Transaction data – the trick to treating bank customers right
By Chris Barry
As we’re a day out from Halloween, I find it amusing to think how a night once steeped in fear and superstition has evolved into a festival that brings family and community together in the name of fun (and spending – the National Retail Federation estimates that shoppers will drop almost $9 billion on Halloween costumes, candy, cards and decorations this year).
At the heart of this evolution has been the tradition of trick or treat, and in the spirit of the season, I thought I’d use that to distill the transformative power of transaction data – often a source of fear and superstition in the banking world. I’ll do that by offering up a few short tricks and treats of my own, based on Quantium’s thirteen plus years of experience as a data science leader in the highly advanced banking sector.
Trick – customer understanding is not static.
Retention and cross-sell are the cornerstones of many businesses, and the challenge for any bank is to not only meet the current needs of their customers but to grow with them through their lifetime. Transaction data is key to this – better enabling the matching of customers with products and services that suit their aspirations and changing life circumstances by leveraging the rich insight from their spending behavior.
Take, for example, the young professional whose sustained cuts to discretionary spending and growing pool of savings might signal a mortgage in her immediate future. Or the recently retired couple whose vacation bookings suggest a likely interest in a credit card with no foreign transaction fees. Chances are that your customers have a problem that you can more easily anticipate and solve through transaction data, ultimately increasing the ease and pleasure of their banking experience by allowing you to have a timely and suitable conversation. Since customer experience is the key to loyalty, retention will take care of itself.
Your transaction data unlocks immense power in identifying these patterns, even beyond the obvious examples listed above. In our previous work, we’ve found that using transaction-based features in your targeting models can double and, in some cases, more than triple the efficiency of your owned marketing and paid media dollar across all banking products.
Treat – use transaction data to drive highly personalized and timely rewards offerings.
Many banks have rewards or cash-back programs, but how many hit the mark and engender true loyalty? Customers are time-poor and already saturated with communications and offers, from airlines to retailers right through to coffee cards. Unlike other verticals, financial service providers can see how customers behave outside of a particular retailer and across verticals. However, few banks are currently realizing the immense opportunities available to them as the center of a rewards ecosystem across all spend.
To fully capitalize on this advantage, it’s important that your data ecosystem can not only categorize each transaction properly but also accurately identify nuances beyond categorization. For example, while two customers may spend the same in the retail category, a single professional may shop high street whilst a young parent shops main street. Going deeper, transaction data can tell you who shops Prada and who shops Skechers. We have seen that connecting with customers in this personalized way significantly increases customer satisfaction and net promoter scores.
Trick – don’t ignore location data, particularly when it comes to fraud detection.
Understanding how the location of a transaction fits in with your customer’s geographical footprint is a vital part of accurate fraud detection. Are they on holiday? Do they travel a lot? Does this purchase represent a dramatic departure from their normal routine and recent transactions? The ability to answer this kind of question can power rapid text- or app-based conversations that allow customers to confirm or flag transactions in the critical moments before fraud might otherwise be perpetrated, or, equally distressing for the customer before you block their card and leave them stranded without funds in a foreign country. Crucial to unlocking this opportunity is trust in your data, and we are generally able to increase coverage by approximately 10% for industry and location classification.
Treat – more accurate transaction classification can give you a vital edge.
Whether you’re seeking to better understand a borrower’s risk profile or hone your marketing strategy for a new financial product, it helps immensely if you can properly categorize each transaction. For example, whilst in the US UberEats carries the Merchant Category Code that classifies it as taxi and transportation, our Q.Refinery platform labels it as fast food – an important distinction when wanting to glean anything meaningful from this transaction. Applying this level of accuracy to your dataset greatly increases its predictive power, especially when it comes to risk assessment – in some cases we’ve seen default predictions increase by 20% thanks to Q.Refinery.
The key Halloween message with which I’ll leave you is this: don’t let your customers live as ghosts in your datasets. Flesh them out – distill their attributes, model their behavior and build your relationship with them around those insights. Banks are better placed than most to build lasting, meaningful and mutually beneficial relationships with their customers, by virtue of their capacity to see a more complete picture of their spending and finances.
Get in touch with me. We can help turn your transaction data into a source of treats for your customers all year round.
Quantium is a world leader in data science and artificial intelligence, and its Q.Refinery product unlocks richer, more accurate information from every transaction, providing the basis for an unparalleled understanding of each customer and delivering far more powerfully predictive targeting and modeling.
Established in Australia in 2002 and now employing over 750 people, Quantium works with iconic brands in over 20 countries, partnering on their greatest challenges and unlocking transformational opportunities.