Your Bank Can’t Personalize What Your Customer Data Can’t See
By Jim Marous, Co-Publisher of The Financial Brand, CEO of the Digital Banking Report, and host of the Banking Transformed podcast
Simple Subscribe
Subscribe Now!
Retail banks have more marketing technology, more customer data, and more ways to reach customers than ever. Yet our latest State of Financial Marketing research shows that the foundation beneath all of that activity remains surprisingly weak.
More than half of institutions surveyed are already using generative AI in marketing, but not one said its customer data was immediately available and AI-ready. We also found a significant gap between what financial marketers call personalization and what customers actually experience.
Most personalization still amounts to selecting an audience and sending it a message. The real opportunity is using customer intelligence to recognize what is happening in a person’s financial life and respond while it matters.
Closing that gap starts with better data foundations, clearer accountability, smarter measurement, and a willingness to invest in capabilities that may be less visible than another campaign or AI-generated asset. Listen to my recent podcast to learn more.
Need to Know:
- Better targeting is not the same as personalization. Customers experience the offer, timing, onboarding, and relevance of an interaction, not the segmentation behind it.
- AI cannot compensate for weak customer data. More than half of surveyed institutions are using generative AI, yet none reported having customer data that was immediately available and AI-ready.
- Marketing budgets should follow customer impact. AI-driven content and creative received significant planned investment despite ranking last among the tactics respondents considered effective.
- Banks are underinvesting in existing relationships. Deposit growth ranks as the top priority, while primary-bank relationships, share of wallet, win-back, dormant accounts, and onboarding receive much less attention.
- The next step does not require a massive transformation. Start with one accountable owner for the customer record, a unified daily view of customer behavior, and basic control-group testing to establish incremental lift.
The Personalization Gap Is Bigger Than It Looks
Hyper-personalization may be one of the most frequently used phrases in financial marketing. I have used it myself. But our research made me stop and reconsider what we actually mean by it.
Roughly one institution in 100 told us it was truly delivering hyper-personalization. More revealing, when we asked institutions to define personalization, about two-thirds of what they described came down to choosing which audience receives a particular message.
I recognize that approach because I used it when I ran marketing at a bank.
The data came from a core system, often sorted by geography and balance tier. We selected a segment, chose an offer, sent the mail, and counted the coupons that came back. We called it targeting. Today, the tools are vastly better. We can reach customers wherever they are, test far more variations, and accomplish in an afternoon what once took a team months.
That progress matters. But the underlying decision has not changed as much as we think.
We are still deciding which list gets which message.
Key insight: The customer never experiences the segmentation. They experience the offer, when it arrives, how the onboarding works, and whether the bank seems to understand what is happening in their life.
That distinction matters because our research shows that financial institutions have invested most heavily in the part of personalization customers cannot actually see.
This is partly a reporting problem. An organization can report strong personalization capabilities while delivering an experience that customers would never describe as personal.
Data Determines What AI Can Do
The reason that gap persists is fairly straightforward: everything underneath the marketing depends on data.
When I talk about fuel for the growth engine, I am talking about whether customer data is unified, current, and usable at the moment a system needs to act.
Roughly two-thirds of the industry told us their data is either scattered across systems or updated in batches. Almost nobody described it as real time, and nobody told us it was immediately available and ready for AI to act on.
That should change how we think about our AI investments.
Generative AI can produce more content. Predictive AI can help determine who should receive something and when. Agentic AI can potentially take actions based on those decisions. But each step depends on having usable customer intelligence underneath it.
Key insight: That is why one of the most striking findings in our research was also one of the simplest: AI-driven content and creative ranked last among the tactics respondents considered effective — even as it was receiving one of the largest planned increases in budget.
I do not take that as an argument to stop using AI. I take it as an argument to invest more deliberately in what makes AI useful.
You do not need perfect data to start. Good solution providers can build effective targeting from imperfect data today, and most institutions should begin using these capabilities now.
But there are grades of fuel.
You can run an engine on regular fuel. It will work. That is essentially what many banks have been doing with customer data for years. What it cannot do is support the kind of real-time intelligence that recognizes something has just happened to a customer and responds while that event is still relevant.
That requires a better foundation.
Put Investment Where Growth Happens
There is another pattern in the research that deserves more attention: we tend to outsource our strengths and protect our weaknesses.
Most institutions bring partners in for media planning. About half use outside help for creative production. Personalization and decisioning, meanwhile, are among the least outsourced capabilities.
There are legitimate reasons for caution, particularly when decisioning involves customer data, governance, and fair lending. But if a capability is weak internally, refusing outside assistance does not make it stronger. It can simply make the gap permanent.
The same issue shows up in the growth plan itself.
Deposit growth is the industry’s No. 1 priority. Yet the activities that help deposits become durable relationships, including becoming the customer’s primary financial institution and increasing share of wallet, rank near the bottom.
Key insight: We are putting enormous energy into acquiring customers while giving comparatively little attention to what happens after they arrive.
The same is true of dormant accounts and win-back opportunities. These customers already exist in our databases. Their behavior is already visible. We know the relationship, at least to some degree. Yet these opportunities barely register on many priority lists.
That is where I would start:
- First, make one person accountable for the customer record, with a measurable target that makes success or failure visible. Then build a unified view of checking, card, and loan behavior that is refreshed daily. Daily data will not get you all the way to real-time marketing, but it gives you a foundation to build from.
- Move a visible portion of the planned generative AI budget increase into the data foundation. That is a harder budget conversation because infrastructure does not produce a shiny output that is easy to show in the next meeting.
- Start measuring incremental lift.
Hold back a portion of the audience in your next campaign and establish a control group. Most of the industry is not doing this consistently. It is one of the cheapest ways to find out whether your marketing actually changed customer behavior.
The broader point is simple: better marketing starts with better customer intelligence.
Bottom line: Our customers want us to know them, understand them, and provide relevant guidance while they are still in the moment. The technology to do that is increasingly available. The question is whether we are funding the foundation that allows it to work.
Before the next planning meeting, I would put two charts side by side: the marketing budget and the investment in the data foundation. Then ask one question:
Are we funding more marketing, or are we funding better marketing?
