By Mike Wallberg, CFA | 30 March 2017
A few years ago, businesses of every stripe suddenly got very excited about the idea of “big data” but few industries possessed the raw source data – and therefore upside potential – of banking. Locked in vaults (virtual and otherwise) live the details of your behavioral profile including account transactions, deposit balances, product breadth, repayment history, referrals, and more. Banks (in theory) know what rates you’ve been paid, and how you responded to offers. They have records of how you like to access your account, what it costs them, and whether or not you own a car or home.
But like the green codes that stream across the screen on “The Matrix,” one first needs to be able to translate the 0s and 1s – to turn “data” into recognizable “information.” So you can recognize and be able act on things like attrition triggers, price rate sensitivities, and demand for individual or bundled products, both internal to the bank or provided by third parties.
Recognizing the very high cost of new customer acquisition – pegged at as much as eight to ten times that of cross-selling – many banks have taken an initial step to tame and translate the seemingly untameable by implementing “descriptive analytics” for their captive clients.
Source: Warner Bros.
“There is much to see in banking data if you have the right tools to view it,” says Dinesh Krishnan.
However, Dinesh Krishnan, Head of Research & Development at Zafin, suggests that these will only prove to be the forebear to the truly powerful tools – relationship analytics – that turn the data into recommendations.
“Descriptive analytics is strictly reporting: Do I know who the customer is? Do I know what their products and their preferences are? This has been largely solved by mid-sized banks and larger by using third-party tools or combing their database with some sort of basic Excel solution,” he says. “However, when you look at profitability, that is the point at which it becomes grey. Profitability may be known at the segment level, but I’m not sure they can do it profitably at an individual customer level.”
The company’s Chief Analytics Officer, Suman Singh, explains that relationship analytics take the descriptive data and make it useful by interpreting it and turning it into recommendations banks can actually use to run their business.
“It’s about creating advanced machine learning algorithms and predictive models to develop customer-specific recommendations. That’s really where relationship analytics differentiates itself from existing dashboards and reports,” he says. “We’re talking about being able to look at a customer’s holdings and ask, ‘What is it that this customer needs? What is the propensity for him to buy, say, a mortgage? Or a line of credit? What is the next best product that the customer has the propensity to add?”
Such analytics could be delivered at both the product manager level and the relationship manager level.
And product or portfolio managers tasked with developing “one-to-many” marketing strategies can meet their twin goals of maximizing both reach and profitability of product offers. Singh sees a good relationship analytics engine as these managers’ new best friend as it can help them direct specific product offers or bundles to specific screened groups or segments based on their propensity to buy.
“By generating actionable insights from the tool, banks can push the right offers to right customers,” he says. “They can download a list of customers who meet particular criteria, and send that to their campaign management tool to reach out with the offer, and then follow up on multiple channels.”
Product managers: An example
You’re a regional manager for a large bank and you want to grow your mortgage lending book. You load the relationship analytics tool and screen on customers within your catchment that have the highest propensity to add a mortgage as their next banking product. Using a data-driven, machine-learning algorithm, it returns a list of 200 clients. You load the clients into your campaign management tool and launch a campaign targeting just those 200 people.
Relationship Managers and the Customer Relationship Score
At the teller or private banker level, where the bank has its most direct contact (and opportunity to sell), relationship analytics can help manage both day-to-day relationship building and support business development.
One tool that Singh’s team has created to do that is the “Customer Relationship Score,” something akin to a FICO credit score. “It is a pattern matching engine that looks at the criteria or drivers – How are they behaving? What transactions are they making? – that all coalesce into a single score reflecting the relationship strength of each client,” he says. “We have created our own composite metric to measure the customer’s stickiness. Fundamentally, it allows the bank to look systematically at customer experience from a strategic perspective of value and opportunity, and morph its interaction model to maximize customer satisfaction.”
He cautions that a low score doesn’t mean a customer is a bad customer, per se. It means there is an opportunity for the bank to help the customer improve their score by changing their transaction behaviours and adding key products. And in an industry where customer expectations for personalization is skyrocketing, the transparency and personal touch can only help. Figure 1 explores the Customer Relationship Score in more detail.
Imagine you’re working in a branch and a customer comes in complaining of a fee error. You go into your system and pull up a profile that shows her history with the bank, including the detail of what accounts she has, what her transaction patterns are, what loans she has and so on. But more than that, the system boils that summary data down into a single Customer Relationship Score (CRS) that represents the value of that customer to the bank. Seeing the customer scores above 800 (good), you reverse the small charge without issue.
Figure 1: Customer Relationship Score
It’s what happens next that gets Singh really excited. The relationship analytics system reads the history and recommends the next product for you to offer the customer – at a price that reflects the customer’s CRS. The customer has a line of credit in good standing, high balances, no overdraft history, and no vehicle loan. So you’re prompted to suggest an auto loan, which you can extend at a preferential rate (calculated and delivered on the fly) given the customer’s high value to the bank.
So what just happened? The bank identified a valuable customer, took care of his original need with commensurate service, identified the product with the highest propensity of take-up to offer him, and offered it to him with pricing that reflected the value of the customer while protecting the bank’s margins.
“It is a game changer”
Krishnan sees this level of granularity in marketing and product management as foundational for banks going forward.
“The banks that have adopted this journey – they are the ones who will be able to be very targeted in their offers to customers with specific products,” he says. “And their conversion rate for each on-boarded customer will go up, lowering the cost of acquisition. Net fee and net interest income will rise. And the banks who haven’t developed a strategy to compete will come under pressure to match that growth.”
For more information about Zafin’s Relationship Analytics, click here to download our eBook!