While there remains a land grab for OTT subscribers, the focus is beginning to shift towards improvements in quality of experience or ‘QoE’. Failure to provide a high QoE, writes Spicy Mango VP of Sales and Marketing John Griffiths, can shift a business from success to failure but the service provider is today chasing fractional improvements not massive shifts.
Big data, data analytics, analytics with artificial intelligence and a plethora of other terms abound not just in OTT but across the IT domain as a whole. But what does it all mean? You can ask 100 people and you’ll probably get dozens of different interpretations. The reason being that each person will see what they need to see and to some extent a single definition is probably unnecessary because of these differences.
Bringing the discussion back to the video and OTT world though, we have seen over recent years a significant growth in tools and platforms promising to analyse the mass of data that an OTT service will generate and provide QoE insights to help the service provider run their business. The vast majority of these tools are client based and sit inside a device or application and summarise the data into a dashboard. This dashboard provides the service provider with a snapshot of device performance.
- But how do they work?
- How as a service provider do you know that what you see on the dashboard is an accurate reflection of your network?
- Even if the dashboard is accurate, do you know where the real issues are and how to fix them?
To address the first bullet is simple. In the main they make assumptions based on pre-configured parameters e.g. if A happens and you see error code XYZ then this is the likely outcome. In many cases this will be correct but still assumes many other elements are configured and behaving correctly.
This does not imply a fault with the tool and indeed they provide invaluable help in identifying potential issues.
The second and third bullets can only be addressed by validating the data sets and sources. It requires a deep dive in to the data to recreate entire sessions through the complete workflow. By gathering all the raw data logs and using a unique identifier such as an IP address or unique session ID (complying with GDPR of course), the end to end customer journey can be mapped.
This process will identify any issues not being picked up by the monitoring tools, (they are only usually configured to look for certain fault conditions), it will validate that the faults being reported are correct, for example is an exit before video start really and exit before video start and it will also enable fault finding. If an error is occurring, where is that error occurring? Is it a timeout in the app or all the reported errors coming from devices that are not officially supported?
Today a service provider is rarely chasing massive operational improvements. The improvements are incremental and a percentage point here and there can have a significant impact on the business performance. If improving the quality of experience reduces churn by just 1% then there will be a boost the bottom line through reduced cost of acquisition and increased revenues from retained subscribers. Not only that but staff can focus on developing new features and services as opposed to feeling like they are constantly firefighting.
In summary, the monitoring platforms are essential to the business operation. However, they are only as good as the data they receive and they do not diagnose the exact nature of the issues. Any OTT service provider should consider how they validate their data to guarantee trust in their dashboard and how they can use that end to end picture to diagnose and fix any issues that are damaging the subscribers quality of experience.
Spicy Mango is an expert media technology consulting and software delivery group to broadcasters, content owners and service providers worldwide and are Silver Sponsors of TV Connect.