Today’s online merchandising solutions offer various degrees of personalisation. When used in e-commerce sites, AI and machine-learning technologies are capable of delivering an enhanced experience that drives towards a retailer’s overall business goal.

Systems handle visitors in different ways. The experience depends on the features available in the particular system, and also on the level of previous interactions by visitors.

Anonymous visitor

When shoppers visit the website for the first time, they are not previously known to that site – there is no previous browsing history nor identity. Even so, without knowing anything about the visitor, a clever solution can produce relevant recommendations. These come from aggregated data, and are based on what is called crowd wisdom, where the behaviour of the majority is used to inspire a person to buy the most relevant products.

Depending on how powerful the machine learning algorithms are, the system can determine the behaviour pattern of the average user in real-time, and offer recommendations from the best-seller list or others corresponding to an aggregated relevance.


Returning visitor

The next level of personalisation is the returning visitor. From this point, we can speak about individual relevance. Without revealing their identity, this person has left ‘behaviour traces’ in the system. A cookie has been saved on this visitor’s platform – be it a desktop, tablet or mobile device – which can connect their visits. This helps the system build an understanding of this visitor. Recommendations can then become more adapted to their enhanced use of the site.

The secret to success here is the combination of aggregated and individual relevance that makes the displayed assortment relevant based on crowd and individual data, for example in displaying appropriate category listings.

returning visitor


Logged-in user

Logged-in user offer the system more personal data, provided they give their consent. This can include location information, age, preferences, and other demographics. Purchasing history data is usually available. Visits across different platforms can also be connected.

This richer data offers more personalisation opportunities (for example by offering recommendations with higher probability of conversion for the person’s likes), as well as adapting recommendations for specific preferences or demographics.

An additional step here can be cross-channel integration. A shirt bought in a physical store, paid with a card is added to the personal account. Next day, if the person visits the online store and logs in, the store ‘knows’ more about the person’s preferences and can suggest products that normally resonate well with the already purchased shirt.


Contact us to learn more.