Using Behavioral Analytics in Banks’ Digital Content

  • Ohad Samet
  • January 6, 2015
  • 1

How can banks know what customers really want?

Many banks plan to further invest in digital in 2015. It makes sense – digital is immediate, it’s always on, it’s a cost saver – and it pleases customers. The key is to give customers what they need, not only what they say they need. Behavioral economics research allows us to do this and to differentiate between declared and revealed preferences – what customers say they want, vs. what they really want – to engage with customers in a meaningful manner.

As we know, the Internet and the digital revolution have changed the way we contact people; a handwritten piece of mail, postcard and phone call have, for the most part, been replaced by texting, IMing, emailing, social media and more. Even voice calling is sometimes considered rude.

The CDC reports that two out of every five US households do not have landlines, and the number is higher for lower-income families. Companies see more mobile and tablet traffic, even for business functions that were traditionally handled in branch – including investments management, account management, and the already established mobile payment processing.

Shifting to digital? It’s not that simple

Not surprisingly, these trends are prompting banks to invest more heavily in digital strategies. What should be included in it, though? Is a site redesign and self-service portal going to shift consumer behavior over night? As it turns out, the answer is no. It’s not enough to create a website – you must present it in a way that makes sense to customers, and seamlessly integrate your digital channels into their day-to-day activity, ultimately giving them what they need and not necessarily what they say they want.

Behavioral economics teaches us that customer adoption is much smoother when their cognitive load is reduced – meaning, a choice creates less innate objection and feels more “natural.” How can we understand what feels natural to the customer without asking them outright? How can we infer what to offer without confusing the customer with too many choices?

Use data to infer customer preferences

The key to answering engagement questions is collecting, analyzing and acting on data that customers create when they interact with your system, even if they never explicitly identify their preferred choice. An important, yet often overlooked advantage of digital communication is the amount of feedback you can glean from customer behavior. Every opened email, every link clicked and every page visited can inform how you engage with your customer. Going digital means much more than just making online communication possible – customers want to engage in many different ways, and allowing them to do so will increase their engagement. Underlying that flexibility, though, must be strong channel integration and a data structure that will allow you to both understand what they want, and respond to their actions in short enough cycles to be able to capture intent.

Responding to revealed preference: two examples

Take a debt recovery platform for example, such as what we do at TrueAccord. Many customers in a debt collection process will not disclose their financial situation or preferences for settling their debt; however, the offers they browse through, and the order in which they browse them, indicate what’s on their mind. There is a distinct browsing pattern for customers who have money and are looking to settle, and those who need a bit more time or an affordable payment plan. By using those patterns, one can construct follow up flows that offer solutions that better fit the customer’s needs. Rather than trying to guess or engage in a confrontational interaction – something customers clearly don’t want – we are able to offer what they need and get a positive response. An effective data collection strategy helps infer the follow-ups, and cross-channel integration enables almost immediate action, capturing intent.

Similarly, the number and order of opens and clicks on messages the customer receives are a good indication for what they care about. One can gauge interest in an offer, detect changes in financial status, and even understand what kind of content customers better react to. This knowledge can then be used to prioritize not only the channel through which you communicate with a customer, but also the type of content, messaging time, and even the name of the agent that messages them. All of these factors contribute to reduced cognitive load, and increase the chances of the customer actually positively responding to you – all else being equal.

Bottom line

When implementing a digital strategy, pay attention to data collection and actionable insights, so that your design will support what customers really want and need, not only what they say.

Ohad Samet is a co-founder and CEO of TrueAccord. For more information, please visit

  Like This Post

One thought on “Using Behavioral Analytics in Banks’ Digital Content

Leave a Reply