The implementation phase of a web analytics project is critical. We regularly see implementations that go to the end of a website project and for which the recipe will be set aside. It is necessary to avoid any risk that would affect the data.
Bad tracking can result in two ways, loss and/or inconsistency of data. In both cases everything that will be done after the implementation, such as data analysis, will be questioned because it will not be based on something coherent. Discover the following six tips for a quick and easy implementation that will give you confidence in your data.
1. Having finalized its tracking plan
The start of any implementation project must be realized by making a tracking plan. This plan will allow developers in charge of implementation to move forward calmly knowing exactly where they are going.
It is essential that this phase of reflection is carried out before working on the implementation, this will allow two things:
- In the first time the technical teams will be able to estimate precisely the necessary work and therefore the dates of putting into production possible. Dates are quite critical since it is quite possible to iterate the releases according to the objectives of the tracking plan.
- The second advantage is the time saved. As reflection is carried out upstream, all that is left to do is to carry out the technical tasks without questioning the formats, the need to raise a button or not.
2. Identify the technologies used
3. Do not stop at the first obstacle
As soon as we are talking about tracking we are actually talking about web development. In this environment, it is common to face difficulties related to technologies. Often we meet web analyst teams telling us that it is not possible to measure some KPIs due to technical complexity. It is a pity to stop there, the measurement of certain information may indeed be more difficult to set up, but it is very rare that it is completely impossible (and “in computer science, everything is possible …”).
In this type of case it is interesting to get closer to the technical teams of the provider directly: it is indeed unlikely that you are the only ones to have this problem, and even if you are an isolated case,
4. Choose the right tracking methods
These three methods are completely different and they each have advantages but also disadvantages. We advise you to choose the methods used according to your priorities.
5. Think about site performance
The performance of websites is a necessity for many actors. Nowadays, when a site is loaded by a browser a lot of calls are made, some sites exceed well over 1,000 calls per page. A call to a Digital Analytics tool rarely exceeds a few tenths of a second; however there may be a hundred of them on the site, a wait of several seconds for the user. However, it should be noted that everything is not so simple: the performance of a site depends only on a sum of calls.
There are techniques to lighten the load. For example, asynchrony (i.e., loading multiple scripts at the same time) and loading scripts at the bottom of the page generally improve site performance, present. Also in some cases we cannot wait until the end of loading the page to send data. It is therefore essential that the digital analytics solution is implemented in the lightest possible way.
If you do not have a video on your site, why would you need to load the code to measure videos? There is no reason to do this and we recommend that you lighten up the code you are going to use. In addition, if you have 30 banners on your site, it is often simpler and lighter to send the information at one time rather than thirty times. All these tricks will help improve the performance of your digital analytics tools and thus your site.
6.Test, test and re-test
This point cannot be overemphasized. This is the most important step.
As soon as an implementation is carried out it will be necessary to test to make sure of two things:
- First, the data will be sent out when you want. If the data is not sent to the servers, it will be lost and can never be retrieved. It is for this reason that we must be able to ensure that everything works perfectly.
- Secondly, it will be necessary to check the good return in the interface of the solution digital analytics and the consistency of the data. Indeed, once everything is sent it may not be enough, if for example the page name is incorrect it will be complicated to find the information you are looking for.
Each site and each technology has its peculiarities. Because of this, code that has worked on 10 of your sites may not work on a 11th. Because of this, even when you implement code that you already know is functional, it is strongly advised to test it again to make sure everything works.
Lastly, even when updating your site it is advisable to check the tracking codes, the change of a button or any other element could have an impact on the tracking. In addition, when updating the site, it is not uncommon for the web analytics script to be removed for various reasons.