The invisible revolution: how startup infrastructure is rewriting the rules of traditional industries

Abstract

Competitive transformation between startups and incumbents is often narrated as a battle of prettier apps, better user experiences, and aggressive pricing strategies. However, the real advantage no longer lies in the visible layer, but in an invisible infrastructure of technologies, data, and B2B2X business models operating under the hood. This white paper argues that future competition will be “modern modular stack vs. rigid legacy stack,” more than startup vs. traditional company.

Drawing on cases in finance, retail and e‑commerce, healthcare, and urban mobility, it analyzes how Banking‑as‑a‑Service platforms, “plug‑and‑play” logistics, healthcare interoperability, Mobility‑as‑a‑Service, AI, and blockchain are redefining value distribution across multiple sectors. It includes examples of large companies that have embraced this invisible infrastructure — such as General Electric, Sacyr, and Nike — and startups that are building it in cybersecurity, quantum computing, or blockchain [1][2][3][4]. The document concludes with strategic implications: which part of the stack incumbents should control, which part they should rent, and how startups can avoid becoming mere “API commodities.”

Background

For more than a decade, the dominant narrative about digital disruption has revolved around the surface: mobile apps, responsive websites, attractive interfaces, and aggressive digital marketing. In this narrative, startups appear as agile actors that “understand the user,” while incumbents are seen as heavy organizations, trapped in slow processes and legacy systems. The comparison is limited to what the user sees on the screen and to pricing aggressiveness, leaving in the shadows the foundations that make those experiences possible — or impossible.

However, evidence from recent years suggests that real competition has shifted to the infrastructure layers. The massive migration of critical systems to the cloud by companies like General Electric — which in 2017 moved more than 2,000 applications and services to Amazon Web Services to free itself from the cost and rigidity of its own data centers [1] — illustrates that the competitive differential is being played out in underlying architecture rather than in the presentation layer. Similarly, Sacyr has developed GeOS, a transversal digital platform that began as a project‑management tool and ended up becoming the infrastructural base for multiple areas of the company, with direct impact on efficiency and collaboration [2].

In parallel, highly specialized startups have emerged as providers of critical invisible capabilities: cybersecurity for industrial infrastructure (Steryon) [1], integrated cybersecurity and intelligence platforms for defense (DefAgent) [2], quantum computing software accessible through familiar interfaces like Excel (Multiverse Computing) [3], or blockchain infrastructure with payment tools and wallets for more than 40,000 companies (Crossmint) [4]. Although the end user rarely knows these names, their services determine which experiences are possible, at what cost, and at what speed.

In this context, “invisible infrastructure” encompasses not only IT, but also business models (BaaS, infrastructural SaaS, data marketplaces), service orchestration (Mobility‑as‑a‑Service, logistics as a service), and data and AI platforms that redefine how value is created and captured. Understanding this layer is essential for executives and innovators who need to decide where to compete: as a final product, as infrastructure, or as ecosystem orchestrators.

Methods

This white paper is based on a qualitative synthesis of recent secondary sources on digital transformation, use cases of invisible infrastructures, and the evolution of business models in regulated and unregulated sectors. Documented examples of adoption of cloud platforms, cybersecurity, quantum computing, blockchain, and B2B2X infrastructures were prioritized, as well as analyses of the impact of regulation on technology startups.

First, cases of large companies that have migrated their technological infrastructure to platform‑based models, such as General Electric and Sacyr, were reviewed to illustrate how incumbents use invisible infrastructures to gain operational flexibility [1][2]. Second, examples of infrastructural startups — including Steryon, DefAgent, Multiverse Computing, and Crossmint — were analyzed with a focus on their value proposition, target sectors, and technologies used [1][2][3][4]. Finally, reports on regulatory barriers affecting startups in sectors such as healthcare and finance were incorporated [5][6][7].

Based on these sources, a conceptual framework was constructed that breaks down invisible infrastructure into three layers (business, technology, and extended experience) and applies it across four sectors: finance, retail and e‑commerce, healthcare, and urban mobility. The analysis focuses on causally connecting infrastructural architecture with the evolution of business models and user experience, rather than describing isolated technologies.

Map of invisible infrastructure: business, technology, and user experience

Invisible infrastructure can be understood as a stack of three interconnected layers that together determine an organization’s ability to innovate in products, pricing, and experience: business layer, technology layer, and extended experience layer.

Business layer: from final product to enabling capability

The business layer includes models that package critical capabilities in the form of a service: Banking‑as‑a‑Service (BaaS), infrastructural SaaS, regulatory compliance platforms, risk platforms, and data marketplaces. These offer startups regulatory‑financial, operational, and data “Lego blocks” that previously required years of investment and licenses. BaaS providers allow the launch of accounts, cards, or credit products without being a bank; data marketplaces offer reusable risk histories or consumption patterns; compliance platforms automate KYC/AML tasks.

The strategic impact of this layer is twofold. On one hand, it reduces the barrier to entry for new models: what previously required building a full‑stack organization can now be purchased as a service and paid per use. On the other, it reconfigures the value chain: whoever controls business infrastructure can capture sustainable rents even when end brands compete fiercely on price and UX. Startups like Crossmint, which offers blockchain infrastructure integrated with AI for payments, wallets, and automated operations to more than 40,000 companies — including global brands such as Adidas and Red Bull [4] — illustrate how it is possible to become “the layer everyone uses” instead of competing on the front end.

Technology layer: APIs, microservices, data, and AI as connective tissue

The technology layer groups modular APIs, microservices, AI‑as‑a‑service platforms, workflow orchestrators, data lakes, event streaming, and digital identity solutions. It is the connective tissue that allows the business layer to be truly composable. General Electric’s migration to AWS, with more than 2,000 applications moved to the cloud in 2017 [1], shows the power of externalizing part of this layer to gain elasticity, reduce CAPEX, and access advanced services (analytics, AI, security) that are hard to replicate in‑house.

Startups such as Steryon, which offers an AI‑based cybersecurity platform for complex industrial infrastructures [1], or DefAgent, focused on cybersecurity and intelligence for defense and the battlefield [2], demonstrate that even functions traditionally seen as “strategic core” can be partially externalized to hyperspecialized providers. Multiverse Computing, for its part, encapsulates quantum and AI algorithms in a platform — Singularity — that is consumed from familiar tools like Excel, enabling companies in energy, logistics, manufacturing, finance, or mobility to apply quantum computing without in‑house experts [3]. The technology layer thus becomes a horizontal accelerator that scales across multiple industries.

Extended experience layer: unified UX with invisible providers

The third, extended experience layer includes white‑label recommendation engines, invisible payment systems, “plug‑and‑play” logistics, and omnichannel systems that unify the experience regardless of who provides each module. The user believes they are dealing with a single coherent brand; in reality, they interact with an ecosystem of hidden providers.

Nike offers a powerful example of how invisible data and analytics infrastructure can translate into tangible experience improvements. The company has deployed data analytics and personalization technologies which, integrated into its digital stack, drove a 36% increase in its digital revenues in the last quarter of 2020 [3]. The customer perceives relevant recommendations and smooth journeys, without seeing the complex orchestration of platforms and providers behind the scenes. Similarly, numerous e‑commerce players rely on recommendation engines and white‑label payment systems to offer “superapp‑type” experiences without building each component.

In digital‑native startups, these three layers are designed jointly, with an eye on continuous business‑model pivoting and UX experimentation. In many traditional organizations, by contrast, they survive as disconnected strata: legacy systems in the back end, rigid business processes in the middle, and experience layers that can only improve marginally without touching the foundations.

Key Findings

Finance: from visible banks to invisible embedded banks

Historically, the dominant banking model has been full‑stack and closed. The bank controlled the product (accounts, loans, cards), the channel (branches, online banking), the core banking system, and compliance processes. This vertical integration allowed it to capture a significant share of the value generated along the chain, but at the cost of high rigidity: every product change meant touching core systems, and each experience innovation required coordination across multiple internal silos. The result has been limited innovation speed, especially compared with the expectations of users accustomed to “one‑click” experiences.

The fintech ecosystem has introduced a radically different logic based on unbundling the stack. Banking‑as‑a‑Service providers enable third parties — from neobanks to marketplaces and commerce apps — to offer financial products under their own brand, relying on a third party’s licenses and core banking system. Other providers specialize in credit scoring, KYC/AML, fraud prevention, or risk management, integrating through APIs. Monetization shifts towards usage‑based models: API calls, processed volume, or monthly subscriptions.

This transition has direct implications for the business model. The traditional bank generates income mainly from its own products and depends heavily on physical branches and legacy systems. The infrastructural fintech, by contrast, lives by enabling others and prioritizes B2B or B2B2C scalability. Startups like Crossmint, whose blockchain stack with AI serves more than 40,000 companies [4], show the power of capturing economies of scale when the business unit is “infrastructure layers” and not individual accounts. In parallel, some traditional banks have begun using such infrastructures to launch digital sub‑brands or embedded experiences, while others remain anchored to their historical core.

From the user‑experience standpoint, the change is profound. Thanks to embedded banking, a consumer can open an account, access financing, or receive a virtual card from an e‑commerce app or superapp without ever dealing with a visible bank. The financial brand disappears from view, replaced by the marketplace or service provider brand. For the bank that limits itself to being infrastructure, this implies giving up direct contact with the end customer but gaining scale and efficiency; for the incumbent that does not integrate into these architectures, it creates the risk of becoming a mere balance‑sheet provider without controlling the experience.

Retail and e‑commerce: “plug‑and‑play” logistics and payments as battleground

In traditional retail, large distributors built their own logistics chains, monolithic inventory systems, and limited payment gateways. This vertical integration required large capital expenditures (CAPEX) in warehouses, fleets, and systems, but provided entry barriers: few could match incumbents’ logistics coverage or acquiring conditions. However, this model made it difficult to adapt to demand spikes, new channels, or emerging payment methods.

The wave of infrastructural startups has eroded these advantages. Fulfillment and logistics‑as‑a‑service platforms allow any e‑commerce company to offer “prime‑like” fast shipping without building its own warehouses or networks. Modular payment gateways enable local methods, BNPL (Buy Now, Pay Later), wallets, and fraud tools as APIs. Costs become variable OPEX, tied to transactions and volume, and the business model resembles that of a shared platform: many merchants leverage the same infrastructure, which gains increasing returns in scale and data.

The consequence for UX is a democratization of advanced capabilities. A small retailer can offer real‑time tracking, next‑day delivery, and multiple payment options with the same level of sophistication as a global giant, using invisible providers. The customer perceives an experience close to the “Amazon standard,” unaware that behind it lies a network of logistics startups, payment providers, and antifraud services. At the same time, some incumbent retailers have reinvented themselves as platforms, opening their logistics and digital capabilities as a service to third parties, turning former rivals into customers and redrawing competition.

Healthcare: interoperability and patient experience as a service

The healthcare sector has historically been dominated by closed hospital systems, fragmented medical records, and heavy administrative processes. Hospital information systems (HIS) and electronic health records (EHR/EMR) were custom‑developed with little focus on interoperability. The result is a patchwork of unconnected databases in which patients repeat tests, professionals lack a complete view, and coordination between centers and insurers is slow and costly.

Infrastructural startups are attacking this problem from the connectivity and data layer. Platforms that integrate data from multiple EHR/EMR systems provide a unified view of the patient, accessible — under permissions — to different actors in the system. Appointment coordination, telemedicine, and remote monitoring tools are integrated via API into clinic or insurer apps, without requiring them to rebuild their entire stack. Their business model is not based on owning the patient, but on selling technological capacity and connectivity as a service to multiple clients: hospitals, insurers, pharmaceutical companies, and digital health startups.

In terms of experience, the patient benefits from more coherent journeys: they can manage appointments, results, and follow‑ups from a unified app, even though multiple traditional systems are connected behind the scenes by these infrastructures. However, operating in healthcare means navigating a complex regulatory maze. Startups must comply with strict regulations on data privacy and security, specific licenses, and requirements that vary by country or region [5][7]. The absence of harmonized and innovation‑friendly regulatory frameworks for digital health in many markets adds uncertainty and costs [6], making regulatory know‑how and compliance capability an essential part of their value proposition.

Mobility and cities: orchestration of fragmented services

In urban mobility, the starting point has been a landscape of single public‑transport operators (bus, metro, train) with closed ticketing and payment systems, alongside taxi or vehicle companies with their own systems. Each operator controls its sales, its channel, and its user relationship, resulting in fragmented experiences: multiple apps, different physical cards, and little route or fare integration between modes.

Mobility‑as‑a‑Service (MaaS) platforms propose an alternative model: a technological layer that integrates bus, metro, taxi, bikes, scooters, carsharing, and other services, and offers a unified experience to the user. Integrated ticketing and payment solutions allow cities and operators to offer “door‑to‑door” trips in a single purchase. The business model focuses on orchestrating fragmented supply and connecting multiple channels (proprietary apps, superapps, municipal portals), monetizing through commissions, licenses, or revenue sharing.

For the user, the change is significant: they no longer buy “a metro ticket” or “a taxi ride,” but a trip from A to B that may combine several modes of transport without friction. The platform calculates optimal routes, updates in real time in case of incidents, and processes integrated payments. Individual operators risk losing brand visibility and direct contact, but gain occupancy and access to aggregated data. Cities, for their part, can use MaaS infrastructure as a lever for more sustainable mobility policies, at the cost of depending on technology providers that are often highly specialized startups or scale‑ups.

Table 1. Invisible infrastructure layers by sector

Comparative Analysis

Proprietary full‑stack vs. shared infrastructure

Comparing full‑stack models with shared infrastructures reveals a consistent pattern: the former optimize control and direct value capture, while the latter maximize speed and adaptability. In banking, the full‑stack model offered stability and high margins but made it costly to launch new digital products or integrate with third parties. In retail, proprietary logistics chains created entry barriers but made responses to sharp demand changes more rigid.

Shared infrastructure — whether cloud, BaaS, fulfillment, or MaaS — introduces increasing returns to scale and learning: the more clients use the platform, the more data accumulates and the better the algorithms and processes become. Nike, by strengthening its analytics infrastructure, saw its digital revenues grow 36% in a single quarter in 2020 [3], a jump that would have been hard to replicate without a scalable technological base. In return, users of shared infrastructures assume dependency on external providers and, in some cases, relinquish direct visibility over the end user.

CAPEX‑intensive vs. usage‑based OPEX

Another cross‑cutting contrast is between CAPEX‑intensive models and those centered on variable OPEX tied to usage. Traditionally, sectors like construction, industry, and energy invested heavily in data centers, proprietary systems, and dedicated hardware. The case of General Electric, which replaced data centers with AWS for more than 2,000 applications [1], exemplifies how even industrial giants are substituting CAPEX with on‑demand services, freeing resources for innovation.

For infrastructural startups, the business model revolves precisely around monetizing that shift from CAPEX to OPEX: they charge per API call, volume of data processed, or transactions. This facilitates adoption by clients who prefer to avoid high upfront investments and aligns costs with revenues. However, the change also introduces a new kind of financial risk: in periods of strong usage growth, the bill for external services can climb rapidly if agreements are not well negotiated; conversely, in demand crises, variable costs automatically fall, providing resilience.

Experience control vs. ecosystem orchestration

A third comparative axis concerns control over user experience. Incumbents have traditionally aspired to control the entire journey: from acquisition to after‑sales service. In healthcare, hospitals and insurers designed end‑to‑end processes; in mobility, each operator controlled its sales channel. The rise of infrastructural platforms shifts this balance towards an orchestration logic: one entity — not always the most visible — coordinates multiple providers to offer unified experiences.

In this new scenario, the strategic question for incumbents and startups is not just “what product do we sell,” but “what role do we play in the ecosystem”: capacity provider, owner of the user relationship, or orchestrator of multiple actors. Companies like Sacyr, which have extended their GeOS platform from project management to other corporate areas [2], are moving toward an internal orchestration role; other organizations, by becoming logistics or BaaS platforms for third parties, expand their ambition toward external orchestration. However, the orchestrator also assumes the greatest coordination, regulatory, and cybersecurity risk.

Table 2. Trade‑offs between traditional and infrastructural models

Case Studies

General Electric: freeing resources by externalizing infrastructure

In 2017, General Electric (GE) decided to migrate more than 2,000 applications and services to Amazon Web Services (AWS), progressively abandoning its traditional data centers [1]. The decision was not merely technological, but strategic: maintaining physical infrastructure and proprietary systems consumed financial resources and talent that GE preferred to redirect to innovation in its core businesses. By relying on AWS’s cloud infrastructure, GE gained compute elasticity, access to advanced analytics and AI services, and a significant reduction in the time needed to deploy new applications.

From this article’s perspective, the GE case shows how an industrial incumbent can reposition itself within the new invisible‑infrastructure map. Instead of competing to build its own cloud, GE chooses to “rent” that layer and focus on developing more differentiating capabilities on top of it, such as connected industrial solutions or advanced analytics for equipment. The trade‑off is clear: it gains speed and flexibility but assumes dependency on a cloud provider for critical functions.

Sacyr: from internal platform to competitive lever

Sacyr, a global construction and concessions group, developed GeOS as a digital platform to improve project management: planning, monitoring, cost control, and team coordination [2]. Over time, GeOS extended to other areas of the company, becoming a transversal infrastructural layer that centralizes operational data and allows real‑time visualization of multiple projects’ status. This evolution transformed how the company collaborates internally and makes decisions.

The Sacyr case illustrates how infrastructure initially conceived as an operational tool can become a strategic asset. By standardizing processes and data on a single platform, Sacyr not only gains efficiency, but also creates a base on which it can offer new services, collaborate more effectively with partners, and potentially open its digital capability to third parties. GeOS functions, in practice, as the “invisible layer” that coordinates a complex network of projects and actors.

Crossmint: AI‑enhanced blockchain infrastructure for third parties

Crossmint has built an “all‑in‑one” blockchain platform with AI integration that offers digital wallets, payments, and APIs for AI agents or other applications to operate in secure environments [4]. Its infrastructure is used by more than 40,000 companies, including global brands such as Adidas and Red Bull, which integrate it to manage digital assets, payments, and web3‑related experiences without internally developing the complex technological stack required [4].

In this case, invisible infrastructure not only simplifies blockchain use for companies but also reconfigures the value chain of digital products. Crossmint captures value as a B2B2X infrastructural provider, while end brands control user relationships and experience design. Integrating AI into the infrastructure layer strengthens its ability to automate flows, detect anomalies, or personalize interactions, consolidating its position as an “enabling layer” in a technologically complex field.

Limitations

This analysis relies on qualitative cases and a selection of sectors — finance, retail/e‑commerce, healthcare, and mobility — which, while illustrative, do not exhaust the landscape of invisible infrastructure. Industries such as energy, advanced manufacturing, or agri‑food are also undergoing relevant infrastructural transformations, often supported by technologies such as industrial IoT or advanced analytics, which are not addressed in depth here.

Likewise, the quantitative data used focuses on specific metrics — for example, Nike’s 36% increase in digital revenues in the last quarter of 2020 [3] or the number of applications GE migrated to the cloud in 2017 [1] — but do not constitute a broad statistical study. It would be necessary to complement this approach with systematic analyses of B2B2X infrastructure adoption by region, company size, or sub‑sector to obtain a more granular view.

Finally, the conceptual framework of “three infrastructure layers” simplifies a reality in which the boundaries between business, technology, and experience are blurred. In practice, technological decisions condition the business model and vice versa, and extended‑experience solutions integrate elements of both. Moreover, the sources on regulation and startups, although current [5][6][7], reflect dynamic environments that can change rapidly, so regulatory conclusions should be periodically reviewed.

Implications

For incumbents, the main message is that competition is no longer decided solely on the visible surface. Conducting an honest inventory of their own invisible stack — which systems they own, which external infrastructures they use, what data they control, and how they orchestrate it — is an unavoidable first step. Based on this diagnosis, organizations must decide which parts of the stack they want to continue controlling (because of strategic or regulatory importance) and which parts are more efficient to rent from startups or large infrastructural providers. Cases such as GE and Sacyr show that externalizing generic infrastructure (data centers, management tools) can free up resources to innovate in differentiating capabilities [1][2].

However, externalization brings risks: dependence on third parties, possible contractual lock‑ins, exposure to security vulnerabilities, or loss of proximity to the end user. Designing agreements that preserve access to data — for example, by establishing portability and own‑analytics rights — and maintain some control over the experience is key to avoiding unintended “commoditization.”

For infrastructural startups, the challenge is symmetrical. Focusing solely on APIs can turn them into easily replaceable “commodities.” Building defensible advantages requires combining several elements: proprietary data that improve with use, network effects among clients, deep integration into critical workflows, and, in some cases, regulation itself as a barrier to entry [5][7]. They must also decide whether to compete directly with incumbents by launching B2C products on top of their own infrastructure, limit themselves to enabling them through B2B/B2B2C models, or embrace a hybrid model. This decision shapes both their technical architecture and their regulatory exposure and talent needs.

Conclusion

The deepest competitive revolution of the last decade is not happening on user screens, but in the underlying layers of invisible infrastructure. APIs, data platforms, AI services, BaaS, fulfillment as a service, healthcare interoperability, or MaaS are redefining how value is designed, delivered, and monetized across multiple industries. In this context, the real comparison is no longer just startup vs. traditional company, but modern modular stack vs. rigid legacy stack.

This invisible infrastructure redraws business models by enabling the shift from final products to enabling capabilities consumed as a service; it makes possible user experiences that were previously unfeasible — such as embedded banking or multimodal door‑to‑door trips; and it alters the balance of power within each sector, granting growing influence to those who control orchestration and data layers. Companies like GE, Sacyr, Nike, Crossmint, Steryon, DefAgent, or Multiverse Computing show that both incumbents and startups can become winners if they choose their role in the stack wisely [1][2][3][4].

For leaders of traditional organizations, the call to action is clear: map their invisible stack, decide what to rebuild, what to connect, and how to position themselves in the new map — as final product, as infrastructure, or as ecosystem orchestrator. Ignoring this silent revolution will not stop its progress; it will only raise the risk that others define, from the shadows, the experience their customers already consider the minimum standard.

References

[1] INESDI, "Cloud computing: principales proveedores y casos de éxito (caso GE)", https://www.inesdi.com/blog/cloud-computing-principales-proveedores-y-casos-de-exito/

[2] CIO.com, "Digital Infrastructure Summit: telón de fondo de las mejores prácticas empresariales para la infraestructura TI del mañana (caso Sacyr-GeOS)", https://www.cio.com/article/3851505/digital-infrastructure-summit-telon-de-fondo-de-las-mejores-practicas-empresariales-para-la-infraestructura-ti-del-manana.html

[3] Vorecol Blog, "Casos de estudio de empresas que han logrado una exitosa adaptación a la transformación digital (caso Nike)", https://blogs-es.vorecol.com/articulo-casos-de-estudio-de-empresas-que-han-logrado-una-exitosa-adaptacion-a-la-transformacion-digital-818

[4] Cinco Días (El País), "Wayra Telefónica invierte en la startup Crossmint para su blockchain con IA", https://cincodias.elpais.com/companias/2025-05-26/wayra-telefonica-invierte-en-la-startup-crossmint-para-su-blockchain-con-ia.html

[5] Realidad Económica, "El impacto de las regulaciones gubernamentales en las startups", https://www.realidadeconomica.es/el-impacto-de-las-regulaciones-gubernamentales-en-las-startups/26292

[6] Avance Digital (Gobierno de España), "Respuesta de la Cámara de Comercio de EE. UU. a la Estrategia Digital (barreras al crecimiento internacional de startups)", https://avance.digital.gob.es/es-ES/Participacion/RespuestasEstatregiaDigital/camara-comercio-eeuu.pdf

[7] Startups Españolas, "Salud digital en España 2025: el laberinto entre la innovación y la burocracia", https://startups-espanolas.es/2025/06/11/salud-digital-en-espana-2025-el-laberinto-entre-la-innovacion-y-la-burocracia/