Viral Product Prediction

Business Opportunity

Exclusive license limited to a specific scope or field, such as context, market, territory, and time.

 

Tech description

Funded by NFR, via BigInsight and with in-kinds by UiO and Telenor the researchers have developed a new method to predict the individual and collective behavior of users/customers in for example the buying of a product or the churn of the customer. The method is based on a new statistical model, it is implemented efficiently and the algorithm has been tested on a set of data from Telenor with a positive result. The fundamental idea is that the behavior of a customer is definitely influenced by marketing campaigns, but crucially also by the observed behavior of other individuals in a social relation to each other. In this case, the process leading to the behavior of the customers can be compared to the spread of infectious diseases. The model is inspired by this similarity, and is able to detect and quantify the viral strength of a product, service or behavior.

 

Advantages

A first proof-of-concept test on a real Telenor product is finalized. Telenor`s product is an app service and based on a training period of 82 days the model were able to predict the adoption 2.5 years into the future. See Figure (blue band, with uncertainty).

Eirik Løvbakken M.Sc.

Eirik Løvbakken M.Sc.

Technology Strategy Manager

Innovation

+47 92 43 82 14

eirik.lovbakken@inven2.com