Regional banks and credit unions have access to more customer information than ever. Between NPS scores, satisfaction surveys, call center feedback, transaction data and campaign results, most marketing teams have a reasonable picture of what’s happening across their institution. Understanding why it’s happening, or what to do about it, is another matter.
That’s where things get challenging, especially for institutions with lean marketing teams and limited research budgets. A survey might reveal that younger members are less satisfied than other age groups, or campaign results might show that one product message performs better than another. Digging into the reasons behind those findings takes time and money, and traditional research isn’t always feasible to inform the decisions that marketers need to make.
As a result, a lot of marketing decisions are made with less consumer insight than we’d like. We rely on the data we have, our experience and the perspectives of colleagues around the table. Sometimes that’s enough. Other times, we’re making educated guesses about what consumers want and how they’ll respond. Synthetic research offers a way to bring another perspective into those conversations, and I think financial institutions should be paying attention.
An opportunity to explore more, more often
Synthetic research uses AI to simulate how defined audiences might respond to questions, concepts or messaging. When developed using credible data and sound research practices, synthetic panels can help marketers explore consumer perspectives without recruiting participants for a traditional study every time a new question comes up.
At Mower, we’ve been testing synthetic research alongside traditional methods, with encouraging results. Our research team recreated previous client studies using synthetic panels and compared the findings with the original research. We saw strong alignment in broad themes, decision logic and differences between audience segments, particularly in work focused on understanding perceptions and evaluating ideas.
Those results don’t mean synthetic research can, or should, fully replace traditional human studies. We’ve also seen limitations, especially when trying to predict precise behaviors or uncover insights that fall outside the information used to develop the synthetic audience. Still, the consistency we’ve seen in certain applications suggests there’s real value in using it to inform decisions earlier in the process.
Consider a regional bank developing a new small-business banking campaign. The marketing team may have a good understanding of its products and competitive advantages, but less clarity about which benefits will resonate with different business owners. Synthetic research could help the team explore several positioning approaches, identify language that feels generic and surface questions worth investigating before investing in creative development or a larger research study.
For a credit union planning a home equity campaign, the opportunity might be comparing how different homeowner segments interpret a product message. That kind of exploration can help marketers refine their thinking and make better use of the resources they have.
Getting more from the data you already have
I’m particularly interested in how synthetic research can complement the customer information that banks and credit unions are already collecting. Take a credit union with strong overall satisfaction scores but lower engagement among younger members. The data identifies an issue, but it may not explain whether those members are less aware of available products, see the credit union as less relevant to their financial needs or simply prefer other providers for certain services.
A synthetic panel could be used to explore those possibilities and evaluate how different messages might be received. The responses wouldn’t establish what’s driving the behavior of actual members, but they could help the marketing team develop more useful hypotheses and determine which questions deserve further investigation.
That distinction matters because traditional research is still the best way to validate what real consumers think and do. Synthetic research can make that investment more productive by helping teams narrow their focus beforehand.
There’s also a practical advantage in being able to revisit a question quickly. If an initial round of exploration suggests that a product benefit isn’t as compelling as expected, marketers can examine alternative approaches without organizing another round of participant recruitment. For institutions that typically reserve qualitative research for major initiatives, that flexibility could make consumer exploration a more regular part of marketing planning.
Understanding where it fits
Financial services is an industry where the limitations of synthetic research deserve particular attention. Consumers don’t always make financial decisions based on a careful evaluation of features and benefits. They may misunderstand terms, delay decisions, stick with an institution out of convenience or respond to personal experiences that are difficult to model.
Synthetic audiences can also behave more like informed, rational consumers than people do in real life. That makes them useful for exploring perceptions and comparing ideas, but less dependable for predicting actual account openings, product adoption or market demand. However, I wouldn’t use synthetic research as the sole basis for a major brand investment, a market-sizing exercise or a decision that depends on precise estimates of consumer behavior. Those situations call for traditional research, real-world testing and other forms of validation.
The quality of the research design matters just as much as the technology. Defining the audience, selecting the right inputs, asking useful questions and interpreting the findings all require expertise. A synthetic panel can produce persuasive responses even when the underlying assumptions are weak, which is why experienced researchers need to guide the process.
A more practical approach to consumer insight
For marketers interested in getting started, I’d look at the decisions already on your calendar. Maybe you’re refining deposit messaging, considering a new commercial lending audience or trying to understand why a product isn’t gaining traction. Your existing data probably provides some clues, along with questions you haven’t had the resources to explore. Those questions are a good place to begin. Use synthetic research to examine possible explanations, compare approaches and identify what warrants additional testing. Then bring those findings together with your customer data, market knowledge and traditional research where the stakes justify it.
For years, many regional banks and credit unions have treated deeper consumer research as something they can afford only occasionally. That’s understandable given the cost and time involved, but it also means important marketing decisions are often made without the benefit of exploring how consumers might respond.
Synthetic research won’t eliminate uncertainty, and it shouldn’t give us false confidence in findings that haven’t been validated. What it can do is make it easier to investigate more ideas, challenge more assumptions and use limited research budgets more effectively.
For financial institution marketers who have spent years wishing they could dig a little deeper before making the next decision, that’s a meaningful opportunity.