Operations Hub software: sync, programmable automation, and data quality
Operations Hub software tests the product's own mechanics rather than HubSpot strategy: data sync with an external system, programmable automation via custom code and webhook actions, and the data-quality steps that standardize and merge records. Scenario questions ask you to predict an outcome - which side wins a conflict, whether a formatting workflow touches historical records, or why a dedupe rule missed an obvious duplicate.
How data sync moves records
A sync connection pairs HubSpot objects with objects in the external app. Only mapped fields move; an unmapped property stays untouched no matter how well populated it is. Sync rules filter which records are eligible at all, so a filtered-out record is never "stuck", it is simply excluded. The connection's conflict setting decides the winner when the same record changed on both sides. The sync history log lists per-record failures, which means an out-of-sync record is usually a mapping, validation, or sync-rule problem rather than a scheduling problem.
Programmable automation and data quality
Custom code actions run JavaScript against the enrolled record's properties and must return a plain object of property updates; anything not returned is left unchanged. Webhook actions post record data to an external endpoint and suit cases where an outside system must act. The Format data action normalizes capitalization, phone numbers, and dates inside a workflow - but only for records that workflow actually processes. Deduplication matches on chosen properties, so normalization usually has to run before matching.
Worked example
A connection is configured so the most recently updated side wins. A rep corrects a phone number in HubSpot at 09:00. At 09:20 the external ERP pushes an unrelated address change on the same record. On the next sync the ERP side is newer, so its version is applied and the corrected phone number can be overwritten. Reasoning: the conflict rule is judged on update time, not on which edit was more correct, so the durable fix is excluding the phone field from the mapping or narrowing the sync rule - not retyping the value.
Traps that cost marks
Assuming a formatting step cleans history retroactively. Assuming unmapped fields sync. Assuming a merge is reversible when duplicates turn out to be distinct records. Assuming a code action's changes apply when the script throws before returning.