Tracking is the process through which users’ information is collected during their navigation sessions. The information is collected by tracking cookies on their computers. These are installed when the webpages are loaded, because tracking cookies are embedded in the source code of the visited website.
Why to track?
The tracking process is desirable because of the following benefits:
Tracking allows advanced analysis of traffic information. It is possible to gather more granular information about users’ behavior. For example, tracking captures, how much time they spent on the website, from which page they left and so on.
Tracking tools help at defining more precise and granular reports.
Testing and Optimizing
Tracking methods are useful to correctly assess the performance of marketing channels. They help at testing hypotheses and optimizing accounts according to measured performance metrics.
Errors or problems might easily occur, but checking tracking data on a daily basis may help at spotting them for an immediate correction.
How does the tracking happen?
When the user puts in a request for a website, its source code is loaded by the browser. This code contains a tracking cookie that is activated and is placed on the user’s computer. If the same cookie was there already, it gets updated.
Cookies from Google AdWords expire after 30 days. This implies that information about a conversion will be counted for the day the cookie was installed, even if it happens 29 days later. This fact biases various reports up to 30 days backwards.
Apart from conversions, the cookie may also track the time on the website, all visited pages during the session as well as bounce rates. These metrics are presented either in Google AdWords, Adlens or in Webtrack.
Multi-Channel is described as follows: whenever a single lead has fired several codes for different marketing channels (e.g. Search, Affiliate, Site, Display Network, SEO) before executing the sale, a problem for the assignment of the conversion premium occurs. The lead is usually assigned to the last cookie that gets fired. However, when using advanced software (e.g. Cookie Loader) we can use rules for assigning a share of the premium to each channel that contributed to the sale.
In order to correctly evaluate (and reward) the effectiveness of different channels, it is important to use internal tracking systems. On the long run such programs will help in the optimization process of accounts.
However, when internal tracking systems are not available or did not work, the click has to be assigned to the last cookie.
- Easy implementation: only inserting some code in the website source code is necessary.
- Free of charge.
- Tracking of multiple conversion codes (e.g. sales, sign-ups, newsletter subscriptions, basket page visits, etc.).
- Possible to monitor the performance of the website along the whole conversion tunnel (i.e. landing page, products pages, basket page, payment page, sale confirmation page).
- The tracking cookies expire after 30 days.
- We can only gather information regarding the click date.
- Only revenue metrics are traceable (contribution margins missing etc.).
Conversion funnel behavior
- Place several tracking codes in the website in order to study the behavior of the visitors.
- Identify possible hurdles in the conversion process and try to fix them.
- Observe the revenue flows brought by different channels.
- Study the average revenue per conversion.
- Try to cluster and study your targeted audience.
During the process of tracking, some problems might occur:
- Conversion data disappearing from one day to the other. This is usually caused by an accidental manipulation of the tracking code in the website source code.
- Conversion data significantly drop. This might be caused by the misplacements of the tracking codes or by the excessive loading time of a pixel. For example, when on “Thank-You-Page” the tracking code is placed at the bottom of the source code, the user may leave the page before the loading process is complete. Thus, always place the tracking-code on top of the source code. Using Firebug we are able to measure the loading time of each pixel.