Missing insert_id or session_id: confirm the problem

insert_id and session_id protect event counts and session metrics. insert_id lets Amplitude discard duplicate copies of a retried event when the copies have the same device_id. This prevents network retries from inflating event counts. session_id groups a user's events into a session for Session Analysis. Missing values can cause incorrect counts and session metrics.

Data Assistant doesn't check these fields. Requirements vary by source. Amplitude SDKs manage session_id automatically for supported client platforms. Events sent through the HTTP API need an explicit session_id to join session metrics. insert_id matters most for retried server-side events. Audit representative events to confirm both fields.

Run this audit if event counts seem inflated or Session Analysis results seem incorrect. Also run the audit when you add a server or HTTP API source. You need access to User Profiles.

Diagram showing Amplitude storing one copy when a retry has the same device_id and insert_id, versus dropping a second real event when the same device reuses those identifiers within seven days

Audit representative events

List each client SDK, server, and customer data platform source in the project.
Choose at least one active test event from each source.
Trigger the test event and open the user's profile in Users > User Profiles.
Select the event, choose Raw, and record its source library, insert_id, and session_id.
For server events that your system retries, confirm that each logical event has a non-empty insert_id. Retries of the same event must reuse the insert_id value. Different events must use different insert_id values.
For client SDK events that belong to an in-product session, confirm that session_id contains a positive session start time. The session_id value must remain the same throughout one session.
For HTTP API events that should join Session Analysis, confirm that session_id isn't missing or -1.
Confirm that session_id: -1 is intentional for out-of-session events, such as some push-notification events.

A sample doesn't measure project-wide coverage

User Profiles and Event Explorer support targeted QA. To calculate missing-field rates across every source, ask your data team to query an authorized export or warehouse copy of the raw events.

How Amplitude handles missing identifiers

Amplitude can't populate insert_id or session_id after ingestion. You must correct the SDK, server, or CDP mapping. For sessions, Amplitude provides one analysis-layer option. An Admin or Manager can define a custom session from another property, from start and end events, or from a timeout. Custom sessions can support Session Analysis without changing every payload. However, custom sessions don't add missing raw session_id values. Custom session definitions also affect session-based charts across the whole project.

Send requirements to the tracking owner

Correcting these fields is an instrumentation change. Send the affected sources to the team that owns your tracking. Copy this message and fill in the brackets:

text

Some of our events are missing event identifiers. On source [SOURCE], event
[EVENT] arrives with [insert_id missing/duplicated / session_id missing or -1].
For retried server events, send a unique insert_id and reuse it on retries.
For HTTP API events that should join sessions, send session_id as the
session start time in epoch milliseconds. This is [inflating our counts /
breaking Session Analysis].

Get help without engineering resources

You can't add these fields to incoming events yourself. Contact these resources in order:

Your Amplitude account team or Customer Success Manager.
A partner or agency that manages your website or app analytics.
Amplitude Support at gethelp.amplitude.com for product questions.

To correct the affected sources, continue with the engineering solution.

Was this helpful?