To find out what your visitor wants, it is useful to compare them to existing customers. The more similar two visitors are, the more likely they are to be interested in the same products.

When visitors first start coming to your website you will generally have very little information to work with. Fortunately we have some data that we keep track of capable of helping you get started in identifying the visitor. This data includes:

  • Device type
  • Operation System
  • Screen resolution
  • IP-Adress
  • Location data
  • Weather information
  • Visit time
  • Visit duration
  • URL of origin (did they enter from Google with a keyword or directly? Did they click on a facebook ad?)
  • Estimated buying phase
  • Estimated persuasion type

This data serves another purpose as a fingerprinting tool. Because a lot of browsers start deleting cookies automatically, this data can identify when it is the same visitor. This is useful because it allows you to still target this visitor with recommendations based on it's known information. Next to that it allows you to not count this single visitor multiple times. This makes sure your visitor statistics are more accurate.

After a while (usually 1-2 months) our algorithm will have gathered a good amount of profile data and is now able to use more metrics to compare visitors. These include

  • What item did they look at first?
  • What items are currently in their shopping cart?
  • What items are often bought together?

Because the algorithm now has insights into the behavioral pattern of your visitors, it can use these statistics to predict their next behavior by matching them to previous customer journeys.

As more information about the visitor becomes known we can create a more detailed profile and thus better recommend products to him or her.

Lookalike targeting is a big part of our self-learning algorithm and will improve the more profiles have been added to your database.

Did this answer your question?