CTV ad stitching: distinguish an inserted ad from a measured impression

An ad is inserted into a television stream. That confirms an insertion step; it does not automatically establish that someone watched the entire ad.

Server-side ad insertion, often called SSAI, places advertising into a stream through server-side processing. Measurement then needs to establish what happened after that decision and delivery work.

The Media Rating Council’s SSAI and OTT measurement guidance addresses the evidence needed in these environments. Ask partners how their implementation supports the reported metrics rather than accepting “server-side” as a complete measurement description.

Map the evidence chain

Separate the ad request, decision, insertion, media delivery, playback start, progress, and completion. Not every implementation exposes each stage to the buyer.

For each reported event, identify who generates it and what observation triggers it. A server delivering media segments and a player reporting playback provide different evidence.

A measurement partner should be able to explain the method, coverage, and limitations in terms the operator can use. An unavailable measurement should remain unavailable rather than being inferred from an earlier step.

Ask what happens when viewing stops

A viewer may close the app, lose connectivity, change channels, or seek away where the experience permits it. Determine how those cases affect impression and completion reporting.

Use a controlled test in a representative app and device environment. Record the interruption time and inspect the corresponding events with the publisher’s help.

Do not expect one device test to validate every app integration. Keep a coverage table showing which combinations were tested and which depend on partner evidence.

Inspect retries and duplicate signals

Distributed delivery systems can retry requests. Establish whether repeated signals are deduplicated and which identifiers support that process.

A report showing more completion events than eligible starts is a reason to inspect definitions, time boundaries, and duplicate handling. It is not immediately proof of inflated delivery.

Also confirm whether reports count events, impressions, sessions, devices, or modeled viewers. These units can differ legitimately, but they should not be mixed without explanation.

Keep creative compatibility in the same discussion

A correct insertion decision can still lead to a poor experience if the rendition, duration, audio, or transition is unsuitable for the stream.

Review the supported creative specifications and test the ad break as experienced on the device. The VAST troubleshooting guide helps with response and media failures, while SSAI review adds the stream and measurement path.

Ask whether verification is supported in the purchased environment and how unsupported cases are reported.

Agree on the reporting statement before launch

A useful measurement sheet names the billable event, source of playback evidence, supported devices, filtering, deduplication, and reporting delay.

Then compare the final delivery against those definitions. The value of SSAI reporting comes from a clear connection between the delivered stream and the event claimed, with limitations visible wherever that connection cannot be established.