Product overview
Make represents an automation as a scenario of connected modules. A starting event supplies data, and later modules map fields into searches, transformations or actions in other services. The important unit of data is a bundle: one incoming email may contain an array of attachments, and splitting that array changes how many times downstream actions execute. Designing the data flow therefore matters as much as choosing the connected apps.
For example, an attachment-processing scenario can split a message's attachment array, upload the files separately, collect the returned file links and send a summary. A router can send data through filtered branches, while an aggregator can bring bundles back together within its configured source operation. These controls are useful when one input produces several outputs. Optional AI steps can extract or classify information, but their output still needs validation before it becomes a record, message or routing condition.
What Make offers
- 01Map fields between connected modules and inspect the bundles produced at each step instead of assuming every step receives one record.
- 02Use an Iterator to split an array and an aggregator to collect results; explicitly include fields needed after aggregation.
- 03Configure router filters, route order and a fallback path. Router branches execute sequentially, rather than simultaneously.
- 04Handle failed modules with an explicit recovery policy and, when enabled, stored incomplete executions for investigation or retry.
A practical workflow
- 01For an email-attachment workflow, verify the actual email trigger and file-upload actions supported by your accounts. Use a controlled mailbox and destination folder.
- 02Retain the message ID and attachment identifiers, then map the attachment array into an Iterator. Decide what should happen when the array is empty.
- 03Upload each attachment with the intended name and destination. Check returned file IDs or links and record how repeated source events will be detected.
- 04Configure the aggregator's source and fields to collect the upload results. Include the message ID explicitly; Group by produces a separate output bundle for each group.
- 05Send the intended summary, inspect real destination changes, and set the schedule, credit budget and error policy before enabling ongoing processing.
Summary based on source material; product claims have not been independently tested.
What makes it different
Make is a useful candidate when you need to reason about branching and changes in data volume on a visual canvas. Evaluate it using your exact array structures, destination actions and recovery requirements. A short scenario with many bundles can do more work, and cost more credits, than a longer diagram processing a single record.
Who should consider it
- Operations teams processing lists, attachments or search results across several business services.
- Teams that need conditional branches and grouped results rather than a single linear chain.
- Owners who can maintain app permissions, usage budgets and a clear procedure for failed or duplicate writes.
What to verify
- Check supported module actions, connected-account permissions and required fields with representative records; the app name alone does not establish a particular action.
- Count downstream executions after searches and iterators. Five attachments can cause five upload actions; this is not the total cost of the whole scenario.
- Distinguish Make credits from provider charges. AI connection type, tokens and some advanced modules change the billing rules.
- Inspect the destination before retrying an uncertain write. Rollback only affects modules that support transactions; it does not undo every connected app action.
Frequently asked questions
Do Router branches run in parallel?
No. Make processes routes in their configured order. A fallback handles data that does not meet the other routes' conditions. Check filter overlap instead of assuming that every input can match only one route.
Why did one input create several actions?
Search modules and iterators can emit multiple bundles. Subsequent actions process those bundles separately. Inspect the bundle count at the preceding module before estimating usage or investigating repeated output.
Does an aggregator always produce one result?
Not necessarily. Group by creates one output bundle per distinct group. Upstream fields are not automatically available after aggregation, and the empty-aggregation setting controls whether an empty result continues. Configure both the grouping and the fields you need.
Is one operation always one credit?
That is the standard rate for ordinary non-AI app operations, with exceptions. Routers and filters have no credit cost, while AI and some advanced features use different rates. With a custom AI provider connection, provider token charges can be separate from Make credits.
Does a failed scenario undo completed work?
Only transaction-capable modules support rollback. Other app actions may already have taken effect. Choose the recovery behaviour explicitly, check incomplete-execution storage settings and capacity, and inspect existing destination records before retrying.
Product facts were checked against the official product website or its first-party documentation on Nov 7, 2026.
