Personalized onsite navigation is a key asset of an ecommerce, however, it is often not known how it can actually be possible to provide people with an experience that is effective on a business and communication level.
Let’s try to shed some light on what drives personalization and what data help us recognize the user.
First step. Registered user or anonymous user?
It is now well known that in order to offer a personalized experience to a user, the user does not need to be registered/logged in. People leave continuous traces of their passage on the web and, thanks to simple cookies and more advanced technologies such as fingerprints, there is a lot of information that is useful to maximize the browsing experience from the very first minutes: for example, it is possible to propose personalized content based on the pages that are viewed or the search keyword from which the visit came . Let’s try to summarize some of the information based on which segmentation could be done, starting with the anonymous and then registered user.
Anonymous:
- visit acquisition channel
- pages viewed or clicked
- geographical origin
- referral
- tags and categories of page views and clicks
- device used
- words searched on the internal search engine
- abandoned trolley
Registered:
- self-profiling
- order history
- wish list
- Opens and clicks on emails or the site
- lifecycle
Each of these pieces of information can define a rule or filter that allows for the creation of increasingly refined customizations based not only on the user’s online behavior (if you saw this product then you might like this other one), but also on latent demand that adds up to several variables at once. How is it possible to do this?
Second step. From history data to contextual data
First, the past and the present should not be information managed in separate silos. Online browsing is contextual therefore personalization cannot be based only on static and historical information but must be dynamic, that is, taking into account what is happening at that precise moment.
This is especially true when running ecommerces with a large product catalog, which does not necessarily correlate easily from the last product purchased. Users buy a certain product, but they may not necessarily want to hear only about it. So you have to try to think about complex segments that take into account who the user is and also what they want at that moment (for example: they bought a TV of a certain brand, they have a certain spending potential, but now they want a new vacuum cleaner and not a remote control!).
As we have already written in this post, there are at least 3 good reasons to focus on personalization: from increased sales, to creating a loyalty relationship with the user. For this we need to be able to intercept ever-changing needs. This is where marketing automation helps us: equip ourselves with solutions that learn in real time and automate based on what is actually happening.
Third step. Move from “who you are” to “how you are”
One example is the Sephora brand, which, for the past 4 years, has been implementing a technology distributed to its physical stores that allows it to classify its customers according to skin type, with the aim of offering them make-up products that are better suited to their natural skin tones.
This results in a series of online shopping suggestions created with the premise of supporting the customer in selecting a product that is closer to the need and therefore more satisfactory. Ranking is done with the support of sales assistants who, by simply taking pictures of female visitors, enrich the database with vital information. (The whole story can be found at this link http://digiday.com/marketing/color-iq-sephoras-shade-matching-skin-care-tool-boosts-brand-loyalty/)
The case of Sephora confronts us with a paradigm shift. Personalization of online browsing is not just a topic of product correlation. Sephora starts from “who you are” to “how you are.” So the further consideration to be made is whether it is not possible to use data to imagine similar user segments, with similar behaviors, which helps us in suggesting interesting alternatives they are not thinking about. Are you such a user? This is the best answer we have come up with to meet your need.
Of course, the greater the complexity we want to achieve, the more effort will be required to be able to collect data, normalize it in order to be able to automate actions that reflect our marketing and sales strategy. Working on multiple variables simultaneously will require structuring a dynamic offering that updates in real time, in a virtuous circle of continuous learning. It is called artificial intelligence and it is the driving force behind personalization not only while browsing the ecommerce site but, as we will also see in all subsequent steps, on direct and third-party channels such as facebook.
