Augmented Intelligence
Nov 13, 2017

Big data analytics: The cornerstone of omni-channel customer experience in retail banking

Retail banks today are undergoing massive transformation to enable omni-channel banking. At the outset, this might seem like an IT transformation that involves changing the underlying architecture to support the omni-channel process. However, the true potential of such an environment can only be realized if robust analytics frameworks are part of that platform.

The omni-channel experience: What does it mean for customers?

By definition, implementing an omni-channel customer experience should enable a customer to seamlessly navigate between channels while encountering a consistent experience. In other words, customers can initiate an interaction in one channel and continue in another channel without losing the context. In figure 1, Customer “A" had registered for auto-pay but receives a notification in his mobile that his credit card bill is past due. 

Figure 1: Customers interact seamlessly between channels

Here, the customer moves seamlessly between channels, getting support from a Virtual Assistant (VA) and completing the payment. This proactively saves the customer from incurring a late-payment penalty and enhances customer satisfaction.

One Gallup survey on banking preferences revealed that over 60% of customers use more than three banking channels.

Or consider some of the key findings from Dimension Data's 2016 Global Contact Centre Benchmarking Report, © Dimension Data 2013-2016:

  • Digital channels now account for 42% of all contact center interactions
  • 80% of organizations now view customer experience as a competitive differentiator

What does this mean for banks?

The volume of customer interactions will increase and get more complex due to more touch points. Omni-channel banking will help banks facilitate this for their customers through new digital architecture. For banks, the goals are twofold:

  1. Enhance the customer experience
  2. Improve operational effectiveness through continuous process improvement

But how can banks ensure that these transformation programs deliver on these goals?

Big data analytics hold the key to unlocking the potential of such complex operations. Enormous volumes of data are generated in the omni-channel platform. But identifying the right insights from this data gold mine—and doing so cost effectively—will lead to higher ROI and return-on-data from such transformation programs.

Let's illustrate this with a simple example:

Customer A tries to pay the credit card bill through a mobile app and fails, but subsequently retries using the web and succeeds. By enabling an omni-channel customer experience, the customer is able to move seamlessly from mobile app to web. But how can we measure the crossover rate? After all, there are millions of transactions happening every second. How can we differentiate between a customer switching channels due to channel failure, and a customer trying to start a new transaction in a different channel?

Figure 2: Omni-channel enablement illustration: Customer switches between two channels

First, each transaction should be uniquely tagged by customer ID, timestamp, workflow activity, and channel. Then, we'll need to specify a time threshold for defining a workflow activity as success, failure, or abandonment. From there, we can define failure rates and channel-switching rates. Now, if the digital architecture wasn't designed to enable the above aspects there is no way one can identify whether there has been a channel breakage or not.

Second, the infrastructure should be equipped to handle big data. The introduction of multiple channels, as well as their interconnections, is going to generate a huge amount of data. And there will be a mix of structured and unstructured data. For instance, there will be clickstream data from web channels and speech transcripts from support calls. These raw datasets will need to be cleaned up and transformed before analytics models can be applied.

Figure 3: Big data architecture

Third, the big data architecture should be used to generate actionable business insights by applying analytics frameworks. In the example illustrated in figure 2, a KPI framework should define the success criteria for an omni-channel environment. Additionally, these frameworks should synthesize the insights from structured and unstructured data to identify process inefficiencies and points at which significant customer effort was required. Analytical models can use machine-learning algorithms to predict customer behavior and promote proactive customer engagement.

In conclusion, for an organization investing in omni-channel transformation, we recommend following design steps (figure 4) below to ensure an effective implementation with higher ROI.

Figure 4: Design steps for an effective implementation of omni-channel transformation

About the author

Mohan Raj

Mohan Raj

Manager, Customer Analytics, Genpact

Follow Mohan Raj on LinkedIn