Analyze your digital
marketing data

First of all, let’s remember that this is a completely free solution. However, it is essential to create an account on Gmail to be able to use it, as it is also the case for all the other services that Google offers. Then you have to create an account for the company/site that will use it. Once the registration form has been completed and the terms and conditions have been validated, it’s time to get serious. It is now time to retrieve the GA script, which will then have to be implemented on the site. Although simple, retrieving this code is not always easy, especially because of the richness of the interface, the abundance of features and information available. To avoid possible problems and unnecessary waste of time, it is preferable to entrust this task and the implementation of the code on the site to professionals, to a web agency in particular.

Once the tool is installed, you have to set it up to be of any use. Google Analytics will indeed collect a considerable amount of information about the traffic generated, the number of visits and unique visitors, their origin, their material, their location… This data will also concern the time that the users spent on the site, the pages they visited, the links they clicked on, the purchases they eventually made, the forms they filled in, the ebooks they downloaded, etc. Depending on the site and the business model, some of this data is more interesting than others. It is therefore essential to make the appropriate settings to have access to it. It can also be interesting to define destination, duration or event objectives, or even to establish “conversion tunnels”, thanks to which it is possible to follow very precisely the path of visitors on a site. Everything is possible, or almost, as long as you know the settings to be made.

Because it is rich, powerful and efficient, GA is able to provide its users with a considerable amount of information, from the most general to the most precise. To do this, it must be carefully configured, in particular to block unwanted traffic. An Internet site is indeed systematically visited by “intruders” and by robots of all kinds. They generate unwanted data which, when taken into account, alter the reliability of the figures and analyses provided. Fortunately, GA makes it possible to exclude such data via a number of filters, which are made available on the platform. Once these filters have been set up, the traffic from bots and other spammers will be excluded from the collected and analyzed data. The summaries provided by Google Analytics will therefore be much closer to the real performance of the analyzed website.

According to Google Analytics, a channel corresponds to a traffic source. To classify the data it retrieves and analyzes, it classifies these channels into groups, according to default models. GA thus groups the different flows into “Organic search” for traffic generated by searches made on search engines; into “Direct” to list Internet users who have directly entered the website url; into “Display” for traffic brought by a banner, for example, or into “Social Network”, to group together visitors who have arrived on the site via social networks Since they are by default, they often only partially meet the needs of each. Hence the interest, even the necessity, to create oneself – or preferably with the help and advice of specialists – one’s own personalized channel reports.

Complementary to GA, the Analytics E-Commerce script is specially designed for sites that sell products or services online. It is not integrated to the “classic” tool and must therefore be added. For a site developed via a CMS, such as PestaShop or Magento, or even via the WooCommerce module of WordPress, it is likely that this Analytics E-Commerce script is already integrated. If this is not the case, it is then necessary to integrate the script yourself, which we provide. Once this is done, you need to make a few adjustments, including the integration of certain values that are essential for the information relating to transactions – number of sales, price, shipping costs, date, etc. – to be correctly extracted. – are correctly extracted. Simple for those who are not afraid of a few lines of code, these manipulations may nevertheless require the intervention of specialists.

Which products have been sold on your site? Which ones are successful and which ones don’t seem to appeal to customers? To help you answer these two questions and more, GA has developed a new module called “Enhanced E-Commerce”. What exactly is it? In a nutshell, this module allows merchant site owners to better monitor their KPIs, the so-called key performance indicators. This new tool helps to understand and analyze the behavior of Internet users. Thus, it is now possible to know on which product a customer has clicked (click); which page of details of another product he consults (detail); which article he has added (add) or on the contrary removed (remove) from his cart; when he starts the ordering process (checkout), etc. These are all very useful indicators for, for example, developing effective conversion tunnels.

If, thanks to GA, we can learn a lot more about the traffic generated by a site, about the behavior of Internet users on that same site, about the profile of those Internet users, etc., we can also learn a lot about the products offered for sale on that site. With the “Product Performance” reports offered by GA, we have two options, two different ways of approaching this product aspect: the “summary” and the “buying behavior”. Clearly, the first one focuses on the sales history (turnover per product, different single products purchased, total products purchased, etc.). The second one focuses more on the steps prior to the purchase (consultation of the list of products, detailed consultations of a product, products added or removed from the cart… The “Product Performance” of Google Analytics allows to know more about the “life of the products”, which is essential for an ecommerce site.

However powerful and efficient it may be, GA does not track all the activity of a website. There is always a small proportion of traffic and events that escape it. It is estimated that 80% of orders placed on a website are recorded by Google Analytics. If a comparison between the actual sales and the figures provided by the tool shows a proportion of this order, it is normal. On the other hand, if this proportion is lower, or even significantly lower, it is likely that there is a problem in the settings. To remedy this, you must first find the origin of this problem. For example, the site may offer several confirmation pages, while the tracking script only follows one. The different payment methods available may also not send to the confirmation page(s) at the end of the sales, etc. The causes of these malfunctions can be diverse, as can the corrections to be made.

Most advertisers choose to reward only the last click. This is an easy solution that not only does not take into account the contribution of the other links upstream of this last click, but also leaves in the dark a lot of important information about the path followed by a customer. However, other models are possible, such as the one that consists in attributing an equivalent share to each of these links. This is known as the linear model. Another option is the decreasing model, according to which the commission decreases as you move away from the last click. Google Analytics allows you to set up your own models, via conversion funnels, multi-channel conversion funnels or the attribution modeling tool. The expertise of an agency in terms of conversation paths can be of great help, especially if it is Google Partner Premium certified.

If everything went well during the installation of the Google Analytics script, the traffic information should be returned after about 24 hours. Unfortunately, this is not always the case, for several reasons. While GA retrieves, summarizes and presents the traffic data, it is Google Tag Manager that collects and processes this data before sending it to GA. Google Tag Manager’s filter settings may not have been properly set, which blocks the transmission of information. The malfunction may also be due to a bad installation of the Google Analytics script. Indeed, even if you take the greatest care, it can happen that you do not place it in the right place, that you implant the code of another account if you have several sites, or that you alter the code by using a word processing software that adds spaces, transforms quotation marks, etc. In this case, it is no longer possible for the script to work properly. In which case it is not the same code at all…

GA makes it possible to know everything, or almost everything, about what is happening on a website, in terms of traffic, Internet user behavior, searches, product purchases, etc. However, without even getting into the question of the precise settings that can and should be made to this tool, there are a number of mistakes that should be avoided. The first is not to implement the GA script correctly. No data will be collected. The second is to make a mistake in the company’s filters, in which case all visits to the site, including those of company members, will be counted. For a small company with three employees, this is not too serious; for a multinational, it is more serious. It is this same logic that makes it essential to filter the robots, which visit all the online sites every day… And then there are also the mistakes that consist in defining the wrong objectives, not associating different modules, etc.

KPIs, for Key Performance Indicators, are key data that allow you to obtain a global vision of the health of a website, but also very precise. Thanks to these indicators, we can know how many people visit the site, distinguish between new and regular visitors, know how they arrived, how long they stayed, how many pages they consulted, which ones they preferred, etc. These KPIs also provide valuable information about the success or failure of a marketing operation, via the information transmitted in real time, about the conversion rate or the devices used by Internet users to visit the site. They also allow, via the information transmitted live, to observe the possible fallout of a marketing operation. In short, these are essential elements to distinguish what works or does not work on a site, what can or must be improved.

What is the bounce rate? It is a percentage that summarizes the proportion of Internet users who, once they have arrived on a site, leave it immediately. The higher this rate is, the less a site manages to retain its visitors. This is not a good sign. However, a site can immediately provide the required information and, in this sense, satisfy the visitors, who certainly do not stay long, but often come back. For some sites, on the other hand, it is necessary that visitors stay a minimum of time, to buy for example. So there is not one rate, but several bounce rates, and the “right” levels depend primarily on the different types of sites. However, other parameters come into play to explain good or bad bounce rates: navigation comfort, long loading time, content organization, etc. In these cases, it is possible to make corrections.