by Balaji Yellavalli, AVP, Banking and Capital Markets
The other day, the CIO of a leading consumer automobile finance company (which also offers other diversified financial services such as credit cards, banking, mortgage loans, etc.) told me that they cannot pull the data on an existing customer, say, a credit card holder, when she approaches the firm with a new transaction, say, a home mortgage or an auto loan. Something we would assume as the most simple or obvious thing that one can do in a post internet, hyper-connected, Flat World! These issues may not manifest as tangible or substantial cost escalations in the short run, but ultimately they are bound to add up and bring down customer profitability.
Why is this concept so difficult to implement, if its benefits are so patently obvious to me as a customer?
I see two main reasons. One, we all know very well: Financial services firms have grown through mergers and acquisitions over the past decade, which have brought "legacy baggage" with them. Business processes, technology infrastructure, organizational politics, turf battles and sometimes even genuine customer privacy concerns have come in the way of sharing data across lines of business.
The other reason is not so apparent: Ironically, it is the evolution of a discipline and function called Customer Relationship Management (CRM) over the past decade or so.
CRM was purveyed as a panacea for integrating all customer touch-points and provide insights into customer behavior. For sure, CRM streamlined customer interactions. The call center or customer contact center, became a ubiquitous part of our daily lives, whether we wanted to add on a new service or to find out why we have been billed erroneously. CRM provided the glue to unify the front-end experience across both these channels and of course any other physical channel, like a bank branch or an ATM.
However, in doing all that, CRM has created for the organization, two new challenges.
One, it generated volumes of transaction data at every customer touch-point, which was often inconsistent with data within another channel or line of business. While data cleansing and data quality solutions addressed some of the problems, there still are far too many exceptions that need to be handled manually and end up not being addressed at all. In other words, while the front-ends of customer interaction points are glued together reasonably well, the middle and back-offices tend to get messy with multiple stacks of data vying for the "single version of truth"!
Second and the more critical challenge: CRM, by definition, is transactional. It tends to be reactive as it "waits" upon a customer to call. When a customer initiates contact (a call, a click or a walk-in), CRM systems, post facto, try and reconcile the true picture of the customer and all dimensions of her relationship with the firm. So, while data analytics and data mining tools have fed off historical CRM transaction data, the foundation or core reference data of a customer tends to remain fuzzy.
The way CRM has evolved, it has ended up being the "cart before the horse" with regard to customer data – i.e., reactive transaction data continuously trying to catch up and build a 360 degree view of the customer.
I’m sure you have experienced this yourself, as a customer and also as a business professional.
I am simplifying the case, but this is where Customer Data Integration (CDI) kicks in – CDI, at the core is the creation and maintenance of a centralized repository of core "master data" – the multiple facets (demographic, household, even updated credit-worthiness data) of information that identify a customer. This data is then "pulled" by various transactional CRM systems whenever a customer "calls" in, so that disparate organizational silos draw from the same master data for the transaction. Once the transaction is completed, if there are any updates to the core, master data, they are checked for consistency and updated in the central data repository and become the master data for the future.
To handle privacy concerns, a well designed CDI system can mask certain "views" of data even internally across lines-of-business. And newer technology developments like Service Oriented Architecture (SoA) ensure that the business process to continuously monitor, maintain and update customers’ master data are agile, flexible and scale with the firm’s growth aspirations.
So the next question is: how painful and challenging is the creation of a CDI? Well, short answer is that it can be quite challenging, but it is a one time exercise as opposed to rediscovering multiple versions of the truth each time there is a customer interaction. And the attendant re-engineering of business processes as well as change management is better handled in a centralized roll out.
CRM worked well when acting reactively to customer contact was acceptable; but in a Flat World, it is necessary to have an accurate view of the entire customer relationship.
The CIO I talked about will hopefully be a happier person, given that he made the decision to adopt the CDI approach to his challenge!
