fr | en | es | de

Site custom variables

Discover the latest tools in the
Analytics Suite, thanks to our new online help.

Contents: Hide

The purpose of site custom variables is to associate specific information of your activity to your website (interface language, loading time, number of articles per page etc).

This information is placed in the xi variables, where i represents the ranking of the variable and has a value of between 1 and n.

Each variable must be declared in the interface, and this is possible by carrying out the following steps:

Should you trace more than one site, you can duplicate objects (level 2 sites, campaigns, site custom variables, rules for alerts, events which are saved in the event log, identified visitor categories) from one site to another; making it easier for you to declare your different objects. However, any duplication will be refused if an object which is to be duplicated already exists (same name or same ID) on a destination site.


We strongly recommend that you check that the analysis is not already available before you create a new one, as there is a limited number of site custom variables. Examples of standard analyses include: identified visitors, search engine keywords, country, visitor category, etc.

The variables (x1, x2, x3, etc) are concatenated together in the main tag's xt_multc variable.


This example takes into consideration 2 custom variables (x1, x2) which are used:

<script type="text/javascript">


xtnv = document;  //parent.document or top.document or document        

xtsd = "http://logxxxx";

xtsite = "xxxxx"; //site number

xtn2 = "";        //level 2 site

xtpage = "";      //page name (with the use of :: to create chapters)

xtdi = "";        //implication degree

xt_multc = "&x1=12345&x2=49367";    //customised indicators

xt_an = "";       //numeric identifier

xt_ac = "";       //category

//do not modify below

if (window.xtparam!=null){window.xtparam+="&ac="+xt_ac+"&an="+xt_an+xt_multc;}

else{window.xtparam = "&ac="+xt_ac+"&an="+xt_an+xt_multc;};




<script type="text/javascript" src=""></script>



<img width="1" height="1" src="http://logxxxx.xxxx/hit.xiti?s=xxxxx&s2=&p=&di=&ac=&an=&x1=12345&x2=49367" >


Types of variables

Variables can be in different formats (text, numerical, date, country, duration).


The "ID/date" variable is a piece of information which can be made up of several different values.

By default you can declare up to 6 ID/date variables.

For "ID" variables, the link between the labels and the ID values which are entered into the tag code can be obtained by importing an Excel or csv file onto this interface.


- The information which can be seen in the "Rank" column is entered automatically whenever a variable has been confirmed. It corresponds to the name of the variable that must be entered in the tag code.
- The rank number can only be used for one variable. Closing the variable does not mean that this number can be used for another variable.
- A set of variables is defined for a particular version for a given period of time. If you add a variable to this set, the set remains the same. However, if you remove a variable, the set is closed and a new version is created. Each time a set of variables is changed there is a break in the history.
We therefore recommend that you limit the number of changes which are made to sets in order to avoid discontinuity of the data that is displayed in relation to the comparison period.
- The "short" label is displayed in the column title of the analysis table.
- The following format must be respected during the tagging process, for the "Date" variables: yyyymmdd.


It is possible to add customised text variables using the following format : xi=[my_text]. The main advantage is that you can retrieve values which can then be read immediately in your analyses. By default, you can declare up to 2 variables of this type.


These variables collect information which can be added together or which can be used to calculate an average. For example, number of articles,and number of nights spent in a hotel can be analysed independently or as a new metric.

By default, you can declare 2 variables of this type.


This type of variable allows you to indicate duration, expressed as a number of days (for example the duration of a stay, delivery deadlines etc). This type of variable can also be used to work out the total values of metrics (or can be used to calculate an average value) in your analyses.


The "country" variable allows you to have an accurate analysis inlcuding a worldmap representation. The format of this variable must respect the ISO code of the country in question, for example FR, BE, CA etc.


The format of this variable includes both decimal points and commas. Decimals will be rounded off to two decimal places should there be more than 2 figures coming after the decimal point or comma, for example 24.4563 will be rounded up to 24.46.