The model is being tested in a pilot project with the Finnish shipping company Tallink Silja Oy. It is based on AI and machine learning techniques, with the Annalect data and technology platform, and consumer behaviour and interests understanding.

Jussi Piri, the CEO of Annalect Finland, said: "We started to develop the model in the spring of 2016. We wanted to increase the value of web analytics by creating a user-friendly model that is better than the currently available analytical methods. We wanted our model to use all user data collected by a website, and we also wanted it to be compatible with data from external sources. The key idea was to make advanced AI and machine learning methods available to all analysts and companies on a large scale."

The company said the model can forecast a visitor’s purchase behaviour in milliseconds when the page loads. It can cluster different visitors into different groups based on similarities in behavioural data.

Piri added: "We can use the findings produced by the model to better forecast and test factors such as media mixes, targeted advertisements, contents and website features."