How it works today vs. a single source of truth
Same cohort — Mastering ToP · Apr 22, 2026 · MRA MN — modelled two ways.
Today — logistics stamped on every contact brittle
The venue / date / hotel / trainer is written onto each attendee individually.
Change the venue → you must edit all 9 records. Miss one → that person's email merges a blank.
Recommended — one Course record “Stamps” info for all attendees single source of truth
The logistics live once, on the Course record. Attendees are associated to it.
✓ Change the venue once → it updates for all 9 instantly. Nothing to keep in sync.
Why the current model is hard to scale & keep accurate
No stable cohort key
A "cohort" is only identifiable by the fuzzy string course_name + date + venue.
course_code / course_id are empty on every record.Dirty, blank-prone identity
Names carry typos and many records are no-date, no-venue, or TBD — exactly the blanks that flow into the emails.
Duplication = drift
The same logistics are repeated on every attendee, so there is no one place to update and values drift apart over time.
Evidence: Apr 22 venue/hotel/trainer copied across all 9 attendees.Tiny, fragile coverage
Only ~0.9% of contacts are course-tagged, so for almost everyone the merge fields resolve to nothing.
Evidence: 101 of 11,290 contacts; venue set on just 73.What "right" looks like
- Use the native Course object (already active on your plan). One record per dated session — holds date, venue, hotel, parking, lead trainer once.
- Associate attendees to the cohort with association labels —
Registered·Attended·Cancelled·No-show— the exact pattern HubSpot's community recommends for events and courses. - Stop duplicating logistics on contacts. The contact keeps who they are; the cohort keeps what/where/when. Update the venue in one place.
- Send cohort comms as workflow-triggered marketing emails keyed to the cohort, not 1:1 sequences. (Personalization in sequences reads from the contact/sender, not associated custom objects — another reason these belong in workflows. Ties back to the "sequences vs. workflows" mismatch.)
What the authorities say
"Each object has individual records which can be associated with records of other objects… information is stored in properties." The data-model usage view even surfaces "the number of properties that have no data, are unused, or are potential duplicates."
"When you are ready to start establishing relationships in your data between one-to-many rather than one-to-one, you will want to start structuring your data with custom objects." Update a shared record once and it "would automatically update and reflect" everywhere it's associated.
Create an Event/Course object and associate contacts with labels for "Registered, Attended, Cancelled, and No Show"; change the relation to many-to-many so "a single contact [can] be associated with multiple courses."
Move high-volume / repeating data off contact properties and into custom objects, where you have far more room and a cleaner structure; audit and de-duplicate properties first (teams often cut 30–40%).
"Accurate CRM data allows sales and marketing to operate from the same source of truth." MDM creates a single "golden record"… and strong data hygiene drives "up to 70% improvement in email deliverability and 25% increase in sales productivity."