glossary Recurring + delayed revenue Integration & Orchestration Systems · Automation & AI Agent Systems

Terminal State Configuration

Direct answer. Terminal state configuration is the setting that governs what a record becomes when an automated sequence stops trying. A dunning sequence exhausts its retries; a follow-up cadence sends its last message; a reminder cycle reaches its final attempt. In each case the system writes a status, and that status is the only thing downstream systems can see. Because the setting is chosen during installation and is not surfaced in day-to-day reporting, the state these records end in can be unknown to the person who owns the outcome.

A bounded sequence has a last attempt

Dunning retries run out. Follow-up cadences send a final message. Reminder cycles reach a last attempt. Escalation chains reach the top. In each case the system writes a status, and that status is the only thing any downstream system can see.

Terminal state configuration is the part of the design that decides what that status is: which value is written, whether the record stays active, whether an event is emitted, and which other systems are told. It is distinct from the sequence itself — the schedule and the messages — which is the part that receives attention during setup.

Identical sequences, different operational positions

Two businesses can run the same cadence and end in completely different places. One leaves a closed record with no owner: the population is unqueryable, so no queue exists and no report can count it. The other leaves an open, queryable state that a person or an automation is explicitly assigned to. The sequences are identical; the terminal state is not.

This is why the setting is worth reading directly. It usually sits in an installation or configuration screen rather than a dashboard, it was chosen once, and the person who owns the outcome is often not the person who chose it.

What to check

  1. For each bounded sequence, find the configured behaviour after the final attempt.
  2. Record the exact status value written to the record.
  3. Determine whether the record remains active, queryable and segmentable in that state.
  4. Determine whether an event is emitted, and whether that event is on the platform's documented event surface.
  5. Identify the owner or queue that receives the handback — and note explicitly when there is none.

Where the end of a sequence emits an event, identity matters for the same reason it matters anywhere else: a producer must ensure `source` plus `id` is unique per distinct event so that a repeated delivery is recognisable as a duplicate rather than counted twice [cloudevents_spec]. Under at-least-once delivery the same terminal event can arrive more than once, so the handler must be idempotent for the outcome to match a single delivery [azure_resilient_event_design]. Where downstream handling repeats a write, whether that repeat is safe depends on the semantics of the operation [rfc9110_idempotent].

What checking the configuration establishes

It establishes It does not establish
Which status a sequence writes when it stops That a different status would perform better
Whether the ended population is queryable at all What that population is worth
Whether a handback owner or queue exists That the handback is being worked
Which downstream systems can observe the ending That any revenue or retention outcome followed

The available terminal states and their downstream behaviour are defined by the specific platform in use and differ between products. Naming the concept does not recommend a setting.

Where this sits

The integration and orchestration systems guide owns the surrounding architecture, since a terminal state is the point where orchestration hands a process back. A missing return route for that handback is a write-back gap. This entry sits in the Glossary. Locate the capability under the Integration & Orchestration Systems domain, and use Revenue Scan when the next question is what public surfaces show.

Source and evidence notes

Explore related entities

[CLAIM BOUNDARY] The available terminal states and their downstream behaviour are defined by the specific platform in use and differ between products. This entry defines the concept and does not recommend a setting, assert that any particular terminal state performs better, or describe any business's configuration.