We’ve seen how using a “good enough” solution for personalization, search and recommendations (S&R) is limiting and potentially damaging to the viewer’s user experience. And we’ve discussed how solutions that allow for continuous change and optimization lead to more robust engagement and lower churn rates. Now let’s take a deeper look into the different types of changes being made and how they reflect the various on-going needs of the business, as well as the limits of relying solely on machine learning (ML) and artificial intelligence (AI) models.
The limits of machine learning
When it comes to improving personalization, search and recommendations, there is a belief that ML and AI models are fully capable of handling all use cases in a way that requires no human intervention. Our experience shows that this isn’t the case.
Companies have business objectives that are often either at-odds with model results or need to work in conjunction with those model results. For example, think about featured content that’s part of a licensing deal. Or consider tent-pole events like the Olympics and other high-profile sports and entertainment events. These types of programming can drive the need for adjustments to be made to what is shown in carousels on screen, often in real-time, and by teams who are on top of the latest content.
Going beyond the content
Additionally, we need to acknowledge that no matter how good a UI or UX design is today, they should never be seen as perfect or complete. That’s because the viewers’ expectations of what a good UI looks like are constantly evolving, too, as is our understanding of how they’re using it. Thus, good enough’ solutions don’t fit into this dynamic because an approach where the UI does not change will not work.
Charting the changes
TiVo’s customers are always in evolution mode. Whether those changes are made weekly or daily, we have the data that not only reinforces this approach but also identifies the many ways customers are changing. Changes range from smaller updates like adding new use cases, to larger efforts to update landing page carousels – because viewers don’t like seeing “comedy for you” over and over and over.
It should also be noted that many of the changes called out in the chart above are not even possible in the ‘good enough’ solutions we’ve discussed. This includes the AB testing, optimization and iteration of UI and UX essential to any usual process in order to constantly optimize and engage. This puts companies in the position where they must rely on a ‘hope strategy’ instead — simply hoping that doing a one-time implementation of personalization/S&R will yield positive results
From identifying the problems with one-size-fits-all personalization, search and recommendation solutions, to addressing those issues through implementation, optimization and consistent change, always remember that this is all part of a long journey.
Ready to take the next step? Discover how TiVo can guide you every step of the way—turning strategy into sustained success.
Don’t Settle for ‘Good Enough’: Schedule a Demo Now.
Explore how TiVo’s personalization solutions drive engagement and reduce churn.
Comments are closed.