Man’s desire to know what the future holds is nothing new or out of the ordinary.
7th century BC: In ancient times, such a human need can be first seen and best exhibited by exploring methods the ancient Greeks adopted to answer their questions for the future. Historian Herodotus mentions in his work at least 18 temples (e.g Delphi) having a shrine providing prophecies for public and private affairs. All of the prophecies were vague but in some cases they turned to be true making people believers. Several stories show the confidence that people of that era had in prophecies. One instance where Delphi’s prophecy proved to be authentic was in 480 BC before the battle of Thermopylae, when king Xerxes and his Persian army were plotting against ancient Greece. The Spartans consulted Pythia (Delphi’s priestess) regarding the outcome of the battle. She indicated they were doomed but also prompted them to hear their fate. Doing so all Spartans lost their lives, but created a piece of history and gained immortal fame. It sounds like Pythia was right, doesn’t it?
21st century AD: Using data, numbers, technology and statistics we have now evolved from prophecies and indefinite speculations to scientific data-based forecasting with the use of analytics. Analytics is also well-established within business helping organisations to improve business performance. Data contains the history of your organisation and analytics is definitely trying to tell you something.
Analytics consists of four pillars as introduced by Gartner’s analytics maturity model. To make these analytics pillars sound familiar and applicable to your financial institution, we can give you lots of intellectual or simple (but not simplistic) day-to-day examples. These four pillars indicate how your organization can go up the maturity curve in leveraging analytics in your business.
Let’s take up a scenario where your organization suffers from customer attrition, you can use the analytics pillars to:
- Pillar I, Descriptive Analytics: As a starting point, you can measure the attrition rate and quantify your losses. You can realise the magnitude of the problem and prioritise it among other business pains you may have to deal with and give it the proper attention.
- Pillar II, Diagnostic Analytics: This pillar will help you examine a complex topic and decompose it into smaller parts that can be better understood. You can diagnose the most significant dimensions of customer attrition or the parameters that gave it a rise. Such dimensions can be geographic regions, branch ids, the time period of the year, channel types, product types, or customer segments where you observe the highest attrition rate. Using these observations you can learn a useful lesson from the past and you can then apply a relative corrective action in the future. But still, this pillar allows you to take re-active actions, after having first suffered the losses.
- Pillar III, Predictive Analytics: Analytical predictions won’t produce an unquestionable future statement, but they will help you arrive at what is most likely to happen based on previously observed and statistically validated patterns. This is feasible by exploring and discovering hidden correlations between data, uncovering customer behavioural patterns, market trends, highlighting sequences of events that can lead as a domino effect to what you ‘re trying to predict, or by applying statistical modelling processes like Logistic Regression, Decision Trees etc.
Predictive behavioural models can calculate the probability (0% to 100%) of each of your existing customers to leave your brand in the next few months, so as to take pro-active actions before it is too late. You can also define a threshold of churn-pressure score which is considered to be high enough to trigger your actions and direct your retention campaigns to, supplying you with a sufficient window of opportunity to retain the customers at churn-risk.
- Pillar IV, Prescriptive Analytics: This pillar will help your organisation to perform and explore numerous business simulations and assess the anticipated outcome of a certain business scenario you are exploring to apply, before you actually decide to deliver it to market, e.g.
- How much can you expect to decrease the churn rate if you manage to migrate customers from a traditional channel to a digital one?
- What is the promotional offer and incentive that can make a customer at churn-risk stay?
- What is the Next Best Offer you can make to each individual customer?
- Can services or product personalisation help you retain more customers? How much?
The intelligence derived as outcome of analytics is the right piece of technology to arm your business with. The four analytics pillars can be used to help financial institutions respond to various business challenges, such as customer attrition, customer acquisition, cross-sell, up-sell, customer lifetime value, asset utilization, non-performing assets, fraud-risk, credit-risk, default-risk, reputational-risk, market-risk, performance management and many others.
Analytics can lead to improved operational efficiency, better customer service, more effective marketing, competitive advantages over rival organizations and better P&L statements for your organisation. The objective is always to improve the business by gaining knowledge which can be used to guide decision making, suggest changes, make improvements, recommend next best actions, or even exploit analytics for innovation.
Financial institutions can also expand the use of analytics to benefit not only their organizations but also their own customers, ecosystem partners, or their customers’ customers. In this way financial institutions can become more customer-oriented, create better services and act like truly trusted partners of their customers helping them grow.
While banks understand the importance of analytics, many of them struggle to realize meaningful returns as the initial investment costs are significant. One way to counter this would be to leverage advanced analytics solutions that use banking data models, new-age open-source technologies and built-in intellectual property to rapidly develop actionable insights. This way can make a difference and help you realize high returns on small analytics investments, while discovering meaningful customer behavioural patterns collecting and integrating data from structured data sources (bank’s internal systems), semi-structured sources (ATMs and corporate website navigation logs), or unstructured sources (data coming from social media or other news feeds, machine sensors IoT).
Finally, could we symbolically claim that Pythia was operating in analytics Pillar III? We can definitively spot 3 pillars in the pic below, the Oracle of Delphi…