Is Banking Really Ready For Big Data?

Does "big data" represent new profit opportunities for retail financial institutions, or are they just chasing rainbows? If bank and credit union marketers don't learn to master advanced data analytics, then this red hot trend could be reduced to little more than just another frustrating buzzword.

The range of analytic marketing competencies among financial institutions is wider than in most other functions. With the introduction of “big data” — i.e. data assets that exceed current internal database capabilities — this gap has only widened. If we were talking golf, only a few would have a scratch handicap while most would be taking 120 strokes and losing 10 balls a round.

Analytic insights into consumers and small businesses can dramatically increase the payoff from marketing and sales. And yet few banks are achieving their potential, even though there’s no longer any need to wait for a comprehensive data warehouse. Most are not capitalizing on new data sources and advanced measurement approaches, choosing instead to continue with outdated analytic approaches that may seem sensible but yield poor returns.

Why can some banks achieve over 1,000% return on investment and double their revenue growth rate within a matter of mere months, while others who use the same marketing channels can wind up with negative returns? Our firm has invested over 30 years figuring out how to help banks shoot “scratch golf.” Here’s a myth/reality checklist to help financial marketing executives identify opportunities lurking in their data analytics programs.

Myth: “Using ZIP+4 level targeting is effective at identifying high priority leads.”

Reality: The presumption that all households within a ZIP+4 region have similar needs, wallet sizes, channel preferences, etc. is significantly flawed. Tools that provide insights at the individual- or household level are now available, and should be a new basic requirement for targeted marketing in retail banking.

Myth: “Defining broad segments — by annual income or funds on deposit — is sufficient.”

Reality: Simplistic segmentation schemes don’t cut it anymore, and should be abandoned. Two households may share a similar profile relative to deposits, but might be totally different with respect to home equity, credit cards, prepaid debit, etc. Such schemes should be replaced with segmentation at the product level in order to obtain a much more accurate prediction of needs, attitudes, and buying behaviors. It may sound challenging, but the lift in ROI is well worth it.

Myth: “Third-party data runs afoul of regulatory guidelines.”

Reality: Discard all analytical tools that no longer pass OCC and  CFPB audits because they use variables
that are not compliant with current guidelines from the FCRA, ECOA, and other consumer protection laws — e.g., variables that correlate with ethnicity, race, gender, age, religion, national origin, marital status, and presence of children.  There are alternative sources of equally powerful, compliant data.

Myth: “If there’s any data worth using, our organization can get (or find) it on our own.”

Reality: Internal data does not predict prospect behavior, and an emphasis on look-alike analyses of variables like product usage and channel behaviors is suboptimal. You need to find the right balance between data derived internally and externally. Solutions that do not make use of third-party data elements provide an incomplete, less potent view of customers. Third-party data can be used identify customers’ financial triggers, pinpointing those who might be new to a geographic area or in the market for a particular financial product. Data elements are available that provide balances held at all financial institutions down to the household-level and across a variety of categories — deposits, investable assets, investment balances, net assets, mortgage balances, etc. Banks utilizing this kind of third-party data can quantify their share of wallet and target the right households to cross-sell.

Some financial institutions now find it possible to move faster and more efficiently by utilizing external data exclusively in many of their marketing activities. Others add internal data elements selectively and continually test whether they are powerful or just add unnecessary cost and complexity.

Myth: “If response rates increase, the campaign is successful.”

Reality: Most direct marketers measure response rate, then calculate cost per account acquired. However,
too many firms erroneously equate high response rate with “success.”  Conventional campaign metrics —
e.g., gross responses, number of accounts opened, cost per account opened, etc. — are problematic
because they don’t often correlate with shareholder value. Campaign performance should be
measured using concepts like Annual Net Income Contribution, Lifetime Value (LTV), Return on Marketing
Investment (ROMI), and Payback Period.  Additionally, campaign performance must be measured
net-of-control, i.e., incremental response and account funding above that of a non-stimulated control group structured to mimic the target population.

Myth: “Shareholder value correlates directly with the number of accounts acquired.”

Reality: You have to assess the quality of the accounts that are opened or reactivated. Many banks only count widgets; but clearly a campaign will be more (or less) profitable depending on the level of deposit and loan balances originated. For instance, data analyses reveal customers with larger deposit and loan balances have longer lifespans and therefore higher lifetime values.

Myth: “Only products featured in a campaign should be attributed to the campaign’s results.”

Reality: Ensure that you are attributing all accounts stimulated by the marketing campaign. Often there is a “siloed” perspective on results — e.g., a small business checking campaign team fails to account for other personal accounts opened by the owner, which are often significant (number of accounts opened and their size).

Myth: “Who needs address-matching technology when brute force comparisons catch almost all matches?”

Reality: Consider whether you need to upgrade the algorithms that match campaign responders to a target file. Because most matching systems use rigid and simplistic logic rules, they lack the sophistication to identify inexact matches and therefore undercount campaign responders.

Myth: “All individuals within the bank’s footprint should be targeted.”

Reality: Defining your target audience as “all people ages 18-55” isn’t really a strategy. It’s essentially the same thing as saying, “We’re targeting everyone with money and a pulse.” That isn’t a targeted market segmentation strategy. Focus on high priority modeled leads. Consider carefully whether to target and measure the campaign at the individual- versus household level. For example, a consumer credit card tends to be an individual decision, while a consumer mortgage is a household decision.

Key Takeaways

1.) Financial marketers must be continually refining their analytic tools. Maintaining best-in-class data analytics requires big strides over a few months, not small steps that take years. The data landscape is evolving fast, and it takes a focused discipline if you’re going to stay a few strokes ahead of the competition.

2.) Refine the selection and accuracy of measurements that determine if marketing campaigns are creating shareholder value.

3.) Upgrade any outdated, inappropriate or suboptimal data analysis tools. Forward momentum can only occur if new “big data” hypotheses can be tested with the latest, best data assets and analytical tools available.

Embracing an advanced marketing analytics strategy can yield a return on marketing investment (ROMI) as high as 300%+ on consumer deposits, 500%+ ROMI on mortgage offers, up to 1,000%+ ROMI  for small business deposits and loans. That’s been our experience at least.

This article was originally published on . All content © 2024 by The Financial Brand and may not be reproduced by any means without permission.