Credit Unions Have the Member Data. Here's Why It's Not Driving Product Recommendations.

Let's imagine this scenario.  A member opens a 48-month auto loan in March. Makes every payment. In month 30, she logs in to mobile banking more frequently than she has in the past two years. Her direct deposit went up. She's started making transfers to savings. She could just appear to be an active member in good standing. However, consistent payments throughout a loan term, increased engagement, and income growth are fairly reliable indicators that she's approaching a point at which a new financial decision is imminent. Maybe a vehicle upgrade. Maybe a home purchase. Maybe a HELOC to finish a renovation she's been putting off.

The credit union's core system will have most of this information, but that doesn't mean the credit union's marketing platform knows any of it. And so, when a person from the credit union does have a conversation with the member, the next best product information just isn't accessible when it actually matters. That's not a data problem. Most credit unions are sitting on better member insight than they realize, just unable to do much with it. 

How Should Credit Unions Structure Product Data in HubSpot?

When we start working with a credit union on their HubSpot setup, one of the first things we look at is how product data is stored and organized. Typically, we find a member object with a series of flat properties. Has mortgage: yes. Has auto loan: yes. Has checking: yes.  Flat presence flags record what a member has. They can't describe the health of the relationship.

That distinction matters because a recommendation has to reason from attributes, not just presence. Knowing a member has an auto loan isn't useful on its own. Knowing they're 30 months into a 48-month loan at a rate that's now two points above market is something you can build a conversation around.

Why We Recommend Modeling Product Holdings as Deal Records

At Geear, we recommend modeling product holdings as deal records in HubSpot. Deal records give you product-level attributes, reporting, pipeline visibility, and the ability to reason across the full member relationship in one place.

Custom objects can also hold product data, and they're the more flexible choice when a product doesn't fit a pipeline model. But deals come with native reporting, pipeline stages, and association tools already built in, which means less configuration and more downstream value for most credit union use cases.

Why Do Credit Union Data Projects Fail to Get Used?

Getting the data structured is step one. Step two is making sure the output ends up where it's actually useful.

This is where most credit union data initiatives stall. The analysis gets done. A report gets built. Someone has to remember to run it before a campaign or open a separate dashboard before a member call. And in practice, that extra step doesn't happen consistently which means the insight exists in theory but doesn't change what teams actually do.

A member service rep fielding a call doesn't have time to cross-reference a report. Take the member who's 30 months into her auto loan. If her HubSpot record surfaces a scored recommendation the moment it opens, for example, for auto loan refinance, based on her rate and payment history, the conversation changes.

Not because the rep went looking for it. Because it was already there when they opened the record.

That's the difference between a data asset and a data tool. A data asset sits somewhere, waiting to be queried. A data tool shows up at the moment of the conversation.

What this looks like in a campaign context

The same logic that makes individual member conversations better also makes campaigns sharper.

Instead of pulling a list of all members without a HELOC and sending the same email to 4,000 people, a credit union with structured product data and a scoring layer can do something more specific: members who have had a mortgage for more than three years, whose estimated equity position makes them eligible, who haven't opened a new product in 18 months, ranked by likelihood to engage. That's a list of 200 people who actually fit the offer and the response rate reflects it.

This is what "relevance" actually means in practice. Not personalization in the first-name-in-subject-line sense. Relevance in the sense that the offer reflects something real about where that member is financially right now.

Where should credit unions start with product data?

If you're trying to close this gap, the most valuable audit you can do is to review how product data is actually stored in HubSpot. Are products represented as records with properties (open date, rate, term, status) or as flat fields on the member record? That one structural question will tell you a lot about how much signal you're currently leaving on the table.

It's also the first conversation we have with every credit union that comes to us asking about Next Best Product. Because no recommendation approach, however sophisticated, can produce useful output if it is just a checkbox.

In this article, we've made one point repeatedly: the member insight already exists, but it isn't always accessible to the credit union teams who could put it to use. Next Best Product closes that gap inside HubSpot, putting scored product recommendations directly on the member record, tuned to each credit union's portfolio and goals.

We're rolling it out to a select group of credit unions ahead of a wider release. If your team has been circling this problem, register your interest, and we'll follow up with next steps.