Personalisation is currently a trendy topic and is recognised as one of the key means to improve engagement and reduce churn for video service providers.
For the last 10 years video service providers have been investing millions of dollars in “user experience”, but …which user experience? Well, lots of service providers have merely been trying to copy and paste the Netflix model or they take very general approaches applied across their customer base.
Humans are not all the same! Just as it has come to be accepted that targeted advertisement is indisputably effective within the internet advertising industry, nowadays personalisation should be a mandatory key aspect of the user experience for every video service.
The challenge all video service providers have now is how to deploy an effective personalization strategy while ensuring maximum impact on ROI for the service.
Until now most video service’s initial personalization activities have focused on content recommendation. Although I still believe that there are a lot of technology improvements to be made to address this challenge (I promise to write a future post specifically focusing on this topic!), I think there are plenty of other areas where personalization can make a real – perhaps even greater – impact (targeted advertisement, personalized navigation flows, personalization of payment methods, genre importance, on-the-go experience, etc).
I believe today’s state-of-the-art machine learning complemented with A/B testing will cause a shift in how personalization impacts video service providers.
Today’s customer clustering algorithm can automatically learn about how video users behave and then answer questions like:
- What content from my service are my customers watching ?
- How long is my service being used by my customers?
- To what extent do my customers enjoying the content they watch?
- Where is my content being watched?
- How often are people binge-watching?
- How is live-TV watched?”
Once you automatically cluster your audience based on questions like these (among others) then it’s time to personalize the user experience for each cluster.
The traditional approach has been for you to rely on your company HiPPO (highest paid person’s opinion) and let them define the user experience for each cluster. But, there is another (much better) option: Use A/B testing to trial real scenarios with your audience and let them speak about what they want. Your customers are talking to you all the time. It’s time to listen!
In summary a data drive approach using machine learning and A/B testing will let you make the most of personalization thereby optimizing the ROI impact for your video service.
Want to know more? Contact us here