Leveraging data analytics for long-term growth

The banking landscape has changed dramatically in recent years. One of the most obvious signs of this change is the wave of regulation that has come in the wake of the financial crisis. It is something that the whole industry has to deal with and is a common issue that we discuss with our bank clients, wherever they are in the world. However, when it comes to the broader market landscape, there are noticeable differences in the types of environment our clients are operating in and where in the business cycle they are.

Banking is a cyclical business. If you take a snapshot of banks around the globe, you would see banks at every stage of the cycle.  At a basic level, it is possible to characterise the industry by regional differences. Credit losses have stabilised on retail banking activities, while growth remains low in the UK. At the same time, in the US, credit losses have declined, and lenders are growing their books again for most consumer products.

In the eurozone, the discussions that we have with banks typically focus on how they can minimise their credit losses. In Latin America and Asia, banks are focused on maximising the opportunities of the growing middle class. Within both of these regions, there are of course differences between individual countries. Asia has the whole range, from the established markets of Japan or Singapore, to the high-growth markets of China and Indonesia. And, in Latin America, where we have made significant investments, the bright spots include markets such as Brazil, Colombia, Chile and Mexico.

These markets are interesting as they demonstrate the cyclical nature of banking: 10 years ago the outlook for these markets was generally subdued. Now, the view of these markets is very optimistic. From a long-term perspective, however, the opportunity for banks operating in these growth markets should not simply be about riding the wave of optimism and grabbing market share during the good part of the cycle. Instead, the focus should be on acquiring the right customers and serving them in the right way in order to drive profitable, long-term sustainable growth.

The same approach is also true for banks in more mature markets. Whether an institution is focusing on reducing bad debt and delinquencies, growing customers or increasing share of wallet, the key is staying close to the customer and getting better at understanding their behaviour and what their needs are.

This is where data analytics, or, more precisely, decision analytics come into play. Knowing and understanding customers and being able to act on this information are what make banks resilient to every phase of the business cycle.  And, this is how we help our clients every day.
Customers First

For years, retail banks have designed products and sold these products based on high-level surveys and scarce information about customer needs. Technology, changing consumer expectations, increased competition and other factors mean that these historical techniques are no longer as effective. Banks need to know which customers will be profitable and how to align their strategy to the needs of every single one of their customers. Essentially, this means moving from a top-down to a bottom-up approach – starting with the customer.

The good news is that banks can use the data they hold internally as well as external information, which can be structured (such as credit bureau and consumer classification data) or  unstructured (such as the data coming from the myriad of online and social networking sites that exist today), to understand a customer’s consumption patterns and financial behaviour.  With a detailed financial profile, banks can understand what products fit with the customer’s life stage and socio-economic group, and they can price the product appropriately.

With more information and more knowledge, banks can be more discerning in the types of customers that they target. This kind of discernment is one of the biggest changes that we have seen in the banking industry in recent years. Five to ten years ago, banks typically did a mass advertising campaign, accepted the qualified customers who applied and then would see which customers were profitable over time. Today, the approach is becoming very different. Now banks can identify the types of customer that will be profitable and target them before they do a marketing campaign.  This new thinking is really about offering a service that the customer will use in a way that is useful for the bank and for them. Banks in developed and developing markets alike are beginning to understand that this kind of discernment is important.
Integrated strategy

The most effective banks have in place a repeatable and scalable decisioning approach that combines information, analytics, decisioning and execution.

Banks need to ensure that they have the maximum amount of information on consumers, which, of course, is compliant and within the regulations of the particular jurisdictions where they operate. Also, they need to have up-to-date data – not only the data that reflects what took place three months, six months or a year ago. Simply having a snapshot of the customer at the point of acquisition is not enough on its own, there needs to be a constant refresh of the behaviours of consumers in the bank’s portfolios. Customers expect their banks to know when they are saving for large purchases or that their needs have changed when they take out a new mortgage.

The next stage is using analytical tools that make sense of this data and that can predict behaviour, for example, a customer’s likelihood of paying back a loan or interest in buying a new product.

It is relatively easy to build analytical models, but to do so in a way that really understands the data and the most relevant attributes is very difficult. Many organisations have tremendous academic talent in-house to build analytical models, but they don’t always have the practical expertise to really understand which data is relevant and how to combine various data sets to maximise the understanding of the customer.

Next, the output of these models needs to be translated into something that makes sense for customers. All of the data, statistics and information need to be translated into an end-customer decision. The bank may be able to predict whether a customer is likely to revolve credit card debt, or whether they are likely to pay it back in regular instalments, but the bank has to give a very specific decision on what credit limit they can offer, and at what rate.

The bank has to find the right balance: the limit cannot be so high that the customer becomes over-indebted, and it cannot be so low that the customer’s spending is restricted. Also, banks need to create an environment where they constantly test these parameters and assumptions. For example, they can use a ‘champion-challenger’ model where strategies can be tested in parallel. In our earlier example, a group of customers with the same profile can be divided in two, with the first half used as a control group with the existing credit limits. The other half can be the ‘challenger’ group, which is tested with a higher set of credit limits. If that group of customers performs better and is more successful, then the higher credit limit can be introduced to all of the customers in that bracket.

Finally, banks need to consider the deployment of these customer decisions. A bank may have all the data on the customer, and the ability to define the best decisions based on their profile, but it has to be available at the point where the customer decides to interact with the bank. The end decision – a credit card limit, cross-sell offer or a mortgage rate - needs to reach the customer, whether they have walked into a branch, or are on the phone to the call centre, or are connecting on-line. The decision needs to be available at that particular point in time – when the bank interacts with the customer – and take into account all the information available in real-time, but this is easier said than done.

Technology is critical in making all this happen, but this is not as complex as many would believe. It is really about being smart about the way in which we look at the information already available. Often the changes that are necessary are not about doing a massive overhaul of current IT systems – it is more about banks adding intelligent automated decisioning capabilities to what they already do.

It won’t be long before consumers across the globe expect their bank to really understand their needs and to offer services that are relevant to them. The banks that are able to develop and deploy their strategies based on the granular knowledge of the customers will be the ones that will reap the most benefits.