Adobe Marketo Engage disables archived campaigns and flags a September API limit

What changed

Adobe's Marketo Engage release notes, last updated 8 September, say that archiving a folder now disables and deschedules all campaigns in that folder tree. The same notes announce that from 30 September the Get Lead Activities and Get Lead Changes endpoints will fail with error 1003 when a listId refers to a static list containing 10,000 or more leads.

Sources: Adobe Marketo Engage release notes

Folder archive becomes an execution control

Archiving was previously easy to treat as an organizational action. The released behavior makes it a control over scheduled execution as well. Adobe says campaigns in the archived folder and its subfolders are disabled and descheduled. Operators should update runbooks so a cleanup task is not mistaken for a harmless navigation change.

Before archiving, inventory active smart campaigns, batch campaigns, recurring schedules, nested folders and downstream dependencies. Name the owner who approves the stop. Preserve the reason and affected workflow identifiers so the team can distinguish intentional disablement from an incident.

Test residual work outside the folder

A disabled campaign does not necessarily reverse completed profile writes, delivered messages or actions already handed to another system. Connected webhooks, CRM workflows and external queues may have their own terminal state. The release notes describe Marketo campaign behavior, not a universal rollback across the revenue stack.

Use non-customer test records to start actions at different stages, archive the folder and reconcile Marketo with every destination. Record queued, completed, cancelled and unknown work. Resume only after the exception queue is understood.

Large static-list API jobs need a migration path

Adobe says the new 10,000-member threshold applies when listId is used with Get Lead Activities or Get Lead Changes. Requests at or above the threshold will return error 1003 after 30 September. An integration can therefore remain syntactically unchanged yet stop retrieving expected records for a growing list.

Inventory jobs using those endpoints and identify whether they pass listId. Measure current static-list membership from the authoritative environment and include lists close to the threshold. Do not infer safety from last month's size; membership can change before the deadline.

Choose the replacement query deliberately

The appropriate alternative depends on the business population and the supported API contract. It may involve a smaller bounded list, a different filter, partitioning that preserves complete coverage or a revised extraction design. Avoid arbitrary splitting that duplicates leads or changes the observation window.

Run the current and replacement methods for the same closed period. Compare distinct lead IDs, activity IDs, timestamps, types and late-arriving changes. Keep rejected requests visible and alert on incomplete pagination or coverage rather than reporting a smaller result as success.

Separate released and forthcoming items

The current notes also describe a refreshed interface rollout and the renaming of Marketo AI to Coworker for Marketo Engage. Script Builder is listed as coming soon, so it should not be represented as generally released on the basis of this page. Operators should preserve those status distinctions in internal enablement material.

Adobe also notes removal of access_token as a query parameter after 31 August. Integrations should already use the supported authorization method, but administrators can still search logs and code for the deprecated pattern and verify that no overlooked job depends on it.

The operator checklist before 30 September

First, update archive permissions and require an impact review for campaign folders. Second, run a shutdown drill that includes nested campaigns and external queues. Third, locate lead-activity and lead-change integrations using large static lists and validate a replacement against a known period.

Assign separate owners for the campaign-control test and the API migration. The first owner verifies scheduled execution and downstream queues; the second verifies extraction completeness, pagination, errors and destination freshness. Keep their evidence in one change record so a campaign operator can see whether a missing audience is caused by intentional archive behavior, an API threshold or an unrelated data problem.

Test the deadline scenario before the production cutoff by using a safe representative list or a controlled request that can demonstrate the expected error without exposing customer data. Alert on both explicit failures and silent count changes. A job that returns HTTP success with an incomplete downstream dataset can be more difficult to detect than one that stops with error 1003.

Finally, monitor error 1003, record counts and freshness after the change. The release gives teams concrete behaviors and a deadline. It does not justify claims about campaign performance or automation efficiency without evidence from the organization's own workflow.