This method consists of answering the following questions, in this order.
- What is the indicator you want to maximise?
- What are your action levers, or what levers do you plan to implement?
- What analyses, or rather what results, will enable you to choose and parameterise these levers?
- What data is needed for these analyses? And how to get it?
As you can see, data is not the starting point, but the ending point of any (good) data project. Here are some ideas to answer these questions.
What is the indicator you want to maximise?
So this is probably the most important question.
Most of our clients have many goals. Just look at the number of objectives set up in your analytics tool.
Remember #1: it should be forbidden to launch any project without knowing which indicator it is supposed to improve.
To prioritize your objectives among them, the best method is to go back to the reason for your website's existence. On your site, are you primarily addressing your customers or your prospects? Do you want to sell or generate leads? Do you want to increase your margin or are you aiming for volume? etc.
Once you have a shortened and prioritized list of indicators, it becomes much easier to identify and, above all, to make good use of the levers that will allow you to optimize them.
What are your levers for action? Which lever(s) do you plan to implement?
Apart from the competitive and economic context, you will have no impact on your key indicator without action. So here we need to establish a list of actions that can improve or, at least, influence your priority indicators.
On the web or in the digital world, the types of actions that can be envisaged are infinite, but they all have one thing in common: a site is always a communication interface between a company (the site owner) and its visitors.
Most of the levers available on the web therefore consist in generating or modifying messages or stories (series of messages) according totriggering events.
These messages can take many forms: site pages, e-mails, notifications, pop-ups, pop-ins, advertising, which are all different levers, whose effectiveness will have to be measured according to their potential (or tested) impact on the targeted indicator.
Quick thought #2: it should be forbidden to implement a solution without measuring as cleanly as possible its impact on the targeted indicator.
What are the results of optimising these levers?
The events that trigger an appropriate change or message generation can be simple or the result of complex analysis. It doesn't matter.
At Potions, we love applied mathematics. But we have often found that the simplest trigger events are the most effective.
The result of a complex algorithm will often confirm obvious business intuitions.
Here's an example.
A user will visit your site on average 3 to 5 times before converting. So, you don't need to follow his navigation on the whole web to know which product interests him... But are you able to adapt your site to his previous visits? And by the way, on your site, do you know how many visits your prospects need to make to convert?
Remember #3: always measure the value of a solution by its impact on the targeted indicator, not by the "beauty" or complexity of the results it uses.
What data is needed for these analyses, and how can it be retrieved?
As we said, the data available to websites is often more than enough to adapt their message to the situation, i.e. to "personalize the experience" of their visitors. Unfortunately, they are often unable to use it wisely without affecting the fluidity of navigation.
For example: to use customer data (CRM), many sites ask their visitors to log in or create an account systematically. This is one more step (sometimes too many?) in the commercial relationship (see our article on log-in).
Remember #4: data retrieval does not justify everything... think above all about the experience on your site!
Depending on the triggering events to be implemented, the necessary data to be retrieved can be a simple click on a button, or the whole set of pages seen on the site for one month. To retrieve this data efficiently without multiplying the round trips with your servers, all solutions developed by Potions work on the fundamental principle of the SHORT CIRCUIT OF THE DATA.
Who are we?
Potions offers solutions for optimizing the performance of websites that rely on edge computing.
This technology uses the databases of browsers to store, analyze, enrich and activate browsing data locally, and without cookies.
This makes their operation simpler (no server development required), faster (no server call, updating each terminal in real time), less expensive (much less setup and maintenance of databases), and more respectful of privacy and the environment.
Contact us to learn more!
