The conversation about how we are going to do this “branch transformation” thing, and the role of technology in that process, continues:
“So, the future of banking is enabling a customer to come into a branch and interact with a bank employee who isn’t in the branch?”
Cornerstone Advisors report: “The Quest for Video Teller ROI”, November 2016
Our colleagues at Cornerstone have published what I believe is a quite fair – and extensive - assessment of the branch transformation technology explosion in retail banking, focusing on the question: “Is there a value proposition for these new assisted self service technologies, and particularly video (remote) teller solutions?”
Their answer, after looking at the question from many different angles, is exactly what you would expect: “sometimes”.
Why is that? Why is the answer “sometimes”? Why not “always”?
In my role at Glory, I visit with bank executives all around the world, talking about their business plans, and the products and services we might develop in support of those plans. These folks, almost universally, are trying to perfectly model customer wants, needs, and behavior and perfectly match that with a channel delivery strategy that puts the FI where the customer needs them at exactly the lowest cost of delivery. They spend millions and millions of dollars in “customer journey modeling” and “innovation laboratories” and “big data analysis” en route to these “perfect plans”.
Then, after they cut the opening day ribbon for their latest, perfectly sized, perfectly outfitted facility that perfectly aligns with their scientific analysis of the surrounding marketplace – and they find it’s not working perfectly: there is more or less work than predicted, deposit and loan activities don’t match expectations (higher or lower), or some other frustration or opportunity happens.
Of course everyone celebrates for exceeding expectations, and we all try to fix disappointments – but in both cases the “perfect model” has failed. The best FI’s learn, of course, and iterate, improve their models, and try again.
But why don’t the models “just work”? Simply because of this: we can measure trends and averages, but we cannot predict life, we cannot predict individual lives, and we cannot predict the next moment in a life that may create an event that leads to a bank visit.
In Glory’s product design process, we talk incessantly about quality. It’s a top priority for us. One rather simplistic measure of quality is called “mean time between failure”. You want a long time between failure events. At Glory we have some ridiculous MTBF standards, from a human lifetime perspective – some parts are expected to have 1,000+ years MTBF. What does it mean, 1,000 years MTBF? It means that sometime between right now and 2,000 years from now, the part will fail. But we cannot tell you which hour, minute, and day that will be. Or which part on which machine somewhere in the world will fail. And that’s life – and that’s why every predictive model will fail right now, but many, and most good ones, will be accurate over the longer term. (BTW - MTBF is just one measure, we use much more sophisticated ways that we evaluate likelihood of failure over time, and our predictive models and proactive management solutions are quite good - thank goodness machines are more predictable than people!)
What does this have to do with bank technology? And that word “sometimes”?
When we introduce the daily chaos in the world of banking – those pesky customers - we get to a “mean time between needs” that is nearly totally unpredictable. There are simply too many random needs that appear at random times to build a perfect model. I’ll give one example from some recent “man in the street” interviews we conducted:
Q: “Why did you come to the branch today”
A: “I brought a power of attorney for a hospitalized friend so I could help him pay his bills”
There is absolutely NO way to model that kind of situation. Ever.
No self-service system can be cost-effectively prepared for that kind of “edge” use case. No video-teller service will be well prepared for efficient delivery of non-linear workflows. The best branch designs, now and probably forever, will always be equipped with the most flexible service solution ever invented: on-site, well-trained, empathetic, human bankers.
So, we must put the right people in places to provide the flexible response required to meet deviations from the standard model, otherwise known as “customers”. Once we have these well-trained, empathetic people in place, only then should technology be applied, by answering two simple questions:
- “Why are we not talking to customers right now?”
- “Why is that customer waiting for someone to help them?”
And, once we understand each of those questions, we add a third and fourth question:
- “What am I going to do about that?”
- “Will that make the customer happy?”
The answers will be different for different FI’s, in different situations, of course. Is there a stupid reason? A rule that doesn’t fit reality? Is there no one trained to help? Am I caught up in a task that might be automated with technology? Is it an unusually busy moment that occurs rarely but could be managed better with technology? Or – is the right answer not technology, but rather, more people (I’m not even sure I’m allowed to suggest that anymore!)
If the answer to the third and fourth questions might be “technology”, we suggest that these should be a scalable, hierarchical set of technologies, thoughtfully applied.
We suggest an approach that starts by finding the answers to the questions above, then considering applying (technology) tools that automate manual tasks, then adding on other solutions as needed, from real-time transaction support systems through to assisted service systems (yes, we build them, and believe they meet certain needs). The technologies that look like answers should then be assessed for impact on total cost of operations and the specific expected return on investment in the total scheme Add to that a rational plan for deployment, strong emphasis on branch choreography, and of course team training to ensure that goals of implementation will be met.
Here is a final checklist on technology decisions:
- Keeps my people talking to my customers.
- Pays for its TCO, or better, in an understandable way.
- We can successfully implement it in a sustainable way.
Our experience is that thoughtful application of a suite of technologies, considered as parts of a holistic solution, will move the “value proposition?” answer from “sometimes” to “yes”, without over-investment, without forcing uncomfortable customer change, and with sensitivity to the specific, complex needs of customers walking in to the branch “right now”.
Mike is Executive Vice President, Global Solutions at Glory Global Solutions, and occasionally feels the need to remind that banking is fundamentally a “people to people” business, even though he works for a FinTech company.