Open the workflow list in any HubSpot portal older than three years and you will find the same fingerprint: pairs of workflows named something like "New Project Record - Part 1" and "New Project Record - Part 2." Nobody designed that. It exists because step four could not see the record that step three had just created, so an admin split the logic across two canvases and wired the second one to a creation trigger.
That workaround is no longer necessary. HubSpot now lets you use records created in a workflow later in the same workflow, which means most of those pairs can collapse into one. The knowledge base article documenting it was last updated on July 13, 2026, and at the time of writing it is still labeled beta, available on Marketing Hub, Sales Hub, Service Hub, Data Hub and Smart CRM at Professional and Enterprise. If you inherited a portal with hundreds of workflows, this deserves a deliberate cleanup pass, not a mental note for next time.
What "use records created in a workflow" actually changed
Mechanically, it is a small addition. After a Create record or Create task action runs, later actions in the same workflow expose a new option in the object dropdown: The created object. Select it and the action targets the record the workflow just made instead of the enrolled record. In the data panel, that record also appears as a data source under the Create record action, so you can pull its properties into any field that accepts data variables. HubSpot's documentation notes the created-object target is not available on every action, so plan to test rather than assume.
Practically, it is larger, because it removes the only reason the two-workflow pattern ever existed.
Why the chained pattern cost more than it looked like
- A timing gap you cannot control. The downstream workflow enrolls on record creation, which is asynchronous. Between "record created" and "record finished being configured," a half-built deal, ticket or project record is visible to reps and countable by reports.
- A doubled enrollment surface. A creation-triggered workflow enrolls every record of that type created by any route: imports, API writes, integration syncs, manual creation. Each of those eventually produces a record that should not have run through Part 2, so admins bolt on exclusion filters that nobody documents.
- Fragmented history. Enrollment history lives per workflow and is retained for six months. When one process spans two workflows, tracing a single record means reconstructing the sequence across two logs.
- Portal budget. Workflow counts are capped per account, commonly documented at 300 on Professional and 1,000 on Enterprise as of August 2026. Confirm against your own subscription, but every "Part 2" occupies a slot you may want later.
Step 1: Find the chains
Start from the workflow list rather than from memory. Filter to workflows whose enrollment trigger is object creation, then apply one test to each: if I turned off the upstream workflow, would this one still have work to do? If the answer is no, it is a continuation workflow and a consolidation candidate. If the answer is yes, because records also arrive from forms, imports or an ERP sync, leave it alone. Collapsing those would break the other entry paths.
Rank the candidates by monthly enrollment volume and by how much the timing gap actually hurts. A chain that assigns an owner and creates a task for a rep is more urgent than one that backfills a reporting property.
Step 2: Rebuild it on one canvas
Put the Create record action where the first workflow ended. Set the required properties for the object type you are creating: email for contacts, company name or domain for companies, deal name plus pipeline and stage for deals, ticket name plus pipeline and status for tickets. The action covers the standard objects plus quotes, orders, projects and custom objects, with the last requiring Enterprise.
Then move everything the second workflow used to do into subsequent actions on the same canvas, pointing each at The created object:
- Set properties on the new record, including ones computed from the enrolled record earlier in the same run.
- Add associations. Records created by a workflow are associated with the enrolled record by default, so you only need explicit association actions for the other relationships, and that is where association labels belong.
- Create the task or note, using data variables from both the enrolled record and the created record in the body.
- Send the internal notification, which is usually the step that made the old chain necessary in the first place.
One option worth knowing: the Create record action can copy timeline activity from the enrolled record to the new one, defaulting to the last 30 days. On a project or deal record that should inherit a contact's engagement history, that replaces a separate sync job.
Step 3: Handle the edge cases before you turn it on
Consolidation moves failure modes rather than eliminating them. Five to check:
- Loops. If the workflow creates the same object type it enrolls, verify the new record does not satisfy the enrollment trigger. HubSpot flags this explicitly, and an unbounded creation loop is the fastest way to turn a cleanup project into an incident.
- Field type compatibility. Copying a property from the enrolled record to the created record only works between compatible field types. Incompatible properties simply do not appear as options, which reads like a missing property rather than a design constraint.
- Number fields. Personalization tokens cannot be used in number fields. If your old Part 2 workflow set a numeric value from a token, that logic needs a different approach.
- Object scope. Leads can only be created in contact-based or company-based workflows, and a custom object must be associated with the workflow's object type before you can create it there.
- Failure of the create step. If the Create record action does not produce a record, every downstream action pointed at the created object has nothing to act on. Add a branch that checks a property on the created record before the notification and task steps, so a failure surfaces as a routed exception rather than a silent gap.
Also watch the data panel ceiling: a workflow can hold up to 50 data sources, and consolidation pulls two workflows' sources into one.
Step 4: Retire the old workflow safely
Turn the downstream workflow off. Do not delete it. Deletion removes your ability to compare behaviour if something looks wrong three weeks later, and the workflow slot is not scarce enough to justify the risk.
Before you flip it, export a baseline. Filter the old workflow's enrollment history to a representative window, use the "only enrollments that reached" filter to isolate the steps you care about, and export the result to a static segment. After the consolidated workflow has run a full cycle, compare record counts, task counts and property values for an equivalent window. Two constraints to work around: HubSpot retains enrollment history for six months, and there is a 100,000 log limit for successful executions per day, after which success logs stop being stored while error logs continue. On a high-volume portal, sample rather than assuming the log is complete.
Give it 30 days, reconcile once more, then delete.
What this looks like in a distributor's portal
Here is a shape we see repeatedly in building materials and industrial distribution. A quote request form fires a workflow that creates a project record. A second workflow, triggered on project creation, associates it to the parent company, stamps branch and territory from the contact's postal code, sets an estimated award date, creates a task for inside sales and notifies the branch manager. Somewhere along the way a third workflow appeared to handle projects created by the ERP integration, and a fourth to fix records the third one mishandled.
The costly symptom is not the workflow count. It is that the branch manager gets the notification before the record finishes being configured, so the project they open is missing the territory and owner the notification implies. Consolidating closes that gap to a single run, and collapses the "why did this record get skipped" investigation from four enrollment logs to one.
Where to start on Monday
Export your workflow list, filter to creation-triggered workflows, and apply the single test above to each one: would this workflow have work to do if the upstream workflow were turned off? Count the ones where the answer is no. That count is your consolidation backlog, and it is almost always larger than the team's estimate.
Then rebuild exactly one of them, the highest-volume chain, using the ability to use records created in a workflow later in the same run. Reconcile it against the baseline before you touch the second. The goal is not fewer workflows for their own sake. It is automation a new admin can read on a whiteboard, because that is the automation that survives the next person to inherit your portal.
