Using Video Metrics to Drive Product Evolution: Content & Business Metrics
This is the first of three posts in our Video Metrics Series.
For media and content providers, video playback metrics are integral to improving products, retaining and growing user base, and opening up new opportunities for monetization. As is true of analytics across the board, data collection is iterative, allowing you to learn and refine as more data becomes available. Tracking the right metrics leads to aggregate intelligence which can be used to inform decisions, and as data collection techniques become more granular can be used to segment users and provide personalized experiences.
The video playback metrics that are important for content providers to track can be divided into three distinct but interrelated categories:
- Content/Business Objective Metrics
- QoE (Quality of Experience) Metrics
- QoS (Quality of Service) Metrics
This blog series will map out and define these three categories of metrics and dive into why they are important for content providers to measure.
Part 1: Content & Business Metrics
This category of metrics is related to the performance of content with audiences. They are typically high-level, session related metrics that allow businesses to gauge content being offered and how it relates to and affects business objectives: monetization strategies, content syndication, etc.
Number of Playbacks
The number of playbacks is a simple measure of how many times a particular video is initiated, but it can provide valuable insight. It allows you to track what content people are playing the most and what is proving less popular. This has utility beyond learning what your users like and dislike. It can also help you optimize your spend if you’re paying for syndicated content. If certain content is underperforming, you can drop it and focus on purchasing other content that is resonating with users.
Percentage Watched (Abandon Rate)
While number of playbacks is important, percentage watched is a more informative metric. It allows you to understand how much of a video users are watching. This can tell you a number of things. If a lot of users are dropping off quickly, it could be an indication that the content is unappealing to your audience, which can inform decisions about what content to continue buying or producing. It may also indicate a Quality of Service issue – for example high abandon rates may correlate with high video-stalling rates or some other issue at the network/transmission level.
Perhaps more importantly, it can inform product and content decisions. For example, if you can determine common drop-off percentage near the end of a video playback, you can optimize the timing of your end cards. Should you initiate the end card during the credits, or just before the credits begin to prompt the next piece of content? This is important insight, allowing you to optimize end card timing and increase the likelihood of users consuming additional content.
Percentage watched is also critical to effective ad queues. If you are monetizing via advertising, understanding how much of a program users are watching can help inform pricing tiers. For example, if 90% of your users will watch the first 5 minutes, advertisements within this time frame should be sold at a higher premium than ads that are offered halfway through the program – a point fewer users will reach. This intelligence indicates popularity of content as well as potential ad audience, which strengthens your ability to sell ad space to advertisers.
Location is a metric that helps content providers determine the logic behind serving local specific content. Using the metrics above and combining them with locational data can provide insight into what content – beyond that content that is only relevant to certain areas, like local news – is performing well in which geographical areas, and optimize your content syndication and content inventory for each particular area.
The above content and business metrics are important to measure because they can help inform the popularity of content with users, indicate what is resonating and what is not, and inform future monetization and syndication strategies.