Automated Anomaly Detection For Rapid Resolution
You’ve got great content, so match that with great viewing experiences
Take the hard work out of improving experience!
Streaming businesses require real time anomaly detection and resolution to provide the best QoE
In video streaming, there is constantly new content being streamed that varies widely in genre, content type, target audience and video duration. Added to this, there are a host of other variables in the streaming workflow that can influence the quality of viewer experience. It is almost impossible for a business to identify issues in real time by relying on manual configuration of thresholds and alerts and the constant monitoring quality of experience dashboards in person.
Manual issue detection systems do not automatically adapt to your changing service
Rule based alerting of issues are useful to inform every time a predefined threshold is triggered, however they are not able to automatically adjust to the changes in seasonal patterns, user behavior, or any other changes in the key and contributing metrics.
Manual issue detection systems require analysis effort
Manual detection of issues requires knowledge, time and resources to know what to look for and what and how to act on them. Even then, this approach cannot scale to all the metrics that matter to your business. Anomalies can occur in seemingly insignificant metrics, or metrics that no one is tracking through manual alerts.
Using MediaMelon’s Anomaly Detection to accelerate issue detection and resolution
MediaMelon’s Anomaly Detection system takes the hard work out of identifying issues. It uses Machine Learning and continuously monitors key quality of experience, engagement and business metrics to trigger alerts in real-time whenever anomalies are detected in the data points. Customers are notified of anomaly alerts through the SmartSight user interface, as well as interactively over Slack, Email or Webhooks for workflow automation.
The above example shows details of a user session where the user experienced a high buffering ratio. MediaMelon’s anomaly detection automatically detected at what point in video playback the user experienced buffering, for how long and what caused the buffering issues.
Key Benefits of MediaMelon’s Anomaly Detection
Resolve issues faster
Analysis is a complex and time consuming activity that risks not noticing the issues as they occur and tends to be retrospective rather than real-time and immediate. Anomaly detection with point of failure analysis is a real-time activity that automatically identifies the root cause of issues and allows the team to focus more of their time on solving issues rather than investigating the source.
Retain and engage your viewers, don’t churn them
Reduce the time to resolution of key delivery issues with Machine Learning powered issue detection across all your metrics with point of failure analysis. Move beyond manual alerts that don’t handle the constant variation in viewing habits that today’s online streaming services have.
Monetize advertising and avoid wasted inventory
Ads are particularly problematic due to the complex ecosystem involved with many points of failure from creative to them appearing in the video player. MediaMelon’s anomaly detection solution analyses all the time, finding the needle in the haystack to ensure that issues are rapidly resolved avoiding unfilled impressions and poor monetization.