Connect Data Feeds

 

Conviva’s Connect Data Feeds provide an easy route to ingest Conviva data into your data warehouse for further analysis. Data Feeds unlock many new use cases, for instance, powering an internal dashboard with audience data, drilling into churn factors, or forecasting system load. Data Feeds are flexible and can be configured as follows:

  • Contain the metrics and metadata you need

  • Sent to your Cloud location

  • Packaged in either Avro or Parquet

Conviva offers a range of Data Feed types across VSI and DPI to suit your use cases:

DPI VSI

Flow Feed: Flow level metrics and metadata with session data.

Content Feed: Stream level metrics and metadata with custom tags.

Event Feed: Event level metrics and metadata.

Ad Feed: Ad level metrics and metadata.
  Offline Feed: Engagement data for content downloads.

DPI Feeds

DPI Feeds provide a regular file delivery of data to a chosen location, ideal to gain insight into user behavior, experience quality, and actionable diagnostics for business, technology, and operational teams. Feeds can be set to minute or hourly frequencies, in Parquet or Avro format. DPI Feeds for Flow and Event data enable application outcome performance analysis at multiple layers of granularity:

Data Feed Contents Use Cases
Flow

Flow level metrics, metadata and session data.

Recommendation Engines
Churn Analysis
Pricing
User Cohorting
Trending and Forecasting
Fraud detection

User Journey Analysis
Event

Event level metrics and metadata.

Full Manual Mode
Feature Analysis
Advanced Behavior Analytics

Flow Feed

A Flow represents the metrics and metadata of a section of user journey. For instance, a Login Flow’s metadata could describe the method the user has attempted to login (OTP, SSO, and others), the Flow metrics illustrate the time it took the user and whether they were successful or not.

To learn more about Flow Feed, click here.

Reconstruct a user’s journey through the app as it unfolds. Each time a Flow completes or times out, the system sends it to the Feed, enabling precise tracking of how the user navigates and the experience being consumed. When the engagement ends and the session terminates, the system adds a consolidated summary of the entire session to the Feed.

For use cases that require near real-time insights, such as powering a recommendation engine, teams should ingest individual Flow data. For scenarios that are less time-sensitive and do not require fine-grained detail, teams can ingest only the Session data.

Each line of the Flow Feed describes one completed Flow. Completed Flows include Flows that have timed out as well as ones reaching their final state. Flow data includes:

  • High level Flow metadata

  • Details of the end user who created the Flow

  • Flow metrics and metadata

Since a session is a special type of Flow, sessions are also included in the Flow Feed, summarizing the user’s engagement with your app.

Use Flow Feed to extract Flow Name/Session ID-wise Metrics and Metadata, including associated device, app, location, and network information at the frequency of hour/minute.

  • Out-of-the-box metrics aggregated at the Session level

  • Custom-defined Flows and custom metrics

Each line of the Flow Feed record contains the information:

Flow Details Location Device App Network Metrics Metadata

  • Flow Name (Custom Flow or Session)

  • Sensor Version

  • Viewer ID

  • Start Time

  • End Time

  • Session ID (when relevant)

  • City

  • State

  • Country

  • Zip Code

  • Device ID

  • Platform

  • Platform Subcategory

  • Device Category

  • Manufacturer

  • Device Name

  • Device Model

  • Marketing Name

  • OS Family

  • OS Name

  • OS Version

  • Browser Name

  • Browser Version

  • App Name

  • App Version

  • App Build

  • App Type

  • ASN

  • ISP

  • All out-of-the-box metrics aggregated at the Session level.

  • All custom-defined Flows and custom metrics aggregated at the Flow level.

  • Metrics added to either MetricsInt or MetricsFloat, depending on type.

All mapped tags

 

 

Event Feed

The Event Feed offers the lowest level of granularity for DPI data. Each line in the Feed represents an app-level event, such as a user activity (for example, a link or button press), a system event (for example, an API call), or a network-initiated event. For an event to appear in the Feed, it must first be mapped in Conviva and marked for export on the Event Feed, ensuring that only the required information is exported. This is important because the Event Feed is extremely large.

To reduce the size of the Event Feed, metadata such as device information is available only through the joining session of the Flow Feed. The Event schema includes only metadata that can change within a session. For example, user location is included because it may vary while the app is being used, making it mutable within a session.

Use Event Feed to extract Session ID-wise event data (event name, category, and details), location, device, and network information at an hourly frequency.

Each line of the Event Feed record contains the information:

Session Details User Location Network Event

  • Session ID

  • Event Time

  • Device ID

  • City

  • State

  • Country

  • Zip Code

  • IP Address

  • Connection Type

  • Event Category

  • Event Name

  • Event Details

 

 

VSI Feeds

VSI Feeds for video streaming sessions include metadata, such as viewerID, and metrics (aggregated from event logs), such as Playing Time. The session time span is from the moment the client starts the session until the client terminates the session. The extracted data represents a snapshot of each session at the end of the day, thereby helping in the session’s performance analysis at a granular level.

Session Summary

Retrieves historical and real-time metrics for building custom dashboards or applications. It provides a detailed, data-driven overview of a single user's viewing experience, showing metrics like plays, time watched, buffering, errors, and quality (SPI), allowing deep dives into specific events, technical issues, and user actions (like restarts) to pinpoint problems and improve engagement, often visualized in a timeline format for granular analysis.

Each line of the Session Summary record contains the information:

Session Details Location Device Network Time Status VSF/VPF/Errors

  • Viewer ID

  • Session ID

  • Asset Name

  • CIR Related Exit

  • isLive

  • Interrupts

  • Average Bit Rate

  • SessionTags

  • StreamURL

  • Percentage Complete

  • Rejoined Count

  • Content Length

  • CIRR Interrupt Count

  • Ad Related Rebuffering

  • Rebuffering During Ads

  • Paused Ratio

  • Number of Bitrate Switches

  • Avg Average Bitrate

  • DMA

  • City

  • State

  • Country

  • Postal Code

  • Device OS

  • Device Hardware Type

  • Device Manufacture

  • Device Marketing Name

  • Device Name

  • Device OS Version

  • Device OS Family

  • Player Framework Name

  • Player Framework Version

  • Device Model

  • Device Vendor

  • ASN

  • ISP

  • IPv4

  • IPv6

  • IP Address

  • IP Type

  • CDN

  • Core CDN

  • Browser

  • Browser Version

  • Last CDN Edge Server

  • Last CDN Group ID

  • Household ID

  • Household IP

  • Connection Type

  • Decision Resource

  • Decision Resource ID

  • Decision Resource Resolved

  • Start Time Unix

  • Start Time Unix Ms

  • Startup Time

  • Playing Time

  • Re-Buffering Time

  • Connection Induced Rebuffering Time

  • Video Restart Time

  • End Time Unix

  • End Time Unix Ms

  • Pause Time

  • Micro Playing Time

  • Micro Playing Interruptions

  • Micro Buffering Time

  • Micro Buffering Interruptions

  • Long Rebuffering Time

  • Long Rebuffering Interruptions

  • Last Playhead Time

  • dt (Starting hour of the hourly interval range)

  • Ended Status

  • Session Ended Status

  • Exit During PreRoll

  • Startup Error

  • Error List

  • VPF

  • VPF Error List

  • VSF Business

  • VSF Business Error List

  • VSF Technical

  • VSF Technical Error List

  • VPF Business

  • VPF Business Error List

  • VPF Technical

  • VPF Technical Error List

 

Ad Summary

Retrieves viewer experience by combining statuses, metrics, and dimensional metadata. It refers to Conviva's advanced analytics and diagnostics for video advertising, providing real-time insights into ad quality (QoE), engagement, delivery, and performance through rich dashboards and data feeds, helping publishers optimize ad revenue and viewer experience by identifying issues like buffering, start failures, and abandonment, and correlating them with specific ad elements. It offers granular, session-level data, allowing deep dives into ad breaks (pre-roll, mid-roll, post-roll) and correlating user behavior with ad events.

Each line of the Ad Summary record contains the information:

Ad Details Device Location Network Time/Status Session/Content VSF/VPF/Errors

  • Ad Title

  • Ad Session ID

  • Ad Manager Name

  • Ad Manager Version

  • Ad Pod Type

  • Ad Stitcher

  • Advertiser Category

  • Advertiser ID

  • Ad Break ID

  • Ad Campaign Name

  • Ad Category

  • Ad Classification

  • Ad Creative ID

  • Ad Creative Name

  • Ad Day Part

  • Ad Description

  • First Ad ID

  • First Ad System

  • First Creative ID

  • Ad ID

  • Ad Is Slats

  • Ad Position

  • Ad Sequence

  • Ad System

  • Ad Technology

  • Ad Type

  • Ad Unit Name

  • Advertiser

  • Device ID

  • Device OS

  • City

  • State

  • Country

  • Continent

  • ASN

  • ISP

  • IPv4

  • IPv6

  • IP Address

  • IP Type

  • Browser

  • Browser Version

  • Household IP

  • Content Watched Time

  • Start Time

  • Start Time Ms

  • Playing Time

  • End Time

  • End Time Ms

  • Startup Time

  • Ended Status

  • Session Ended Status

  • Stream URL

  • Avg Bit Rate

  • Segment ID

  • Rebuffering Events

  • Rebuffering Time

  • Content Length

  • Content Asset Name

  • Content Session ID

  • VAST Media File API Framework

  • is Live

  • Session Tags

  • Viewer ID

  • Household ID

  • Error List

  • Fatal Error Codes

  • Video Playback Failure Errors Business

  • Video Playback Failure Errors Tech

  • Video Start Failure Errors Business

  • Video Start Failure Errors Tech

  • Startup Error

 

 

Conviva Data Feeds help engineering, analytics, and product teams extend insights beyond what dashboards or aggregated metrics typically provide and understand how users behave by showing which experience metrics influence engagement. Instead of relying on time-series data that only records event counts, Data Feed helps in measuring how long customers spend in each part of the user journey (for example, login, search, browsing, playback, or purchase).

Use the metrics and dimensions to identify patterns, link performance issues to business outcomes, and create custom metrics through Pulse UI. These detailed datasets support analysis across different geographies, devices, networks, and app versions, making it easier to understand how user experience impacts churn, engagement, and revenue.

Connect Data Feed Delivery Options

Configuration of Data Feeds includes:

  • Type (Session, Flow, or Event) of data and fields to export

  • Export frequency

  • Export file format (CSV, Avro, and Parquet)

  • Export destination (FTP/SFTP, Amazon Web Services (S3), or Google Cloud Services)

Common Use Cases for Connect Data Feeds

  • Understand Customer Sentiment and Improve Retention: Combine Conviva's Data Feeds with CRM data, contact center logs, or app-store ratings to identify behaviors linked to customer churn. By analyzing session, flow, and event-level data, isolate issues that contribute to dissatisfaction.

  • Optimize a Key Metric: Connect Data Feeds support detailed analysis of how various dimensions—such as device type, geography, or network—affect a selected metric. For example, examining purchase completion rates on mobile devices in a specific region. Identify the influential factors contributing to system performance and accordingly estimate the effect of improvements.

  • Identify Fraudulent or Abusive Activity: Use Conviva Feeds with CRM or identity data to detect abnormal or high-risk behavior. For example, concurrent logins from distant locations, or patterns indicating restreaming attempts. Investigate event-level details to respond proactively.

  • Support Personalization and Improve Journey Completion: Use Conviva Data Feeds with CRM-based segmentation to get meaningful customer cohorts based on behavior or preferences. Identify users who pause mid-journey and trigger targeted prompts to help them complete key actions.

Conviva VSI Connect Files

The Conviva Connect session summary files contain the daily and sub-daily delivery frequencies of offline historical logs of your Conviva data across the supported Conviva data models.

To access Connect file data definitions, login to Pulse or reach out to your Conviva representative for access.

Note: To include the Conviva Household ID in your content and ad summary data, reach out to your Conviva representative.

 

 

Get Started with Conviva Connect

  1. Enable Conviva Data Feed

    To enable Conviva DPI or VSI Feed, reach out to your Conviva Representative or contact Conviva Support.

    Conviva will configure your delivery frequency and destination options.

  2. Get the Conviva Data Feed

    3. The method of delivery is either to a data store, such as a Google Cloud Storage bucket, an Amazon S3 bucket, or SFTP file transfer.

    • The Conviva Data Feed output file format is CSV (comma-separated values). We use the standard described in RFC 4180.
    • The Conviva VSI Feed file name is in the format: ContentSessionLog.csv, AdSessionLog.csv, IdentitySession.csv

    For more details about the delivery destinations, refer to:

    For more details about managing the S3 Bucket and AWS Role with the KMS encryption for SSD upload, refer to:

    For more details about how to use Conviva Connect Content Summary reports, refer to:

    • Conviva Content Summary Report (Contact Conviva)

    For more details about how to derive the Streaming Performance Index (SPI) data based on Content Summary reports, refer to:

    • SPI Calculation based on Content Summary (Contact Conviva)

    For more details about how to generate Quality Impacted Assets (QIA) report, refer to:

    Conviva replaces the SSD files. To continue using the SSD file delivery, refer to:

    • Conviva Video SSD (Contact Conviva)

    • Ads SSD (Contact Conviva)

 

Connect Data Feed Connect Data Feed Connect Data Feed DPI Feed DPI Feed DPI Feed VSI Feed VSI Feed VSI Feed API API API Connect Connect SFTP S3 Bucket Google Cloud Storage GCS Bucket