From frontstage to backstage: how the operating model explains the difference between incumbents and startups in fintech, healthcare, and retail

Summary

This white paper explores the less visible side of digital transformation: the operational backstage that sustains the user experience in three data‑ and regulation‑intensive industries—fintech, healthcare, and retail. While public debate usually focuses on mobile apps, branding, and marketing campaigns, these “frontstage” layers explain only a small part of startups’ competitive advantage. The central thesis is that the structural difference lies in how emerging organizations reconfigure value, data, and decision flows in the back office.

Using a three‑layer framework—business model, technology architecture, and internal process design—we compare banks versus fintechs, hospitals/insurers versus healthtechs, and traditional retailers versus digital‑native merchants. We integrate recent evidence on sector regulation, startup failure rates, and organizational culture to explain why some approaches scale and others do not [1][2]. We show that an attractive interface without backend changes yields only marginal improvements, whereas redesigning the backstage enables leaps in response times, personalization, and transparency. Finally, we discuss practical implications for incumbents and startups, and collaboration opportunities focused on operating‑model redesign rather than superficial commercial agreements.

Introduction: from ‘frontstage’ to ‘backstage’

When comparing startups with traditional companies, the conversation often remains on the surface. Sleek app designs are contrasted with outdated interfaces, social media campaigns with TV ads, or digital onboarding experiences with paper‑based forms. This approach, though intuitive, is insufficient to explain why, despite investing millions in redesigning interfaces, many established organizations still deliver cumbersome, poorly personalized experiences.

To understand the real difference, it is helpful to distinguish between frontstage and backstage. The frontstage is everything the user sees: the app, the website, the point of sale, the brand tone, the service channel. The backstage includes information systems, data flows, business models that define incentives, internal processes, and team autonomy. It is, ultimately, the hidden operating model that determines what can be promised and delivered in the frontstage [1].

In fintech, healthcare, and retail, recent evidence points in the same direction: the true competitive advantage of the startup ecosystem lies less in visual design and more in how it reconfigures value, data, and decision flows in the back office [1]. In Latin America, for example, the creation of specific regulatory frameworks for fintech in countries such as Brazil, Mexico, Argentina, Colombia, and Chile has enabled platforms to serve unbanked segments through novel business models and lightweight technology [1]. However, that same freedom has contributed to very high failure rates: roughly 75% in fintech and about 80% in healthtech, partly because many backstages cannot sustain growth or regulatory compliance [2].

This article decomposes that backstage into three layers and analyzes it across three industries with a clear goal: to show that user experience (UX) is a symptom, not the cause, of the underlying operating‑model design. Understanding that causal relationship is key to directing digital‑transformation investments to the points of greatest structural impact.

Comparative framework: three backstage layers

To compare incumbents and startups beyond the surface, we propose a three‑layer analytical framework:

1. Business model and revenue sources
2. Technology and data architecture
3. Internal process design and team autonomy

Each layer defines a set of constraints and possibilities that together determine UX—from response times to personalization levels and bureaucratic friction [1].

First, the business model and revenue sources shape incentives. A bank that monetizes mainly through credit volume and cross‑selling financial products has very different priorities from a fintech focused on underserved niches, dynamic pricing, and “as‑a‑service” offerings [1]. A healthcare system based on fee‑for‑service billing incentivizes episodic activity and intensive use of procedures, whereas subscription or outcome‑based models, common in healthtech, favor lifecycle patient management and prevention [1]. In retail, the gap between a model based on volume and supplier negotiation and one centered on direct customer relationships defines how much is invested in understanding the end user and capturing interaction data.

The second layer is technology and data architecture. Traditional firms typically run on legacy monolithic systems, batch integrations, and fragmented databases. This hinders a single customer view and limits the ability to orchestrate real‑time omnichannel journeys. Startups, by contrast, tend to use cloud architectures, microservices, and APIs to connect specialized modules, enabling rapid iteration, intensive use of alternative data, and algorithmic personalization [1]. This difference is critical in highly regulated sectors, where traceability and security must coexist with agility.

The third layer is internal process design and team autonomy. Organizational culture is a decisive factor. Startups embed technology into their DNA and treat processes as hypotheses to be tested; as a result, 90% already use advanced digital tools in key processes such as recruiting, compared with 65% of traditional firms [2]. Hierarchical structures and rigid processes can stall adoption of technologies that, on paper, are available. Conversely, small, cross‑functional, empowered teams accelerate redesign of critical processes and open‑innovation adoption [2].

These three layers manifest directly in UX. Credit approvals that take days versus minutes, continuity or fragmentation in patient care, or the ability to buy in one channel and return in another depend far less on button colors and far more on how business models, data systems, and internal processes are orchestrated.

Methods

The analysis is based on a synthesis of recent secondary sources and well‑documented cases of digital transformation in fintech, healthcare, and retail [1][2]. Instead of conducting primary fieldwork, we build a unified conceptual framework and apply it comparatively to three prototypical journeys:

• Credit application and approval (fintech)
• Searching, scheduling, and following up on a medical appointment (healthcare)
• Discovery, purchase, and returns in an omnichannel environment (retail)

In fintech, we incorporate evidence on the development of specific regulatory frameworks in Latin America that have facilitated the expansion of digital platforms aimed at financial inclusion, while raising new challenges in supervision and consumer protection [1]. For digital health, we use a 2021 study on success factors in eHealth adoption, highlighting the role of trust, perceived security, and user acceptance in the effectiveness of these solutions [2]. In retail, and transversally, we consider data‑protection and privacy regulations (such as GDPR and equivalents elsewhere) as preconditions for safe, trustworthy omnichannel experiences [1][2].

We complement this regulatory backdrop with statistics on average tech‑startup longevity (6.7 years) and sector‑specific failure rates—75% in fintech, around 80% in healthtech—to illustrate the fragility of models when the backstage does not scale adequately [2]. We also include studies on organizational culture and digital transformation, such as the “Siemens 2020” program, which achieved a 15% productivity increase via training and multidisciplinary collaboration [2], and experiences at companies like IBM in continuous digital upskilling and open innovation [2].

On this basis, we construct three sectoral “mini‑studies” using the same three‑layer framework, then derive cross‑cutting patterns and practical recommendations. The goal is not an exhaustive technology inventory, but an evidence‑based explanation of how invisible backstage decisions translate into observable differences in user experience.

Key findings

Fintech: granting a loan

In the credit application and approval process, divergence between a traditional bank and a fintech begins at first contact. In a bank, the application often starts at a physical branch or via a web form tied to a legacy core banking system. This core is designed for accounting and transaction recording, not digital‑journey orchestration. Information is often validated manually or semi‑automatically, with approvals cascading through multiple hierarchical levels. Systems are typically integrated via batch processes, which means key data—such as recent income history or credit status—may be outdated at decision time [1].

A fintech, by contrast, designs the flow to be initiated and completed on a mobile device, with a modular decision engine consuming data from multiple sources via APIs. Beyond traditional variables (income, credit history), it can integrate alternative data such as digital‑consumption patterns, repayment behavior on e‑commerce platforms, or even social‑network signals, where regulations allow it [1]. In Latin America, evolving fintech‑specific regulatory frameworks have enabled such companies to operate as platforms serving unbanked populations with fully digital onboarding, using biometrics, remote document verification, and connections to government registries [1].

On the business‑model side, traditional banks work with standardized products, regulated margins, and a focus on cross‑selling services (accounts, cards, insurance). Profitability depends on scale and a large customer base. Fintechs tend to specialize in specific niches: small‑business loans, alternative consumer finance, low‑ticket/high‑frequency credit, or services for users without formal banking histories [1]. They monetize via dynamic pricing, service fees, and increasingly Banking‑as‑a‑Service (BaaS), providing infrastructure to third parties that want to embed financial services in their own apps.

Technologically, the contrast is just as stark. Banks rely on centralized legacy cores with tightly coupled modules and strong constraints on change. Internal and external integrations often run as nightly batch jobs. Data sit in silos: card, loan, account, and digital‑channel data are not always reconciled in real time [1]. Fintechs start from cloud‑based architectures and microservices exposing functionality via APIs. The scoring engine is decoupled from the transactional core, allowing risk models to evolve without rewriting the central system and enabling fast testing of new business rules.

The UX result is measurable: processes that may take days in a bank—weeks if extra documentation is required—are reduced to minutes in fintechs, with automated decisions and immediate communication of outcomes. Users know conditions, costs, and terms upfront, can simulate scenarios, and can self‑manage changes. Document friction drops to a viable minimum thanks to automatic identity verification and connectors to public and private data sources. This difference is not explained by a “nicer app” alone, but by a backend enabling real‑time decisions at low marginal cost per operation.

However, this agility has hidden costs. Fintech faces a failure rate near 75%, due to intense competition, regulatory and compliance challenges, and more [2]. Documented cases show promising startups failing because they never reach product‑market fit, underestimate customer‑acquisition costs, or do not internalize security and supervisory requirements from the outset [2]. A lightweight, distributed backend can become a liability when operational robustness, cyber‑resilience, and continuous adaptation to prudential rules are required. Balancing speed and backstage solidity is particularly delicate here.

Healthcare: scheduling and managing a medical appointment

In healthcare, the “find a doctor, book, attend, and follow up” journey exposes a structural gap between traditional systems and digital‑health startups. In a conventional hospital or clinic network, doctor search often depends on call centers, informal recommendations, or static website listings. Scheduling is frequently managed with local systems per specialty—or even spreadsheets and paper diaries. Health records are fragmented by institution and sometimes by department, preventing professionals from having a complete view of the patient [1].

By contrast, a healthtech or telemedicine platform integrates the process end‑to‑end. Users search for professionals filtered by specialty, availability, language, insurer, or modality (in‑person/virtual) via app or web. The scheduling system connects to a cloud‑based electronic health record (EHR) so each consultation adds to a single patient file. Automated reminders reduce no‑shows, and triage algorithms direct patients to the most appropriate professional or channel based on symptoms [1].

The business model also shifts. Traditional systems are anchored in fee‑for‑service billing and rigid contracts between hospitals, clinics, and insurers. Patient relationships tend to be episodic: the goal is to optimize schedule and bed occupancy, not continuity of care. Digital‑health startups experiment with subscriptions, outcome‑based payments, and professional marketplaces where the platform captures a commission per appointment. Some models pivot to prevention and continuous monitoring, aligning incentives with the reduction of costly acute events such as avoidable hospitalizations [1].

Technologically, traditional players struggle with closed, poorly interoperable systems. Each institution—and often each department—runs its own software for clinical records, labs, and imaging. Interoperability is limited, making it hard for information to follow the patient. These limits are not only technical; they also reflect strict health‑data protection rules requiring high standards of security, informed consent, and granular access control [2]. Startups, by contrast, typically design cloud‑based, integrated EHRs from day one, using interoperability standards (e.g., FHIR) to share data among authorized providers at different levels of care.

The resulting UX shows how much continuity of care depends on a well‑designed backstage. In traditional systems, patients repeat information at every visit, carry printed test results from one center to another, and have only a partial view of their own history. In healthtech platforms, continuity materializes as appointment and medication reminders, mobile access to health records, result summaries, and remote monitoring. A 2021 eHealth study finds that user trust and perceived control over data are key to system acceptance [2]. Video consultations are of little use if the underlying database does not consolidate information or equip professionals to make informed, coordinated decisions.

Yet healthtech startup failure rates are even higher than in fintech, around 80% [2]. The combination of tight regulatory scrutiny, clinical complexity, and interoperability needs means that a sophisticated technical backend is not always sufficient. Without deep understanding of clinical workflows, payer incentives, and local legal requirements, the promise of smooth UX collapses when scaling. The Theranos case, though extreme and fraudulent, illustrates what happens when an innovation narrative is not backed by a robust clinical and scientific backend [2].

Retail: omnichannel experience

In retail, the “discover, compare, buy, receive, and possibly return” flow tests the backstage’s ability to orchestrate multiple channels. A traditional omnichannel retailer usually starts from a monolithic ERP managing inventory, procurement, and billing. Over this core, over time, a legacy e‑commerce site, in‑store POS systems, and sometimes a heavy CRM are layered. Integrations are rigid and often one‑way, producing outdated catalogs and limited real‑time inventory visibility [1].

A DTC startup or digital‑native e‑commerce player designs its operating model in reverse: it starts from the direct customer relationship and the need to capture data from every interaction to feed marketing, product, and operations. The business model emphasizes micro‑segmentation, recurring sales (subscriptions), and lifetime‑value optimization rather than pure volume and supplier bargaining power [1]. This justifies significant investment in advanced analytics and experimentation (A/B testing, dynamic personalization) that traditional structures tend to defer.

The technology architecture of these digital natives relies on “composable commerce.” Instead of an all‑in‑one system, they assemble a lightweight CMS, a specialized checkout engine, payment providers, logistics platforms, and analytics tools, all connected via APIs. Each component can be swapped out without rewriting the rest. This allows, for example, changing logistics providers without altering the storefront, or rapidly adding new payment methods in specific markets [1].

From a UX standpoint, the difference shows up in concrete ways: consistent pricing between physical and online stores, real‑time visibility of size/model availability, easy buy‑online/return‑in‑store flows, and detailed shipment tracking. In traditional retailers, the promise of “buy anywhere, receive anywhere” often collides with a backstage unable to reconcile inventory and orders across channels. Returning an online purchase in‑store may require manual steps, approvals, and ERP exceptions, introducing friction and uncertainty for customers.

Digital‑retail startups, with fully traceable real‑time inventory and orders, can automate much of reverse logistics and offer simple return policies. Still, they face constraints: they are heavily dependent on third parties for logistics, payments, and often the e‑commerce infrastructure itself. This dependence can become a bottleneck during rapid scaling or regulatory shifts in data protection and privacy, which are critical to retaining consumer trust [2]. Cookie policies, consent for data use, and limits on targeted advertising directly shape how granular UX personalization can be.

Comparative summary of the three industries

Backstage differences across the three sectors can be summarized as follows:

This comparison reinforces the idea that visible differences for users are manifestations of deeper backstage decisions, not the other way round.

Comparative analysis

Business‑model layers: incentives and risks

In fintech, healthcare, and retail, traditional business models tend to optimize internal metrics (loan volume, number of procedures, inventory turnover) rather than holistic experience quality. This is no accident: incentive systems and cost structures are designed around legacy products and processes. Free from that legacy, startups can build models centered on customer lifetime value, prevention, or recurrence, which allows them to justify UX and data investments incumbents see as “extra cost” [1].

However, user‑centricity does not eliminate risk. Failure rates of 75% in fintech and 80% in healthtech [2] show many models fail to balance value proposition, financial sustainability, and regulatory compliance. Aggressive pricing can erode margins before scale is reached; prevention‑oriented health models without clear payment mechanisms can clash with fee‑for‑service systems [1]. In retail, DTC strategies driven by heavy digital‑ad spend can become untenable as acquisition costs rise or data‑usage rules tighten.

Technology architecture: monoliths versus modular platforms

Technologically, the monolith‑versus‑modular contrast defines each actor’s room to maneuver. A banking core or retail ERP provides stability and compliance but hampers rapid change. Each modification may require long test and deployment cycles, discouraging UX experiments or new data‑source integrations. In healthcare, lack of EHR interoperability impedes cross‑system journeys, hurting continuity of care [2].

Startups, by betting on microservices and cloud services, gain flexibility but introduce other complexities. Dependence on multiple infrastructure, payments, analytics, or logistics providers means managing an ecosystem of third parties, each adding risk vectors. The apparent lightness of these architectures can be misleading: as they grow, orchestrating dozens or hundreds of services and APIs demands data and integration governance as rigorous as that found in incumbents—yet built almost from scratch [1].

Processes and culture: agility versus resilience

The third dimension is cultural and procedural. Startups treat internal processes as hypotheses to validate, not fixed procedures. Small, cross‑functional teams are organized around products or journeys, aligning promises to users with delivery capabilities. This stance is reflected in widespread adoption of digital tools: 90% of startups already use advanced solutions in processes such as recruitment, compared with 65% of traditional companies [2].

Traditional organizations, despite their inertia, can transform their backstage if they tackle culture and processes deliberately. The “Siemens 2020” program, focused on training and multidisciplinary collaboration, delivered a 15% productivity boost in German plants, showing that agility can be increased without sacrificing robustness [2]. Cases like this show the gap is not just available technology, but how teams are structured and autonomy is distributed. Continuous digital‑skills programs, such as IBM’s, and open‑innovation mechanisms help incumbents incorporate startup‑like capabilities without giving up sector expertise [2].

Trade‑off comparison table

Trade‑offs between startups and incumbents can be summarized as:

This table underscores that no model is “perfect”: sustainable competitive advantage requires combining strengths from both worlds.

Case studies

Case 1: instant‑credit fintech in Latin America

Consider a Latin American fintech specializing in consumer loans for users with limited credit histories. Leveraging favorable regulatory frameworks in Brazil and Mexico, the company designs 100% digital onboarding supported by biometric identity verification and connections to alternative data sources such as utility‑payment history and online shopping behavior [1]. The business model mixes transaction fees with revenues from licensing its scoring engine to merchants that want to offer point‑of‑sale financing.

Its architecture is built on microservices and APIs: one service for registration, another for scoring, another for portfolio management, plus multiple connectors to data providers and payment gateways. This modularity allows launching new products in weeks and adjusting risk criteria almost in real time. For users, the result is a flow where applications are processed in minutes, documents are uploaded from mobile, and loan tracking is fully transparent in the app.

As the company scales, backstage tensions emerge: it must strengthen fraud and compliance controls, formalize previously informal processes, and guarantee high service availability. The same flexibility that fueled its growth becomes an orchestration challenge. The firm is forced to invest heavily in data governance, security, and regulatory reporting, converging toward standards similar to a traditional bank while trying to preserve the seamless experience that differentiates it.

Case 2: prevention‑focused digital‑health platform

Imagine a health startup offering a monthly subscription that combines unlimited telemedicine…