For a good decade or more, when it came to knowing the customer, the banking spirit was willing but the technology was weak. Now, new strengths in analytics make it possible for banks to rise to their long-held aspirations and develop customer relationships of sustainable value.
It was inevitable that breakthroughs years ago in data management would create high expectations, perhaps too high. Vast data warehouses promised banks the ability to create “customer segments of one,” enabling them to tailor products, prices and marketing strategies accordingly and boost cross-sell for greater profit.
But with some notable exceptions, industry cross-sell still hovers down around 3.5 products. Even among the exceptions, profits and loyalty are questionable if customers succumb to solicitations for products they don’t need. And all of us are familiar with the anecdotal evidence of segmentation schemes gone awry, like the childless seniors pitched life insurance policies. Or home equity loans marketed to the deeply indebted. Or the direct mail sent to customers who interact mainly via computers and cell phones.
The main problem has been the sheer volume of customer data and the high cost and difficulty of assembling that data rapidly and precisely enough so that banks could act on it and measure the results. The mushrooming of customer privacy and data protection laws didn’t help. But now the promised era seems close at hand, thanks to recent developments in analytics.
To begin with, the recent economic downturn took many once profitable customers out of that desirable category and left banks competing harder for a smaller population, basically forcing them to focus more on segmentation analytics. It also spurred regulators to limit bank fees and customers to eliminate some accounts and avoid fees.
The upshot is that banks, now in fiercer competition for a smaller population of high-value customers, are seeking positive ways to increase customer engagement, loyalty and profits. That strategy, however, requires precise handling of enormous data, such as knowing (not guessing) what products which customers will value enough to willingly accept the bank’s price as a fair and positive value exchange; knowing how best to offer those products to each targeted customer; and anticipating correctly how customers will react to different products, prices and sales tactics through different channels.
That’s a lot to know, and for years, despite all the information available in the data warehouse, it remained largely unknown. The tools didn’t exist to organize the data and turn it into knowledge and then embed that knowledge into the bank’s interactions with customers. Existing tools took months to deploy while the data they depended on decayed in reliability. Support costs for the tools and data infrastructure, meanwhile, ran into the millions of dollars.
Enter the new era of analytics. Now, in cloud-based processes that take weeks instead of months or years, no infrastructure or equipment investment and minimal Information Technology (IT) support, banks can use analytics that score customers. These scores mimic FICO in their detail by presenting a scale of 0 to 999 rather than just a simple “profitable vs. unprofitable” option. They also indicate the current value of the customer to the institution.
Because current value is a significant but insufficient indicator of potential value, the analytics further segment customers on a combination of their financial contribution and “persistence.” Persistence captures the customer’s level of engagement with the institution – not just product usage but all transactions and interactions, financial and non-financial. It reveals what has often been overlooked by analytics, namely that a customer relationship is not defined solely by a “buy” or “don’t buy” decision but also by interaction at every touch point, such as contacting the call center, swiping a debit card, accessing a safety deposit box or visiting the bank’s website.
Such detail makes it possible to map every customer to a grid which the x axis ranks by Persistence and the y axis by Contribution. The grid itself can categorize customers in discrete enough segments so that the right sales tactics are easier to divine and the bank doesn’t waste resources on tactics that will obviously fail with certain segments – either fail to attract them or fail to earn a return on their business.
For example, it makes apparent that targeting Takers, who rank high on persistence but low on contribution, in a campaign involving costly branch resources, is a bad idea. But High Performers, who rank high on both counts, are worth the strenuous efforts to reward and retain. In between are Advocates who, though highly profitable, show lower persistence; the right campaign for them is one that includes a special incentive for them to act.
But even with this richer level of insight, campaigns are not perfect, customers are not static and competitors are not asleep. It is not enough for analytic tools to deliver intelligence. They must also, in real time, measure the impact of the campaigns on each targeted customer, capture that information and update customer scores, and perhaps reassign them to a different segment. Captured information from each campaign is immediately reworked into the analytics and subsequent customer experience.
That’s what “knowing the customer” is all about. It’s about competing in the same real-time environment in which customers and competitors make decisions and spend money.