Since last year, a so-called 'fintech winter' has severely tested the resilience of startups and investors alike, with consumer fintech among the hardest hit. Just two years ago, money poured into consumer fintech: personal financial management (PFM), lending, investing, and more. Today, with lackluster performances in public markets and a lack of profitability across the sector, the picture is markedly different.
Many investors believe that the inability of consumer fintechs to reach profitability is a simple unit economics problem. Customer acquisition costs (CACs) have ballooned as digital channels become oversaturated. By and large, consumer fintechs have not managed to increase average revenue per user (ARPU) at the same rate to maintain, let alone improve, their unit economics. While we largely agree—and recognize that digital acquisition isn’t getting any easier—this narrative misses another pivotal factor lurking further down the income statement. Beneath the contribution margin line, there is a back-office efficiency problem that is not only eroding profitability but also stymieing growth among consumer fintechs.
In the race to delight consumers and displace traditional finance incumbents, fintech companies have historically disproportionately prioritized crafting impeccable front-end user experiences. This focus, while pivotal for customer acquisition and conversion, often comes at the expense of back-office processes and operations. Fintechs often depend on their sponsor bank partners for back-office processes like KYC and AML. While competent at managing regulatory complexity, these legacy banks are traditionally lower-tech and move much more slowly than their fintech partners. Fintechs gather data and funnel it to the banks quickly, but the reverse is not true, creating an asymmetric data flow between both parties.
As fintechs scale, this imbalance between front-end optimization and back-end inefficiency manifests in manual, operationally intensive, and de-linked back-office processes that fail to keep pace with the speed of growth. Ironically, these under-optimized processes eventually result in poorer user experiences. For instance: slower to resolve customer complaints due to fragmented and offline internal communication between departments and delayed access to critical customer and product information; more drop-off at onboarding due to manual KYC, AML, and fraud reviews; slower-to-settle transactions due to manual ongoing monitoring reviews.
Compounding these challenges is the intricate and continually evolving regulatory landscape. In particular, bank-fintech compliance has become a major focus for regulators who are getting more aggressive in enforcing robust and effective compliance programs and risk management (e.g. CFPB and fair lending, OCC’s establishment of Office of Financial Technology, Fed’s creation of Novel Activities Supervision Program). Sponsor banks and fintechs don't want to be caught unaware and want to show they are taking these compliance changes seriously. Therefore, they often self-police and tighten approvals, which increases costs (e.g. investing more in manual reviews).
To address these complex requirements, fintechs typically resort to hiring large teams focused on repetitive and mundane tasks (e.g., reviewing user documents, generating regulatory reports like CDD/EDD, and doing online searches). Throwing headcount at the problem stifles the efficiency and vital operational leverage fintechs require for sustainable growth.
These challenges are exacerbated by a customer quality issue, particularly among neobanks.These platforms often cater to lower-credit-quality customers with higher default and fraud rates. Managing a riskier customer base means more intensive reviews at onboarding, increased transaction monitoring, and risk management to catch anomalies on an ongoing basis, as well as greater resource allocation for fraud detection and prevention. All are harder and costlier in this back-office-inefficient environment.
For a tangible reflection of these issues in action, we decided to look at two of the most prominent public consumer fintechs to test this thesis: Moneylion and Dave.
MoneyLion and Dave offer an example for understanding how back-office inefficiencies and customer quality concerns play out at scale. According to their 2022 income statements, both showed strong topline and gross margin performances, as well as Sales & Marketing (S&M) expense and CAC improvements.
Despite these healthy growth rates, strong product margins, and reduced CACs, Dave and MoneyLion lost $135M and $99M, respectively (before other exp/income), charged off $54M and $84M. As a result, they are both unloved by public markets, currently trading at or below 1x LTM revenue. We believe the root caue is two factors thwarting their path to profitability: 1) A lower-quality customer base, with higher default and fraud rates and 2) a set of operational expenses that are both unsustainable and sticky.
These financial snapshots serve as cautionary tales for the broader fintech landscape, signaling challenges that extend beyond surface-level metrics.
While still unprofitable on a GAAP basis, MoneyLion showed signs of improving its profitability in the first two quarters of 2023, including a reduction in headcount expense as a percentage of revenue. This reflects a range of factors, not least of which is the rise in interest rates, which has provided an inorganic boost to profitability. However, it's crucial to recognize that such improvements are not necessarily sustainable, and investors have largely discounted these temporary advantages in their valuations.
A lack of strategic focus on back-office often results in creating manual and labor-intensive operational practices that lag behind the technological efficiency of the consumer-facing operations. This OpEx is not only expensive, but sticky and hard to shift, largely thanks to regulation.
The complexity of fintech regulatory requirements cannot be overstated. Companies must continually adapt to a diverse set of rules concerning data protection, consumer privacy, risk management, cybersecurity, and more. These regulations come from various bodies, each with their own set of laws, practices, and frequent rule changes. Failure to monitor and enforce these shifting guidelines exposes fintechs to substantial fines and reputational damage. Furthermore, the fear of regulatory action trickling through sponsor banks down to their fintech clients creates even more incentives for fintechs to monitor their customers and create robust compliance processes or risk potentially losing their bank partner.
The result is all too familiar: more headcount to tackle problems manually. For example, financial crime compliance in the U.S. alone costs financial institutions an estimated $46B, the majority of which is spent on manual compliance processes. Despite these financial and personnel commitments, compliance remains a huge challenge. Financial fraud and Anti-Money Laundering (AML) fines rose by 30% and 50%, respectively, in 2022, highlighting the shortcomings of existing solutions. And even when teams are in place to manage these processes, human errors can be devastating.
Moreover, the regulatory landscape continues to evolve, and advances in AI are making it easier to commit financial fraud. Not only will compliance remain a stubborn cost center that does not scale with revenue, but the associated risks and challenges are also likely to intensify.
Finding streamlined solutions to these back-office problems is a matter of urgency. Specifically, we see these scaling issues playing out in a few key areas: onboarding, ongoing monitoring & risk management, and customer service. However, we also believe each of these areas presents exciting opportunities. It's crucial to recognize that both fintechs and sponsor banks share these responsibilities to different extents, and the success of these opportunities hinges significantly on the agility of sponsor banks and their propensity to share data and expedite or hinder these processes.
Problem: The onboarding process exposes deep inefficiencies in fintech's scaling, marked by steep costs, wasted human potential, and subpar customer experiences, despite frantic efforts to patch these holes through outsourcing and overburdened teams.
Onboarding reveals the first cracks in fintech's scaling model, specifically in the form of manual KYC, AML, and fraud reviews. This is a costly endeavor, both in terms of human and financial resources. With the average KYC/AML/fraud analyst costing ~$90,000 annually and individual compliance checks ranging from $13 and upward of $130, the impact on unit economics is harsh (i.e. vs. a ~$50 ARPU). However, it's not just the high costs that are concerning—it's also the inefficiency. Less than 15% of an AML investigator's workday is dedicated to actual investigative tasks—the rest is consumed by manual processes and administrative overhead. Further, manual reviews tend to result in high customer drop-off rates and poor satisfaction scores.
To combat this inefficiency, fintechs often resort to hiring outsourced analysts or consultants. While this approach may seem like a quick fix, it introduces its own set of problems. The onboarding and training process for these external teams can stretch into months, and maintaining quality assurance becomes a significant challenge. Additional tooling or automation to improve efficiency usually falls on the internal engineering teams, further stretching already limited resources. Furthermore, unlike at banks, processes at fintechs evolve quickly such that activities like updating credit models or refining AML review steps can look different on a quarter-over-quarter basis, further intensifying the challenges of outsourcing.
The intensive manual component of these tasks not only drives up costs but also cements the need for a sizable, and therefore expensive, workforce.
Opportunity: Amid a sea of fintech innovators targeting onboarding, the true edge lies in proprietary data moats. Harnessing unique data insights not only refines onboarding and slashes fraud rates but also streamlines operations and elevates the customer journey by minimizing false positives, delivering a seamless experience while maintaining robust security.
The first step to onboarding customers is verifying identity, and the second step is assessing them for risk, fraud, and AML signals. These key steps of the process have been inundated with innovation, with hundreds of companies vying for a slice of the market by targeting KYC, account opening, document verification, fraud detection, AML, orchestration, and more. Yet, very few of them can consistently and accurately detect emergent fraud and AML patterns, and therefore require manual human intervention to manage them, which typically creates more manual reviews and human errors. Thus, amidst this crowded landscape, we believe the key to solving this problem is having proprietary data moats that address specific fraud vectors.
Sentilink, an ID fraud provider, exemplifies the potential of leveraging proprietary data. Their unique approach of buying written-off bad debt to discern behavioral patterns and infusing them into their algorithms, showcases the transformative power of proprietary data in detecting fraud and establishing a competitive edge. As Sentilink developed its consortium approach, its algorithm’s potency amplified, making it a formidable barrier against fraudulent attempts AND a highly defensible platform vs competitors. Baselayer (Torch portfolio company) is employing a similar approach within B2B SMB fraud. The ever-evolving nature of fraud, where fraudsters continually adapt and modify their tactics, can be both a curse and a blessing. While it poses challenges, it simultaneously presents golden opportunities for vendors to detect and address emergent fraud patterns, as Sentilink has done with ID fraud.
There are also emerging early-stage innovators within the onboarding landscape like Minerva which is harnessing the power of proprietary data not just to detect, but also improve, the decision-making process in money laundering cases and more accurately assess and understand risk. Companies like Parcha, Greenlite and Accend are deploying AI frameworks to enhance accuracy and efficiency in back-office processes related to compliance and identity verification. While these businesses currently provide services that primarily rely on machine learning models, we believe the winners in this space will distinguish themselves by effectively utilizing proprietary data to forge superior insights and more accurate predictions.
However, not all vendors in the identity space are created equal. A sizable portion lacks proprietary data, essentially repackaging and redistributing data from other sources, and over time will find it challenging to stand out and stay relevant.
Problem: The persistent reliance on manual processes for ongoing monitoring & risk management hampers efficiency and introduces significant room for human error. The continuation of these outdated methods restricts scalability and exposes fintechs to regulatory risks and inefficiencies.
Activities like transaction and AML monitoring, as well as compliance controls testing, are still predominantly manual processes in many organizations. While tools like Unit21 and Sardine streamline more transaction monitoring and ongoing fraud & AML prevention processes, many fintech companies remain anchored to manual systems where staff must continuously update expiring documents, track alterations in country-specific risks, and stay abreast of changes to sanctions lists, among other tasks.
Compliance controls testing — which usually involves manually sampling a percentage of accounts on a quarterly basis to ensure they meet financial crime requirements — also falls into this manual-labor intensive category.
These operational necessities are manual largely due to regulatory imperatives, which, again, results in sticky headcount expenses that make it challenging for fintechs to scale.
Opportunity: Amid a landscape riddled with manual compliance hurdles, emerging tools for ongoing monitoring and risk management present a golden opportunity. By targeting the most cumbersome aspects of compliance, these innovations are poised to redefine operations, merging heightened standards with marked operational cost reductions.
Tools for ongoing monitoring & risk management are also emerging as a necessary category to automate cumbersome compliance processes and ensure adherence to pertinent financial regulations and laws. The significance of these tools lies in their ability to eradicate the most manual and disjointed elements of the compliance routine, drastically streamlining operations.
For instance, Cable is reshaping the compliance landscape by automating effectiveness testing and the user sampling processes. ComplyCo is assisting financial institutions in seamlessly meeting their regulatory communication and documentation responsibilities. Refine Intelligence is applying AI to monitor money laundering decision making and obtain clear evidence and explainability of AML alerts. There are other platforms like Ethos focused on scaling model validation capabilities and accelerating the model approval process. The advent of such tools not only promises heightened compliance standards but also heralds significant cost savings by trimming down manual tasks and headcount needs.
Problem: Customer service is a costly, labor-intensive maze, marked by extensive training needs, fractured communication, and a consistently lagging customer experience despite major OpEx investments.
CS is a notoriously manual function and represents a significant and growing portion of a fintech’s OpEx. Financial decisions are highly considered and frequently complicated purchases, necessitating large CS teams to address questions and resolve complaints. Training agents on not only the fintech's range of products but also the regulatory landscape is both time-consuming and costly. Despite significant headcount investments in human resources, the customer experience often remains subpar because agents frequently don’t have immediate access to the information needed to resolve customer issues promptly and internal communication is fragmented and poorly integrated with the rest of the organization.
Despite pouring a lot of human capital into customer service, fintechs grapple with both escalating and sticky costs and diminished user satisfaction due to these inefficiencies — many of which stem from regulatory complexities.
Opportunity: In the fintech landscape, where trust is paramount, evolving beyond a siloed customer service approach is vital. The future hinges on integrated platforms that fuse CS with the broader organization, accelerating resolutions, fostering transparency, and driving operational synergy across departments.
The importance of responsive and effective customer service cannot be overstated, especially in fintech where trust plays such a pivotal role. Seamless resolutions to customer complaints and inquiries have become a primary metric of success for fintech firms. While platforms like Zendesk are the go-to for many fintech companies, there’s an evident gap in how customer service interacts with the broader organization. This often siloed nature of customer service operations can lead to delayed resolutions and a fragmented user experience.
The next frontier in fintech customer service is the development of platforms that seamlessly integrate CS teams with other departments. Such an interconnected platform would empower customer service representatives with real-time access to information, allowing them to either immediately address a complaint or seamlessly escalate it to the relevant department—be it marketing, finance, or any other. This not only promises faster resolution times but also cultivates an environment of transparency and collaboration within the organization. In essence, a tool that fosters this level of inter-departmental cohesion not only enhances user trust and satisfaction but also streamlines internal workflows, leading to operational excellence.
Several startups are addressing the urgent need for integrated and enhanced customer support in fintech. Companies like Parabolic and Ada provide an AI-powered customer service platform. Narrative is building an automated, intelligent operations platform with applications in streamlining customer service. Casap is focused on automating the card disputes process which can significantly lighten the load for customer support teams. These innovations signify a shift toward more integrated, intelligent, and customer-centric approaches in handling service requests and streamlining operational challenges.
While we believe these areas of opportunity are hugely promising, back-office operations are not typically top of mind for C-suites, and scaling successful solutions in this space comes with challenges. We believe the following strategies are critical to the success of back-office innovation:
Any software solution targeting the back-office must tackle pressing, immediate issues, acting as a wedge into broader organizational needs. It's not enough to solve just one problem, no matter how acute. The ideal tool offers a roadmap to expansive functionalities, catering to diverse challenges within the fintech environment.
Further, as is the case with all venture investments, investors are searching for scalable solutions with considerable market potential. Solutions that limit themselves to singular functions within the fintech back-office, such as KYC or customer service alone, often find their market potential capped. In a space that demands both innovation and breadth, solutions need to think broader.