In our last webinar, we took a deep dive into how data-driven customer insights and card-linked marketing can help banks wow customers with relevant content, delivered at the right time in the right place through the right channel. Our speakers, Sára Hanniker, W.UP’s head of data science, and Tom Peace, business development leader of financial services and retail at global loyalty and benefits giant Collinson, talked optimised customer journeys, next-level digital experiences, contextual use cases and harnessing AI and ML to drive better business outcomes. Couldn’t make it? Worry not. Here are five key takeaways from the discussion.
Card-linked offers are not just merchant offers
Card-linked marketing is content marketing in a very targeted way. Or at least, that’s what it should be. Card-linked offers, or CLOs, have seen a massive growth globally, fuelled by a strong customer demand and best-of breed suppliers. Today’s consumers expect highly targeted offers that are tailored to their purchase behaviour and extremely convenient to use.
“Let’s look at an example. A bank has a customer who has never previously bought in the baby category but over the course of 6-8 weeks they have started buying nappies, a pushchair and other baby products. The bank can assume based on that person’s spending behaviour that they’ve just become a new parent: there are too many transactions for it just to be a gift and there is obviously a pattern,” Tom Peace said. “This is a point in the customer’s life when they may need to start thinking about their future financial well-being. So the bank, having seen this purchase behaviour, can start talking to them in a highly relevant way about life insurance, for example. They might be doing so through a customer service representative so they’re reaching the right customer at the right time with the right offer in the right way,” he added.
And they’re also about more than just purchase behaviour
Card-linked offers actually power the alignment of customer profiles and other forms of data, turning bulk sales messages into personalised offers. These can be offers from the bank on specific banking products or they can just be advisory products that address a customer need based on customers’ purchase behaviour, demographic information, location and so on. “The days of the generic offer, even those sent to a particular segment like millennials or retirees, are over. Banking offers now need to be extremely highly targeted,” Tom warned.
He continued: “Some of the banks we’re dealing with use customer data marrying purchase behaviour with the customer’s profile and even location through a card-linked offer programme.” What does this look like in practice? Let’s say that a bank has pre-identified a group of people who are foodies. They like to eat out and have a taste for the finer foods. The bank also sees transactions from this group in shopping centres typically between 2 and 5 p.m. on weekdays and assumes that this is the time when foodies are usually out shopping. With this information in mind, they can offer them 25% cashback at a restaurant, based on a pre-agreed deal, that could be push notified through the banking app. Win-win-win.
Open banking will open the path to hyper-personalisation
But fine-tuning offers based on customers’ spending behaviour or geolocation is only the tip of the iceberg. Thanks to open banking, banks can now pull all kinds of data and figure out when and where customers are travelling, what device they are using and even what they’re browsing for. And that’s nowhere near the end of it. Customers will soon be able to link up things like Fitbit with their banking app and allow banks to access information about their health and fitness levels. Meaning that under PSD2 rules, financial institutions can collect massive amounts of data with the customer’s permission and take targeting to a whole new level.
Let’s say that your bank knows that you’re a die-hard gym enthusiast, who works out religiously, based on the data you’ve uploaded to your Fitbit account. Now it might then see that over the course of 3-4 days you don’t go to the gym at all, and your foreign transactions show that you’re travelling in Germany for business. Your bank can offer you a one-day gym access pass to a gym that’s just down the road from where you’re located and deliver it as a QR code via email. So now we’re taking not only the geolocation, profile and the purchase behaviour, but we’re actually overlaying that with some other data sources, in this case, Fitbit.
Untidy data is no data
As rosy as this all sounds, there’s a catch. Well, not so much a catch but rather a very important rule. What banks want to achieve is smart interactions with customers. They have the data and they have the communication channels but they tend to think of what’s in-between as a magic wand and a spell. “The truth is that you actually need to put a lot of hard work into building those interactions. For each data source, be it transactional data, card data, online banking data, mobile banking data, geolocation data or device data, you need to build up a specific data pipeline first. So you want to start with cleansing, enriching and aggregating information,” Sára Hanniker pointed out.
She continued: “Let’s look at mobile banking as an example. Your clicks and events make up raw data that needs to be combed through. Why? Because you might have accidentally clicked on a function and triggered an action that you did not want to trigger. This is why mobile banking data needs to be aggregated at function, session and customer levels.” Once you’re done with all this, then you can start setting up customer profiles and micro-segment customers based on lifestyle, banking habits, product preferences, online behaviour, hobbies and more, using clustering techniques and business rules.
When it comes to customer engagement, proactive is better than reactive
Important milestones in customers’ lives are just as crucial to keep an eye on as everyday life situations. When building customer profiles, financial service providers look at longer periods of time to find out how a customer lives, what they like, what their hobbies are and so on. Once they get a sense of what’s ordinary to them, they can easily recognise what’s not. And tap into a whole world of opportunities for creating real value for customers.They can detect an online hotel booking then trigger marketing events. Or spot a huge deposit made to an account that’s unusual for the account owner. Or see charges related to house refurbishment or car repair. And when it comes to the biggest life events, you don’t just want to react to them but forecast them.
Now from an analytic point of view, that’s a bigger challenge. “That’s where machine learning comes to the rescue. Some events are easy to detect because we’ve already tracked relevant transactions and enriched them, and some of them are not, so we’ve had to come up with complex algorithms to identify them,” the data scientist explained. But one thing’s for sure. If you have your customer profiles, events and labels in order, you’re all set for creating card-linked offers that engage and delight customers and grow your bottomline.