Video Session Feed: Start Here

Orientation for data analysts building a pipeline on the Conviva Connect Video Session Feed: what it is, which feed to choose, the schema, metric SQL, and delivery setup.

The Video Session Feed, formerly known as Session Source Data (SSD), is a session-level export of your Conviva Video data. Each row describes one video play or attempted play, with the viewer, device, network, and geo metadata alongside the experience and engagement metrics for that session. It is built for analysts who want to land Conviva data in a warehouse and build their own reporting, diagnostics, and models on top of it.

The links below take you to the schema, the metric calculations, the delivery setup, and a worked end-to-end walkthrough, in the order you need them.

Conviva Connect vs legacy SSD

Conviva Connect is the current product: a new schema plus a more robust delivery pipeline, and all new customers onboard to the Connect schema. Legacy SSD is still maintained for existing legacy customers and its files are now delivered by the same Connect pipeline underneath, but it keeps the legacy schema, so the field names and layout differ between the two.

If you are onboarding today, use the Conviva Connect schema described on these pages. If you already receive legacy SSD files and want to compare or migrate, see SSD Schema Comparison: Legacy vs Connect for the field-by-field mapping.

Daily vs hourly at a glance

SSD is delivered on a daily or an hourly schedule. They are not interchangeable, and the hourly feed has a deduplication requirement you must handle before you aggregate.

Daily: lifetime metrics, one row per session with final totals for ended sessions, timezone-aligned to your customer day. This is the clean one-row-per-session feed.
Hourly: interval metrics for tracking ongoing sessions. Each file carries every session active in that hour with its totals up to that hour, so a multi-hour session repeats across files. Consumers doing cross-hour aggregation must deduplicate on ConvivaSessionID plus StartTimeUnixMs, keeping the latest row.

Read Choose your feed: Daily vs Hourly before you build anything. It covers the daily and hourly semantics and the hourly deduplication rule.

Next steps

1. Choose your feed Daily vs hourly semantics and the hourly deduplication rule. 2. Schema and column dictionary Field, type, mode, and description for every column in the Video Connect schema. 3. Metric calculations Content Summary: the SQL for core, QoE, and engagement metrics, plus the lifetime-versus-interval math. 4. SPI calculation Derive the Streaming Performance Index (SPI) from Content Summary data. 5. Delivery setup Configure delivery to Amazon S3, Google Cloud Storage, or SFTP, and manage AWS roles. 6. Build your first pipeline End to end: land the files, dedup the hourly feed, compute per-day metrics, and validate against Pulse.

Delivery destinations

Conviva Connect can deliver SSD files to your storage of choice. Set up whichever destination fits your stack:

Amazon S3 bucket (and managing the AWS role for KMS-encrypted uploads)
Google Cloud Storage bucket
SFTP
BigQuery