jeff Chester, the executive director of the Center for Digital Democracy, a lobbying organisation that defends digital privacy:

This idea that a cookie is non-personal information no longer really applies in this digital age. You don’t need to know a person’s name to know a person, to understand their likes and their dislikes, the contents of what they need, what they put in their shopping cart. It’s really personal now.

In short … Choose the monitor rather than the (smoke)screen


12Digital identity is a collection of digital traces that can be displayed on a computer screen – or monitor – and “tracked” by programmes behind the screen.

More monitoring

13Digital identity requires constant surveillance or monitoring. This surveillance necessarily involves “tracking” techniques and methodologies.

Still more monitoring

14Given the technical, legal, social and behavioural issues at stake, empirical “learning by doing” should be replaced with specialised training, acculturation and support that allows users to understand, analyse and circumscribe their digital identity. Such training and support should form part of most school and university curricula.

The digital identity toolkit

  • 6 allows you to check whether a username (or the name of a business or brand) already ex (...)

15A whole range of tools exist to measure the perimeter of one’s digital identity, implement reputational strategies or simply check the availability of a name or brand on a given service, site or social network6. We present below a few of these tools and the questions they can help us answer.

How do I know what people say about me? Search engines

16The easiest way is still to type my name into Google or another general-use search engine. This will identify sites within my control (articles or posts I have written or websites I have contributed to), but also identity traces that escape my control, i.e. what other people say about me, my brand or my business.

How do I know what people know about me? Aggregators

17Several sites act as “aggregators”, in other words they retrieve information connected to my name from different sources and present this information synoptically. The most well-known of these sites was, which displayed my address (retrieved from the Yellow Pages), my photos (taken from Flickr or Google Images), my articles (published on my site or blog), the domain names I own (or can buy), my videos (retrieved from YouTube), and my “contacts” (extracted from social-networking sites that authorise this practice).

18The effectiveness of these tools is naturally correlated to my actual online activity and presence, as well to the privacy settings I have chosen for my different accounts.

And if I really want to know everything about a person? Neighbourhood search engines

  • 7 The cost of these services varies from $2 to over $50.

19In this case, you need to turn to what we might call “neighbourhood search engines”. Though currently available only in the United States and on a fee-paying basis7, and provide very detailed information on a given individual:

  • their address and social security number
  • records of their house-moves and changes of address over the last 10 years
  • their date of birth and extracts of any criminal record (including various types of fines)
  • the dates of their marriage(s) and/or divorce(s)
  • the names and telephone numbers of their neighbours
  • the value of any properties they own as well as the date they bought it/them and the name(s) of the people they bought from
  • the size of the lot on which their house is built
  • a satellite image of their home
  • the average age of the people living in their neighbourhood
  • the number of murders, rapes, vehicle thefts and other thefts in their neighbourhood
  • the ethnic composition of the neighbourhood, the level of education and the languages spoken
  • the licences they possess (driving licence, sailing licence, flying licence, motorbike licence, etc.).

20There’s no need to hire private detectives to find out this information. In the United States, such information is freely accessible in different government databases. All that remains is to collate the information systematically and provide it on demand.

21For those wishing to rigorously police the perimeters of their online identity and track mentions of their name in even the most obscure discussions forums, a more fine-grained or systematic approach is needed. This requires systematic tracking and tools that are compatible with the types of content one most wishes to target. There are now professional software solutions specially designed to perform this kind of monitoring or surveillance of one’s name or company name.

What do specialised e-reputation tools have in common? Working in a similar way to meta- search engines or aggregators, they exploit defects in – or the absence of – privacy settings on websites. They systematically explore what is known as the “invisible” Web, which ranges from discussion forums to government databases.

The e-reputation market

22Google the definition of “e-reputation” and you will find an impressive list of advertising links to companies specialised in “image management and protection”, “influence”, “rumour clean-up” and “brand buzz analyses”. Online reputation has become a market, and a very competitive one at that. Among the array of services on offer, I can take out a yearly subscription which, for €15 a month, will keep me abreast of what others say about me (or my business) on the Web. A subscription of €100 a year will allow me to work up a “good” reputation, or even – and this is a very popular option among North American parents – to be kept informed of my children’s online activity (their friends on Facebook, the photos they post, the messages they send and receive, etc.).

  • 8 Search request “define: e-reputation” on and Le Manifeste des évidences, http://www.cluet (...)

23Assuming that “markets are conversations8”, and given that conversations can make or break reputations, it was inevitable that online reputation would eventually become a market.

24To get a clearer idea of the main groups of players in this market, it’s useful to take a look at Aref Jdey’s diagram of the e-reputation market.

How can I counter online defamation?

25A person who wants to control their e-reputation, or who is the victim of online libel or slander, has several lines of defence open to them.

26Si vis pacem, para bellum. If you want peace, prepare for war. An initial approach, once you have chosen to expose parts of your life on the Web, is to regularly track mentions of your name. Generally, a little ego-surfing – typing your name or company name onto a search engine – is all it takes to stay abreast of possible attacks and defuse them before they get out of hand.

27Various firms or agencies made up of legal experts, computer scientists and communications staff also offer services to “clean up” their clients’ online reputation.

28The first cleaning method involves explicitly requesting site managers, and above all search engines (see, for example,​websearch/​troubleshooter/​3111061?hl=en), to remove the incriminating content, sites or remarks.

29The second method involves “drowning” or burying negative results. This consists of creating “false sites” that talk up a person, their business or their products. Care is taken to ensure that these sites are well referenced or interconnected, and thus, gradually, the negative results are relegated to the second or third page of search engines. As very few users look any further than the first page of search results, this also amounts to “cleaning up” one’s reputation.

30The third method involves taking legal action, or hiring a lawyer to write a strongly worded letter that arouses fear of prosecution in those responsible for the attacks. This method is used as a last resort or when the first two methods have failed.

31Finally, insurance companies are now starting to offer contracts directly geared towards e-reputation protection, thereby confirming that these issues have reached a new level of commercial maturity.9

The digital right to be forgotten?

  • 10 Serge Tisseron, “Le droit à l’oubli sur Internet : une idée dangereuse”, Libération, 4 December 20 (...)

32Despite the praiseworthy efforts of governments, particularly in France, to promote the digital right to be forgotten, it is highly likely that this will go no further than charters or other recommendations devoid of any contractual value. The implementation of such a right to be forgotten runs into several major obstacles: technology moves much faster than legislators or regulators, and legislative plans, which by their very nature are limited to a national territory, are rendered ineffective by globalisation. Lastly, the disruptive logics of digital technologies often thwart any attempt to predict the future: how could anyone have predicted the rise and phenomenal popularity of Wikipedia or Facebook only a few years ago? What’s more, users’ willingness to reveal large swathes of their private lives is often the implicit condition for popular sites to remain free of use. The segmentation of profiles that this enables means that these websites have commercial value, and major market actors are in no hurry to forgo this cash cow. Even if they were to do this one day, the psychiatrist Serge Tisseron recently pointed out that “the right to be forgotten might also encourage people to forget the law, and particularly image-protection rights: anything goes because anything can be deleted10.”

The four pillars of digital wisdom:

Protect and reserve your name

Get involved

Decide what information to keep private

Stay vigilant


1 Territoriality means that laws are applicable and valid only within a national territory.

2 This is a common corporate expression and is invoked by market actors each time their activities are legally or politically called into question. It stresses the fact that anonymising search requests would lessen the quality of the user experience.

3 Olivier Ertzscheid, “Bing. Je fais confiance à Microsoft …”,, 28 May 2009,, accessed 30 October 2012.

4 Garett Rogers, “Google now expires cookies after 2 years, not 31”, ZdNet, 16 July 2007,, accessed 30 October 2012.

5 Christophe Auffray, “Microsoft s’engage à anonymiser les données de Bing après 6 mois”, ZdNet, 19 January 2010,, accessed 30 October 2012.

6 allows you to check whether a username (or the name of a business or brand) already exists among a very wide range of participatory sites and social networks, for example.

7 The cost of these services varies from $2 to over $50.

8 Search request “define: e-reputation” on and Le Manifeste des évidences,, accessed 30 October 2012.

9 Sabine Blanc, “E-réputation et bénefs assurés”, owni, 27 March 2012,, accessed 30 October 2012.

10 Serge Tisseron, “Le droit à l’oubli sur Internet : une idée dangereuse”, Libération, 4 December 2012,

Table des illustrations

Titre The e-reputation market map
Crédits @ Aref Jdey Source :​2009/​08/​24/​e-reputation-market-map-carte-du-marche-de-la-e-reputation/​
Fichier image/png, 336k