My primary bank has dodged a few bullets — six, actually — in the past 24 months. But I’m pretty sure it has no idea. The bank still gets my direct deposit and many of my day-to-day transactions. That said, I am not the loyal customer I might seem, despite my steady transaction stream. I rarely engage with the communications they send. I never visit their branches. And this year, thanks to COVID-19, I no longer need their ATMs since many people don’t want to handle cash.
The bullets it has dodged? Those are the six other bank apps I’ve accumulated on my phone over the past two years or so, most of which are neobanks. That doesn’t even include Venmo and Paypal, which have moved from rarely-opened novelty apps to essential payment utilities.
So why haven’t I moved beyond kicking the tires with one of these interlopers and made a switch? It’s because my early experiences with them — while friction-free, easy and digitally slick — were also generic and devoid of any hooks that fostered a relationship. After giving each of them an honest chance, I concluded that none were offering anything more revolutionary than basic bank utilities and that I would never be anything more than an anonymous customer to them. It’s just not worth the pain of switching — at least not yet.
Banks on both sides of this equation are the losers here. While my traditional retail bank has enjoyed a stay of execution, it is still in a precarious situation and still hasn’t changed the way it interacts with me. Meanwhile, the neobank challengers have wasted precious marketing dollars by giving away less-than-compelling incentives.
What’s sorely missing on both sides is digital empathy. Banks of all stripes desperately need more marketing and product interactions that evoke a sense of connection. In time, profits will follow.
NO EASY TASK
There are certainly challenges to this goal, but it is essential. A bank that isn’t responsive to its customers’ behavior and fails to interact with them in relevant ways won’t have those customers for long.
The promise of personalization is high, but given the breadth of impact, the value is difficult to measure. That can make it a tough sell in the budget process, especially when marketers are allocating precious investment dollars.
But a growing amount of evidence shows that it is worth the effort. Banks that establish digital empathy with their customers through personalization enjoy benefits at every stage of the customer lifecycle — from reducing the cost of acquisition to improving primacy and driving relationship depth. For example, Curinos has demonstrated that creative personalization can more than double the performance of the existing decision engine. (See Figure 1.)
Figure 1: Creative Optimized for Customer Existing
Creative optimization had the biggest lift on the segments who previously were the least-engaged.
Note: No offers were included in any of the messages. Emails were only educational content with an invitation to apply. Customers were equally eligible for all content.
BANKING BASICS
Setting aside for a moment the wide variability of salesforces, let’s contrast a typical bank’s digital marketing interactions with those initiated by a typical banker. Where digital marketing engines blast out “personalized journeys” that are mostly linear and trigger-based communication streams that are, at best, tailored to broad segments (such as mass affluent or young professionals), a banker engages in a rapid test-and-learn exercise to build a direct relationship with the customer.
Among other things, she will greet you sincerely and ask you what you’re looking to accomplish in your visit. She’ll probably ask some questions about your current situation and financial goals. She might throw out a bad joke to see if you respond to humor. She’ll mention a service or product feature to gauge your interest or point out some bank resources you might find helpful. Every bit of stimuli she puts out there, she’s watching — assessing how it landed and using that information to either change her approach or dive deeper into a topic. These micro interactions elicit feelings of connection and increase the relevance of the bank and its solutions to the customer.
It is difficult to replicate the banker’s interactions in digital channels at scale and it is impossible to do if operating teams use the same tools and approaches to drive personalization that they rely on today. Today’s operating models — even the agile ones — and martech stacks are designed around a linear workflow that require teams to predefine segments and then preconceive “personalized journeys” for each of them, respond to any triggers and/or incorporate the next recommended actions.
All this hard wiring requires massive resources. Analytical teams must first determine how to segment customers and then for each segment quantify the most critical areas of relationship-building opportunity. Creative teams then generate distinct executions for each segment opportunity and digital marketing teams program test cells, work with analytical teams to read results and then scale up the winning approaches.
Not surprisingly, this quickly hits a point of diminishing returns — and keeps digital relationship-building stuck in automaton land.
A BETTER WAY
Advances in marketing software offer a path out of this linear logjam. A more human approach is possible with continuous test-and-learn capabilities driven by machine learning and AI. Instead of prefabricating and predetermining customer journeys and interactions, new technologies enable marketers to let machine learning drive the personalization based on a rapid test-and-learn approach to digital relationship building that is rooted in the assumption there are many possible “best” treatments.
With machine learning at the core, teams can invest their time and energy in building a robust library of treatments for any objective — such as driving primacy during the first 60 days. This library of executions, or executional components, flex different variables — from the topic communicated to positioning, proof point emphasis, tone, imagery, copy length, sequencing and incentives. A sophisticated machine learning engine can then capitalize on the treatment library to run tests with micro-segments at a scale that simply isn’t possible for human teams. It can also observe downstream behavior to “learn” what’s working (or not) and automatically adapt campaign messaging going forward. Such an approach dramatically improves the relevance of each customer interaction without requiring a scale-up in personnel.
