BIG DATA, OMNI-CHANNEL AND TESTING: TURNING BUZZWORDS INTO PROFITS

By Will Weidman, Senior Vice President at Applied Predictive Technologies

It’s nearly 2015 and omni-channel is still at the centre of attention in the financial services world. At this point, it’s not a question of ‘should we develop an omni-channel strategy?’ but instead, ‘how should we optimise and refine our omni-channel strategies going forward?’

In this new environment, financial services firms should be focusing on two main objectives, the first of which is to driverevenue through non-branch channels. The mobile and digital revolution is often viewed as a necessary adjustment, due to changing consumer preferences, but it has also proven to be a significant source of revenue,if a bank leverages these channels fully.We have seen innovative banks try a variety of omni-channelstrategies to drive new revenue streams, including mobile/online cross-sell campaigns, fees for new digital services (e.g., mobile deposit availability), mobile onboarding, video engagement campaigns, and more.

Second, banks should focus on making smart investments in new technology to reduce costs without hurting growth potential or customer satisfaction.According to NCR, it costs approximately£2 less for a bank to fulfill an ATM transaction than a teller transaction.  To take advantage of the lower cost of new channels, retail banks are experimenting withinnovative approaches such as new ATMs with enhanced capabilities, iPads or self-service kiosks in the branch, new branch staffing models (for instance, universal bankers), and monetary incentives for customers to transact in non-branch channels.  To succeed, banks need to identify which investments reduce cost while not limiting their ability to grow and retain relationships.

Unfortunately, it can be very difficult to determine exactly which tactics will work to profitably achieve our two main objectives, as customers are transacting in a variety of complex manners across multiple channels. Let’s take an example: a major bank that executes hundreds of acquisition and cross-sell campaigns per year needs to determine the optimal mix of direct mail, email and mobile cross-sell offers. Many executives still believe that direct mail is very effective, while others believe that the mobile channel is just as effective, or that the cost-savings of mobile deployment will make up for any decreases in revenue caused by cuts to direct mail. The bank has tried a few mixes of outreach in the past year, but has not seen a real impact on KPIs as a result, due to actions taken simultaneously by competitors, other programs the bank is implementing in various channels, and general “noise” inperformance data. Their executives are wondering, how did customer behaviour change due to the shifted outreach strategy? Can we simply eliminate direct mail all together? How are other channels besides mobile/online benefitting from the digital outreach?

The fact of the matter is, as customers interact across more channels and banks increase the volume, frequency and variety of communications in tandem, the analytical environment simply becomes too noisy to tease out any real answers using correlations or modelling tactics.

The proven approach to accurately answering such questions in an omni-channel environment is to use a “test and learn” methodology toconduct a variety of lean, agile tests with relatively small subsets of customers,markets or branches and subsequently measuring the incremental impact of each action on key metrics.For instance, the bankin our example could switch a subset of customers to digital outreach, and compare the sales performance of those customers to the ones who continued receiving direct mail as before, therefore enabling the bank to understand the true financial impact of the new approach. Additionally, a more holistic answer could be achieved by dividing customers into multiple test “cells” and then varying the mix of communications in each cell, which would allow the bank to understand exactly which mix is most profitable for each type of customer.

Some of the most successful financial services institutions in the world are applying a test and learn methodology to drive revenues and strategically lower costs through a variety of innovative new programs. These firms are using their available data to the fullest extent to better understand the impact of these ideas: in addition to financial success metrics (account generation, monthly balance, AUM, revenue, etc.), companies using streamlined data models are able to integrate new data feeds, such as mobile app logins, online deposits/transfers and ATM usage, into test analysis. Further, by identifying significant drivers of performance (e.g. demographic trends, branch characteristics, customer relationship history) and targeting initiatives to customers or areas that fit such criteria, banks can constantly refine their approach and improve the ROI of each program.

We have seen this approach driving tens of millions of pounds annually for some of the largest companies in the world. Organisations can build out a funnel of tests across channels, functional areas and divisions to optimise ROI on nearly every decision made. The most sophisticated testers use learnings from each test to revisestrategy and generate new testable ideas going forward, creating a continuouscycle. Conveniently, this model works extremely well in the ever-changing omni-channel environment, where there are new ideas and concepts to test each day. Additionally, given that many omni-channel ideas are extremely risky due to the potential of customer attrition, large capital investments, or product cannibalisation, testing has never been more essential to business strategy.

The new omni-channel environment is complex and rapidly changing, so banks need a robust method for making data-driven decisions about omni-channel initiatives. In-markettesting is the only way to gain an accurate understanding of the success of each omni-channel action to provide decision-makers with one version of the truth.

Comments are closed