OpenTV Platform supports business intelligence and analytics so that the operator can monitor and analyse their service and user behaviour. It allows them to:
-
Gather consumer usage and playback quality data from all devices
-
Monitor playback quality of experience in Bitmovin Observability
-
Visualise data in the included Tableau dashboards
-
Measure the performance of rails, pages, and A/B tests in Operator Console
-
Access the underlying data warehouse for deep-dive analysis
-
Receive data exports and data lake tables into their own tools (for example Microsoft Fabric / Power BI)
-
Act on the data with the wider OpenTV products to promote content, resolve delivery problems, and optimise commercial offers
Quality of Experience
OpenTV uses Bitmovin Observability for playback Quality of Experience (QoE) on clients that use the Bitmovin Player, including OpenTV ION. It is enabled by default for ION.
Bitmovin Observability provides:
-
Real-time monitoring of plays, concurrent views, startup time, buffering, bitrate, and errors
-
Drill-down to individual sessions and errors, with filters by device, platform, country, ISP, CDN, content, and app version
-
Alerting on quality thresholds (for example startup time, buffering ratio, error rate)
-
Basic and Premium options, with 15-day or 13-month retention
Bitmovin QoE data is also loaded into the OpenTV data warehouse and data lake, so that it can be reported on next to TV Analytics data (see Bitmovin QoE data below).
Customers on older clients that report playback metrics through the OpenTV Analytics agent keep the Quality of Experience dashboards in Tableau. These dashboards cover session quality, buffering, bitrate changes, frame drops, errors, and time to start.
TV Analytics
The TV Analytics module allows operators to measure the performance of their service with consumers.
The data analysed covers:
-
Content
-
Time viewed
-
Live / Video On Demand / Catch-up / DVR
-
Series / episode
-
Genre
-
Channel
-
Video format / quality
-
Completion
-
-
Consumers
-
Sessions
-
Devices
-
Daily / hourly usage
-
Max concurrent
-
Signed-in and guest (anonymous) users
-
-
Devices
-
Form factor (mobile, tablet, big screen, desktop)
-
Device model
-
Technical specifications
-
Application
-
Platform
-
-
Connection
-
Connection type
-
Wi-Fi signal and standard, CDN download speed
-
Broadcast tuner signal strength, quality, signal-to-noise ratio, and lock (hybrid devices)
-
TV Analytics data is available for reporting within 1 hour of occurring on the device.
Rails & A/B test analytics
Operator Console includes screens that show how rails and pages perform:
-
Views, clicks, plays, completions, purchases, and engagement time, from each rail and globally
-
Comparison of A and B groups in an A/B test
-
Drill-down by device, form factor, segment, entitlement, country, app version, and time
-
Comparison against earlier periods
This lets editorial teams test a rail layout on part of a segment and decide, based on data, whether to roll it out. See Editorial curation and Rails Builder.
Video Platform Operational
The Video Platform Operational module allows operators to monitor the operational behaviour of OpenTV Platform.
The data analysed covers:
-
Content
-
Metadata and images, including content with missing metadata or images, linked to product availability
-
Ingest and workflow activity
-
-
Accounts
-
Activity
-
Devices
-
Products
-
Entitlements and rights
-
Recordings and recording quota usage
-
GDPR subject access requests
Operational data is available for reporting within 1 hour.
Email notifications are sent for metadata ingest, metadata export, and asset workflow events (started, completed, failed). These events are also stored in the data warehouse.
Data Collection
Data is collected for OpenTV Analytics from:
-
Client applications, via:
-
The Bitmovin Player analytics collector (playback QoE)
-
The OpenTV Analytics Agent, integrated by the application vendor, which supports iOS, Android, and HTML/JavaScript
-
-
The OpenTV Analytics Collector API
-
The OpenTV Platform Activity API
-
OpenTV Platform itself (for example metadata, workflow, accounts, devices, entitlements)
-
Bitmovin Observability (QoE data, loaded periodically)
Redshift Data Warehouse
Collected data is processed into the OpenTV Analytics Redshift data warehouse.
The warehouse holds the collected data in a data model. This supports:
-
Event-level reporting
-
Aggregated views for fast reporting, built as fact and dimension tables that work well with BI tools such as Power BI and Tableau
The warehouse can be accessed by Tableau, or by SQL from other reporting / analytics tools.
Tableau
The dashboards are delivered as a series of Tableau Cloud dashboards with pre-defined visualisations of the data. Operators can also build custom reports in Tableau from the published data sources.
Tableau is an off-the-shelf product with a wide range of dashboarding and visualisation features, including:
-
Dashboarding
-
Alerting
-
Metrics
Data exports and data lake
For use in their own data platforms, operators can receive OpenTV data in open formats in Amazon S3. Exports are set up on agreement with NAGRAVISION, through a service request.
Scheduled data exports
Selected tables are exported every 24 hours to Amazon S3, in CSV or Parquet format. Available exports include:
-
Devices
-
Playback events and metrics
-
Errors
-
Live, replay, and VOD viewing sessions
-
Channels
-
Programmes
Iceberg data lake
The main data warehouse tables are published as Apache Iceberg tables (Parquet files) in Amazon S3, refreshed daily. They can be read directly by modern data platforms without copying the data through SQL.
The tables cover:
-
Facts: content consumption, playback sessions and metrics, activities, purchases, subscriptions, and daily account, device, and error summaries
-
Dimensions: accounts, user profiles, devices, channels, EPG events, and editorial content (with localised attributes)
A step-by-step guide is available for connecting the data lake to Microsoft Fabric, which makes the data available in Power BI. Other tools that read Iceberg or Parquet from S3 (for example Amazon Athena, Spark, Google BigQuery) can use the same route.
Bitmovin QoE data
Bitmovin playback QoE data is loaded periodically into the data lake and the data warehouse:
-
Full session-level data is kept for 15 days
-
Aggregated QoE views are kept for 12 months
This means QoE can be joined with account, device, and content data in a single query or report.
Access logs
Operators can also receive the HTTP access logs for their OpenTV Platform API, operator, and admin endpoints. The logs are replicated into an S3 bucket in the operator's own AWS account, for audit, troubleshooting, and security monitoring (SIEM).
Fair Use Policy
DWH SQL access is governed by your NAGRAVISION SLA, which includes terms for using the services as intended.
SQL access is designed for aggregated reporting on the TV service, or detailed reporting on specific consumers / devices / content. It is not designed for bulk replication of the data. For bulk data access, use the data exports or the Iceberg data lake described above. Additional Redshift nodes can also be provisioned (at extra cost).
Things that are 'fair':
-
Aggregated queries on aggregation tables across the full time period
-
Aggregated queries on raw data with 'fair' time periods / groupings
-
Running the Tableau dashboards, or creating similar dashboards in other tools
-
Fetching data for a single account
Examples of things we know break fair use include:
-
Dumping whole DWH tables via SQL (use the data lake instead)
-
Non-aggregated queries across large datasets (for example
select * from activity) -
Aggregated queries across 'raw' data over long time frames (for example months / years)
-
Inefficient SQL that has a large cost for the result and can easily be optimised
To help manage fair use, the Redshift database uses a Work Load Manager that prevents some queries from running. Its rules implement the guidance above, and include:
-
A maximum of two million rows returned per query
-
A maximum execution time of 6 minutes per query