In early 2020, no one (least of all the largest commercial banks) expected anything but doom and gloom.
Jamie Dimon, CEO of JPMorgan Chase, noted in his annual shareholder letter, “We don’t know exactly what the future will hold — but at a minimum, we assume that it will include a bad recession combined with some kind of financial stress similar to the global financial crisis of 2008.”
Fortunately for everyone, those dire predictions have so far proved wrong, thanks to the Fed slashing rates to record lows while pumping billions into the economy. Indeed, for all the tumult and stress, banks may someday look back on the COVID era as magical, at least from a profit perspective.
Per the FDIC, over the last 18 months, net profits from all insured institutions topped $421.7 billion – the highest ever over any prior 18-month stretch.
Yet despite this banner performance, risk-averse bankers still see storms looming, with threats seeming ever present:
• “Shadow banks” like hedge funds and money market funds siphoning away trillions in deposits.
• Fintech disruptors steadily eroding customer bases, especially among younger demographics, such as Gen Alpha and Gen Z.
• And the elephant in the room: Rising rates now spurring the first uptick in delinquencies in years.
This rocky terrain has banks laser-focused on identifying and retaining their most valuable customers. However, a challenge emerges: Determining the true profitability of a given customer is rarely straightforward.
While bankers intuitively believe they understand who their most profitable customers are, genuine insight requires looking past superficial metrics to uncover less apparent correlations. Gaining that insight is where advanced customer profitability modeling can prove invaluable, providing a toolkit to cut through the noise and optimize the portfolio.
For years, such sophisticated modeling was feasible only for megabanks, leaving smaller institutions behind and lacking the technical expertise and resources. However, financial technology advances have democratized access to analytics, allowing community banks and credit unions to leverage next-generation platforms and providing a pathway to clarify customer value and transform portfolio management.
Murky Profitability in A Sea of Data
Distilling oceans of data into clear profitability insights remains a struggle – particularly as consumers spread their banking across multiple institutions.
According to research from Rivel, a firm that does wide-scale semi-annual online surveys sampling thousands of bank customers, the share of wallet retained by a consumer’s primary bank has declined 4% in just two years, evidence of rising multi-banking trends. Such fragmentation places even more importance on analytics to retain the most valuable relationships.
Factor in the Pareto Principle, which posits that the lion’s share of bank revenues (~80%) are driven by a small fraction of customers (~20%), and the importance of retaining top-tier customers becomes glaringly evident.
It may seem obvious that keeping your best customers happy should be a priority, but lots of institutions fumble with this fundamental task.
“Many retail or commercial bankers lack clarity on bank profitability, let alone relationship profitability,” says Erin Guthman, SVP of Advisory Services at PCBB, a banker’s bank that provides services to community institutions. “Profitability modeling opens their eyes to who their truly valuable relationships are. It lets banks optimize pricing and marketing efforts toward the customers that can drive the most value.”
“Many retail or commercial bankers lack clarity on bank profitability, let alone relationship profitability.”
— Erin Guthman, PCBB
Profitability modeling is not new, by any means. It was invented in the late 1980s as an outgrowth of credit risk modeling from the 1970s. Management consulting firms like McKinsey, Bain, and BCG promoted proprietary advanced statistical techniques and published foundational thought leadership papers, highlighting tools they created to pinpoint a bank’s most valuable current and potential customers.
Since then, big banks began hiring their own quants with PhDs to sift through mountains of data to calculate each customer’s lifetime value, a projected future revenue less the costs of serving each relationship. The goal is to identify the ideal intersection of revenue generation and cost-efficiency.
Today, the most sophisticated modeling takes a multifaceted view of profitability drivers. Mosaic customer profiles integrate demographic, behavioral, and even psychographic inputs alongside basic transaction data. These data provide a nuanced understanding of the highest-value segments among both existing and prospective clients.
The beauty of profitability modeling is how it reveals hidden truths that defy conventional wisdom. For example, some customers who generate sizable revenue may not actually be highly profitable once service costs are incorporated. In contrast, others who don’t seem to drive a lot of revenue can be highly profitable. With that information in hand, a bank can create a profitability roadmap to guide strategic prioritization. Modeling spotlights where to focus engagement, service, and marketing investments for optimal ROI. Statistical techniques like regression and decision trees strengthen predictive accuracy.
Profitability modeling also enables scenario testing, such as forecasting the impact of boosting deposit rates for target segments. Such modeling platforms incorporate product usage, risk profiles, channel preferences, and revenue/cost data to calculate lifetime value.
But today such analyses encounter a thorny issue: how long is a “lifetime” in bank profitability modeling? Industry benchmarks for retail customer relationships used to be 10 to 15 years, but the number is now a downward-moving target, thanks to changing demographics, increased competition, and ever-changing customer preferences.
Is it worth investing in acquiring and cultivating a relationship with a “whale” customer who might bank with an institution for months instead of years or decades?
Enhancing Profitability Throughout the Customer Journey
Data can answer that question. Fortified with profitability scores and modeling, banks can transform strategy across the customer lifecycle to maximize value. Onboarding presents opportunities to fast-track profitable relationship development through optimal cross-sell offers. New customers with profiles that match a bank’s most valuable relationships can be identified for priority relationship building.
Additionally, profitability metrics are regularly used to inform engagement and retention efforts. The highest-scoring relationships warrant prioritized nurturing to promote loyalty. Proactive outreach to discuss financial goals and recognize loyalty helps sustain critically important relationships. Moreover, profitability data rationalizes pricing exceptions to accommodate and reward top relationships.
Michael Zagorski, Chief Credit Officer at Belmont Bank & Trust in Chicago, a client of PCBB, describes the successful results of his firm’s use of the tools, “It has dramatically improved our loan pricing. We can scientifically determine pricing that is compensated for the risk and cost of different deals.”
- Four Ways Banks Sabotage the Onboarding Process (And How To Fix It)
- Today’s Deposit Battles Demand Personalized Marketing Strategies
The Crucial Role of Data and Technology
But once you’ve identified the individual drivers of profitability, i.e., who are your most valuable customers and how can you retain them, why stop there?
With profitability quantified, more banks are testing scenarios to guide strategic decisions. Changing deposit rate assumptions shows potential bottom-line impact under different interest rate scenarios. Tweaking acquisition spending forecasts profit lift from heightened targeting.
Technology empowers small institutions to benefit from advanced analytics, once accessible only to megabanks. As PCBB’s Erin Guthman says, “Profitability platforms level the playing field. Small community banks and credit unions have gained sophisticated capabilities to optimize their portfolios.”
A Primer for Profitability Modeling
So, what are the steps if an institution hasn’t already decided to invest in profitability modeling and is now ready to take the plunge? Here is a primer for bringing analytics into a customer strategy, step by step:
• Audit available data: Compile sources of customer insights across your organization to assess potential model inputs. Identify any gaps requiring additional data integration.• Build predictive models: Leverage statistical modeling techniques to analyze how customer attributes correlate with profitability. Regression, decision trees, and clustering algorithms provide predictive power.
• Generate customer lifetime value scores: Apply models to your customer base to quantify anticipated lifetime value based on profile attributes and product holdings.
• Identify opportunities: Profitability scores spotlight engagement opportunities such as cross-sell potential. Lower scores prompt exploration of root causes.
• Enable human conversations: Arm bankers with profitability insights to have productive conversations grounded in a data-driven understanding of customer needs.
• Guide strategy with data: Incorporate profitability metrics into acquisition targeting, onboarding, pricing policies, and engagement prioritization.
• Iterate: Continuously enhance model accuracy by validating predictions against emerging results and refining algorithms. Like most things, profitability modeling is a matter of trial and error.
Although the steps look discrete and straightforward enough in theory, data collection and analysis are messy tasks. It is crucial to work with data partners with track records of success. Consider that the largest commercial banks have dedicated large departments of software engineers and data scientists who have devoted years to such efforts and are still working on refinements. Using data to glean actionable insights is significantly more challenging than it looks.
Need A Bank Tech and Data Toolkit? Listen to the webinar.
The Human Side of Profitability
Advanced analytics provide invaluable data-driven insights that can optimize customer portfolio performance. However, an over-reliance on data modeling without accounting for human expertise carries risks.
Recent research by Bain & Company found that consumers increasingly use a mix of financial providers rather than consolidating with a single primary bank. This fragmentation allows traditional banks to focus on more personalized and engaging experiences to build loyalty.
Bain’s analysis shows banks that get the digital experience right by providing seamless, personalized digital services earn higher loyalty scores. However, their research also found loyalty is maximized by blending excellent digital experiences with human expertise for more complex needs.
The critical insight is that human judgment and relationships remain essential, even in a digital world, especially for personalized advice. Bain’s findings underscore the need to combine advanced analytics with experienced bankers’ expertise.
Don't Give Up On What You Know:
Bain research emphasizes that consumers still value human judgement for personalized financial advice — even as they grow more digitally focused.
Further supporting this, Rivel’s consumer research shows that service satisfaction correlates more with retention than factors like higher deposit rates or technology. This finding underscores that no algorithm can replicate the nuanced human judgments cultivated from years of customer interactions. Even with advanced analytics, human expertise and relationships remain essential — especially for personalized advisory services. Profitability scoring augments but does not replace human intuition.
There is No Silver Bullet
Keeping perspective is crucial. If a super-banker like Jamie Dimon, with all his connections and access to Chase’s unlimited economic data, can’t accurately forecast seismic economic shifts, how can we expect profitability modeling to foresee all challenges? The answer is it can’t. There are too many variables. Analytics enable efficiencies, but relationships build affinity.
Technology now allows smaller institutions to compete and grow through advanced modeling once exclusive to megabanks. However, customer loyalty is built with human bonds. Profitability metrics are indeed helpful in identifying opportunities, but bankers on the ground still provide guidance that no algorithm can yet match.
In isolation, modeling becomes a futile exercise. That said, the combination of humans leveraging both machines and vast quantities of data is likely a winning formula. After all, banking is not just about profit; it is fundamentally about trust — a quality that has proven elusive for technology to replicate.