With Videoland, RTL Netherlands is anticipating sudden growth in Video-on-Demand (VOD). With its thousands of films and series from home and abroad such as Zware Tulp, Divorce, Gooische Vrouwen, Wolf of Wall Street, The Blacklist and Broadchurch, Videoland has the Netherlands’ largest and most varied online offer.
Videoland either rents videos individually (‘transactional VoD’ – TVoD) and then Videoland Unlimited offers unlimited films and series for a fixed monthly charge (‘subscription VoD’ – SVoD). Cmotions was asked to provide the Consumer Intelligence team support in evaluating campaigns and a range of CRM and direct marketing issues.
One of the important questions is whether different groups of customers can be identified on the basis of behaviour. Using a cluster analysis on viewing and clicking behaviour we were able to identify a number of customer groups, including:
We profiled each segment with reference to various customer brands. Subsequently we ran an analysis that provided insights into customers’ cancellation behaviour.
We also ran an analysis into the viewing and clicking behaviour of consumers before cancelling their subscription. We found that a of new customers cancel their subscription straight away on the first day. From that point onwards, the percentage of cancellations remains equally distributed throughout the free month trial period.
It also emerged that when a customer is hooked right in the middle of an exciting American film or series during week three or four of the month trial, they are more inclined to extend their subscription.
We set up several direct marketing campaigns using the scores at cancellation. The scores for each individual are recalculated every night on the basis of a special ‘survival’ analysis.
Finally, Cmotions ran a sentiment analysis on the subtitling of all Videoland’s films and series. We can use text mining to derive what types of films – based on the spoken words – sound more or less alike. This information is useful for purchasing new content and for Videoland’s recommendation engines.