The Enemy of Digital Transformation in FSIs: Regulators
Also in this issue: Can Customer Centricity Overcome Profit Motive and AI Limitations?
The Enemy of Digital Transformation in FSIs: Regulators
Like any other service, winning in the finance or insurance industry is based on identifying residual pain points among target clients that are significant enough so that fixing them would make acquisition less expensive while increasing retention. Digital transformation speeds up this process through a more effective operating model and tools/analytics for the identification, confirmation, and testing of pain points, and then scaling solutions to address them. The most digitally advanced FSIs could translate this advantage into better service and pricing, eventually winning the market while customers are increasingly delighted.
FSI regulators in the US do not favor the market’s “invisible hand” because this means they wouldn’t need to exist to protect consumers. Surely, it would be great if they could prevent traditional FSIs and fintechs from misleading and defrauding consumers, but that requires too much hands-on work. That is what journalists are for, whether it is the LA Times investigating Wells Fargo’s phantom sales in 2013 or CoinDesk publicizing FTX’s cozy relationship with Alamadea in 2022. Plus, even if the regulator discovers some fraud later on, pursuing it could make them look like they were asleep at the wheel.
What regulators prefer instead is restricting the supply of financial services and insurance offerings by constraining companies around products, prices, and targeting, which severely limits the upside of digital transformation. That approach is much easier and guarantees successful outcomes. The more rules and procedures required from FSIs, the more likely they are to violate them, which leads to fines and regulatory press releases.
To be fair to FSI regulators, their peers in other industries are not fundamentally different. The most hilarious but also sad example is in the automotive industry. Tesla cars have fewer accidents per mile and receive remote software upgrades—what an amazing innovation that regulators should champion and learn from, right? Instead, they are calling virtual software updates an old-fashioned “recall” and requiring Tesla to send customers physical letters.
Protecting or Harming Vulnerable Groups?
The supply impediments by regulators have been ongoing for decades. Frank Rotman, who was Chief Credit Officer at Capital One in the late 90s (before this role officially existed), remembers in a recent interview how they couldn’t use valuable inputs such as an applicant’s profession in their underwriting models.
The reason for this is because certain professions are disproportionately represented among vulnerable minorities. The regulator’s underlying thinking is that avoiding model inputs that are not proportional to the general population would ensure that financial services would be distributed equally across all groups.
For the same reason, some US states don’t allow the use of credit score data for underwriting property insurance. Lower credit scores are not equally distributed across races, so by not using them, property and casualty (P&C) carriers offer everyone the same terms, right? Imagine getting married when you can’t see your partner and can only ask questions that are equally represented in the country's population. Pretty sure the technical term for this is “crapshoot”. There would be much fewer marriages if the participants-to-be could help it, and FSIs' resulting behavior is not much different:
Using less data → Worse predictive model → Higher risk → Avoiding broader swaths of riskier groups altogether.
In other words, instead of empowering FSIs to supply vulnerable groups with a pricier solution, so they could then compete to improve pricing based on more data, regulators make FSIs and fintechs avoid high-risk groups of all ethnicities and sexes altogether. What is the point of digital transformation if the focus is on using less information and just following regulatory procedures?
You could say, “Great, those discarded segments represent tens of millions of potential customers, if traditional FSIs don’t want to engage them without a few data points, fintechs and insurtechs will; they are much better at data analytics anyway.” They are not - startups have both less data and less advanced models. Many startups have tried targeting specific race- or sex-based groups - some went bankrupt and others are struggling to grow. Winning a share of those groups but only in a high-risk sub-segment would be even harder.
One insurtech, Root, was so determined to help the vulnerable groups that it voluntarily stopped using credit scores for setting auto insurance rates, but it did not help to capture that market. Root’s losses and expenses almost tripled vs. the industry average by 2022, and it has been trying to survive.
No Insurance For You
Poor consumers are particularly vulnerable when it comes to obtaining property insurance. They often lack the financial resources to repair or replace a damaged property. This predicament is especially dire in areas prone to flooding and fires. But why should this be a challenge in the U.S., with hundreds of P&C carriers possessing decades of data on risks and pricing strategies? While some states may prohibit the use of credit scores, couldn't carriers simply adjust prices to accommodate the higher, unpredictable risks?
The reality is that they can't, as regulators determine price increases. Say what you will about the Soviet Union and the efficacy of its planned economy, but at least growing up there, we had just one nationwide agency to set prices (GosKomtSen). In contrast, every state in the US has its commissioner who decides its price control policy:
And what motivates the commissioner’s decision? Perhaps he analyzed competitive forces in his state, assessed the digital capabilities of carriers to price risks better, and concluded that they are not utilizing the best data and analytics for underwriting risk analysis. But of course not:
"I'd like to see a zero (percent increase). But I'm willing to listen, if they want to come back with numbers that are more reasonable…. The people have trusted me to fight for lower cost for consumers since 2017. I have done just that and will continue to do so as your Insurance Commissioner.”
Perhaps insurance companies stay, or maybe they leave the state, but winning the next election is more important. However, insurance companies do leave those states. Instead of transforming their digital capabilities to better price risks in flood or fire-prone areas. It is easier to avoid that pain and engage in less risky business.
Regulatory Life Hack
This newsletter might seem unfair to regulators. After all, how could they understand the data and analytics capabilities of FSIs if they don’t even know how to conduct basic vetting of their C-suite executives? The Information recently uncovered two such faux pas:
We recently revealed the OCC's first Chief Fintech Officer fabricated his work history. Today, we have published a deep dive on the FDIC's first Chief Innovation Officer, based on dozens of interviews and documents.
Instead of digging through complex underwriting models, there's a simple regulatory life hack: insure everyone at low prices. This way, consumers keep acquiring property in the riskiest flood and fire-prone areas, and government-offered insurance covers its loss every time.
A similar life hack is applied to government-issued student loans. The Department of Education and other stakeholders know the earning potential of different degree majors, but they don't want private lenders to use that data because it's not equally distributed across races and sexes. This means private lenders, including fintech companies, target graduate degrees in prestigious schools, while the government takes care of the rest. Add to that student debt forgiveness, and you get college degrees enjoyed by 40% of Americans instead of 10% in 1970, while colleges charge 3X the rate of inflation. Win-win, and without tedious digital transformation.
Again, to be fair to FSI regulators, they are not more inept than their peers in other sectors. For example, by limiting the supply of alternatives and discouraging innovation, school regulations created an environment in urban centers where spending 5X per student has no impact on the outcome, resulting in zero math proficiency for everyone:
What this means for FSI digital transformation is an additional significant constraint. Since the end purpose of digital evolution is P&L impact, heavily regulated solutions might be an area where continuing investing in analytics and user experience is a wasteful affair. Instead, prioritize your FSI’s transformation activities in regions and capabilities where regulators don’t intend to score political or ideological points.
Can Customer Centricity Overcome Profit Motive and AI Limitations?
Even if we imagine that regulatory bureaucracy is gone, and FSIs and fintechs are completely free to develop the best business models, would customer centricity be a major factor? That might seem like a silly question with the obvious response “yes”, but it is because your definition of customer centricity is a decade old. Let’s level up.
1990s to Today: The History of Customer-Centric Evolution
Financial services and insurance companies have been becoming more customer-centric for decades. Openly discussing how to sell clients “dogshit products” for a 5% commission and a trip to the Bahamas was the norm in the "Greed-is-Good" era of the 80s. However, FSI executives eventually realized that this approach didn’t foster long-term client retention. The introduction of digital capabilities accelerated this shift. By the late 90s, Progressive Insurance was already offering customers a price comparison service.
In 2007, USAA, one of the early champions of customer experience in financial services and insurance, partnered with TrueCar to offer clients a single-stop experience, enabling them to buy a car, obtain a lease, and acquire auto insurance through a single website workflow.
By 2012, USAA was already utilizing customer analytics to personalize offers and integrate them across different channels:
“By using "big data" (generally a mix of internal static transaction information and more fluid, or "unstructured" information from social media and prior consumer communication), USAA hopes to determine if a balance query made on the Web is a prelude to a purchase, or another financial transaction - such as a transfer to a wealth management account or an insurance product - and respond with marketing or information that can be used later on a mobile device.”
The bar for customer-centricity was significantly raised when digital natives like Uber and the 2010 wave of fintech began scaling across the consumer population at large by 2015. It now meant quick refunds on incorrect charges via chatbot messages, intuitive and elegant user experiences, and best-in-class pricing. This bar seemed unreachable for traditional FSIs, leading many experts to conclude that the days of incumbents were numbered due to the imminent unbundling by startups.
Indeed, startups like Wise for money transfers, Mercury for banking, or Next for small business insurance have created a delightful client experience at best-in-class prices, attracting new customers to their platforms. However, this didn’t necessarily mean that existing customers of traditional FSIs were leaving them in droves. Leading players like Chase, Capital One, Progressive, and others faced that challenge head-on by accelerating their technology investments to close the gap. By now, the difference in customer experience between them and fintechs has become marginal.
Of course, traditional FSIs still operate more bureaucratically than fintechs and may not be as effective with digital investments. However, they also have legacy cushy margins to play with. So while Bank of America may be celebrating billions of logins and alerts, they still can ensure that important digital features work well enough to prevent the attrition of profitable customers.
The scale of large FSIs combined with gradual digital transformation meant lower costs and higher margins, which traditional FSIs could repurpose into client benefits, creating virtuous market growth cycles.
What Is That Next Level of Customer Centricity?
How is that theory working out so far? Could FSIs develop a virtuous customer-centric flywheel that increasingly delights customers while maintaining a higher growth rate and positive unit economics?
The results have been mixed. For example, investing in more digitization will keep adding to short-term profitability, but it won’t necessarily make FSIs more customer-centric. Since 2021, NatWest has delivered run-rate savings of around £250 million a year through digitizing customer journeys. Sounds great, except that their customer feedback is one of the worst among competitors.
Some experts would say that "Generative AI" has the answer to making FSIs truly customer-centric. It will ensure a much higher degree of personalization, greatly expanding ease of customer support, and offering products to previously rejected consumers.
But which of those next-level capabilities would be in a consumer’s best interest? Making customers buy and use more financial and insurance products is generally better for FSIs' bottom line, but do they materially improve customer financial well-being? There is a related running joke about BNPL fintechs who always claim that their customers are better off compared to being stuck with credit card debt, except they never prove that their customers got rid of that debt.
No, the ultimate customer-centricity is not in offering the best-priced and easiest solution to meet their short-term need, but the one that adds to a consumer's long-term financial well-being.
Better offers -> more educated clients -> even better offers
Like parenting, ultimate customer centricity involves leveling them up over many years - teaching and pushing them to become more effective at managing finances by changing their decision-making process and habits, such as:
Adjusting spending patterns to get out of debt
Encouraging savings and investments to maximize returns
Promoting habits and property protection to minimize insurance claims
On everyone’s favorite topic of helping poor people, this would mean a bank forcefully telling them which spending they should cut rather than offering a free balance transfer so they could keep spending money they don’t have. To be truly customer-centric also requires FSIs to leave alone customers who are already doing well with their decisions and habits and be forceful with the ones screwing it up.
This presents three challenges for FSIs:
A typical FSI can't furnish a relevant cross-up-sell offer despite a decade of transaction history. Mastering strategic insights across a broader array of inputs might have to wait until AGI is a reality in 2070.
Nudging consumers doesn’t work while forcing them would lead to a massive backlash. “Why do my latte and avocado toast transactions keep getting denied?!”
This might be bad for business. The more effective a consumer becomes, the less margin cushion they leave for an FSI, and the less loyal they get if something is not best in class.
Yes, the ultimate customer-centricity could be destructive to unit economics. It is like paying a dietician $100 per hour to lose weight – if they are good at their job, you shouldn’t need their services after a year because you lost weight and formed healthy and persistent habits. That is why even the world’s best fintechs don’t consider sending customers alerts to use their service less.
Early Attempts At Ultimate Customer Centricity
In a recent survey of US consumers, Chase shared a myriad of potentially helpful digital features: fraud alerts, credit monitoring, and budget tracking. Score Planner was the most interesting tool since it provided actionable recommendations to increase credit scores, including paying off accounts and reducing spending across a customer’s accounts.
What is unknown is whether Chase offers such suggestions for the accounts of its competitors or if internal profits are indeed being sacrificed for customer benefit.
Cyber insurance might be another such area where insurtechs Coalition and At-Bay are not just measuring clients' fitness for underwriting but forcefully leveling them up as part of the ongoing engagement:
But that's just half the battle. How do they keep pricing up after the client has leveled up and other carriers are now willing to underwrite them? They don’t. The average premiums declined about 20% in 2023. Being truly customer-centric means potentially reducing the total addressable market (TAM). Would investors be okay with such a management philosophy?
Telematics in auto insurance also illustrates such a potential dichotomy. After a decade, Nationwide claims to have achieved strong adoption, leading to three notable benefits: 1) 24-34% customer savings, 2) 2-3% higher retention, 3) 10% reduction in handheld distractions, i.e., a potential trifecta of great pricing, happy customers, and fewer accidents. However, it is still too soon to know whether a decade-long focus has resulted in a sustainable business model: as of 2022, Nationwide Auto dropped from the 8th to the 10th rank in written premiums, with a combined loss and expense ratio of 115%.
While such attempts continue, it might be another decade before it is clear whether ultimate customer centricity is congruent with a sustainable business model. In the meantime, an average FSI should keep mastering the ability to send customers offers that they are likely to use, even if it is bad for their financial well-being.
