When the form kills the pitch: why a single digital flow decides who wins between giants and startups

The Hook: the form that decides a unicorn’s fate

It’s 11:47 p.m. on a Sunday. A user in Mexico City is trying to open an account with a “revolutionary” fintech. She comes from a social media campaign, dazzled by a video explaining how generative AI will give her personalized financial advice and how the subscription model will save her from the “abusive” fees of traditional banks.

Three minutes later, she’s facing a form that asks for: a selfie with her ID, a proof of address in PDF, a personal reference, a payroll file in CSV, and acceptance of a contract that doesn’t even fit on a mobile screen. She drops off at field number 7.

The next day, exhausted, she crosses the street and enters a traditional bank branch. She takes a number, waits 40 minutes, signs paper documents and walks out with her account opened. The fintech, with its cloud‑native architecture, AI/ML, and gospels about the circular economy, has just lost a customer for the same reason the bank almost did: a poorly resolved high‑stakes flow.

What we’re being sold as a “war of business models,” “ecosystem hybridization,” and “exponential disruption” is being decided at a microscopic point: the moment when the user’s hand trembles between continuing to fill in the form or closing the tab.

That micro‑instant—a form, a payment, a complaint—is the minimal element I’m going to use to rewrite the entire story of incumbents vs. startups in banking, retail, healthcare, mobility, education, and manufacturing.

Genesis: how we ended up measuring the revolution with the wrong metrics

Over the past decade, both traditional industries and startups have filled their annual reports with big words:

• Generative AI to redesign drugs and personalize treatments, as already seen in pharma according to PwC.
• Subscription and freemium models as the holy grail of recurring revenue, from fintech to edtech.
• Circular economy with promises to be “100% circular” by 2030, like IKEA’s commitment.
• Digital platforms that integrate products and services into ecosystems, monetizing data and network effects.
• B Corps that claim to align social and environmental impact with profitability.

Meanwhile, some macro data sets the backdrop:

• The cryptocurrency market in Q2 2023 grows only 0.14% in total market cap, while Bitcoin and Ethereum rise ~6–7%. Lots of noise, little tide.
• The real estate market in 2023 keeps rising in price despite fewer mortgages and fewer transactions: the Housing Law doesn’t correct scarcity and only increases social tension.
• The smartphone market drops 12% year‑on‑year in Q1 2023, even as giants like Samsung and Apple concentrate more than 40% market share.
• The telecommunications sector, by contrast, grows 5.6% in 2023 and a single player concentrates almost 59% of revenues.
• The labor market in Latin America and the Caribbean shows persistent segmentation and inequality across 28 countries, despite economic differences.

The dominant narrative is clear: technology, platforms, and subscriptions are reshaping entire sectors. Banks vs fintechs, brick‑and‑mortar retail vs ecommerce, hospitals vs healthtech, transport operators vs mobility platforms, universities vs edtech, factories vs Industry 4.0.

But when you look at operational data—not the PowerPoint—an uncomfortable pattern emerges: major changes don’t fail for lack of AI, venture capital, or “circular economy.” They fail at the simplest interface. Where one extra data point is requested, where a flow doesn’t account for the reality of the user or the regulator.

The BMW–startup collaboration in its Garage is celebrated because it manages to bring innovation into production. The Blockbuster–Netflix case is a classic because the lack of vision on digital buried a giant. Yet in both stories the minimal piece that materializes the vision is ignored: how a customer enters, uses, and exits the service without losing their mind.

That flow, that form, is the cell I’m going to dissect sector by sector.

The Invisible Conflict: nobody wants to talk about friction, only about features

Let’s look at what’s being discussed today in boardrooms and pitch decks:

• Banks and fintechs talk about open banking, BaaS, regulated cryptoassets.
• Retail and ecommerce debate omnichannel, marketplaces, last mile, circular economy.
• Healthcare is wrapped in rhetoric about telemedicine, clinical IoT, blockchain for medical records.
• Mobility is sold with dynamic pricing, matching algorithms, shared fleets.
• Education drapes itself in edtech, AI tutors, à‑la‑carte courses, digital credentials.
• Manufacturing is being rebuilt with IoT, digital twins, Industry 4.0.

The conflict almost nobody talks about is brutally simple:

Most innovations don’t fail because of the business model, but because users can’t stand the process to access the supposed value.

That process translates into:

• An onboarding flow.
• A recurring payment.
• A complaint or cancellation.
• A change of plan, rate, course, or doctor.

The optimistic story says: “startups win because they reduce friction and corporations lose because they’re bureaucratic.” The cynical story says: “incumbents survive thanks to regulation, capital, and distribution networks, even if their UX is from another era.” Both stories are too coarse.

My contrarian thesis: the real axis of comparison is no longer startup vs. incumbent, but tolerable flow vs. unbearable flow. And that axis cuts across all sectors.

Evidence and Insights: six sectors seen through a single flow

1. Financial services: onboarding and payments that don’t forgive mistakes

Business models.

• Traditional banks live off interest and fees in B2C and B2B models, heavily leveraged on physical assets and strict regulations.
• Fintechs play in B2C and B2B2C, with subscriptions, freemium, pay‑per‑use, often asset‑light and built on cloud infrastructure.

Technology.

• Banks: legacy systems, monoliths, long release cycles, and compliance as a justification for every delay.
• Fintechs: microservices, cloud‑native architectures, open APIs, AI/ML for scoring and real‑time personalization.

User experience.

Fintechs do simplify interfaces. But KYC/AML regulations don’t go away. The conflict:

• If they copy banks, they replicate absurd forms.
• If they ignore regulation, they end up fined or shut down.

The winner will be the one who designs a regulated onboarding flow that feels like a three‑minute chat, not a police interrogation.

Pattern the reports ignore: a bank with bad systems can camouflage them with in‑person processes and humans who “fill out the form for you.” A fintech can’t; all its value is played out on a 5‑inch screen.

2. Retail / ecommerce: the cart that turns into a ghost

Business models.

• Physical retail: direct sales, frequent vertical integration, CAPEX in stores and inventory.
• Ecommerce and marketplaces: commissions, advertising, logistics services, B2C and B2B models built on platforms.

Technology.

• Incumbents: heavy ERPs and CRMs, sometimes obsolete.
• Startups: APIs, public clouds, constant A/B testing, dynamic pricing.

User experience.

The narrative says ecommerce is “naturally superior” to physical retail. But look at the concrete flow: checkout.

• In a physical store, payment might be clumsy, but the customer has already invested time in getting there; their abandonment threshold is high.
• In ecommerce, a five‑step form and a single CVV error are enough to lose the sale.

With a global smartphone market dropping 12% year‑on‑year in Q1 2023, but still dominated by a few manufacturers, any poor mobile optimization magnifies losses: millions of users on small screens, expensive mobile data, and little patience.

Startups boasting about AI recommendations and circular logistics can still lose on the basics: a payment that fails once on 3G in an outlying neighborhood. Traditional retail, clumsy as it may be, still collects cash without lags or frozen screens.

3. Healthcare: the paperwork that separates promise from treatment

Business models.

• Hospitals: direct services, insurance, heavy dependence on physical infrastructure.
• Healthtech: subscription, pay‑per‑use, digital platforms that reduce CAPEX and promise remote access.

Technology.

• Many hospitals: legacy systems, fragmented electronic health records.
• Startups: IoT, telemedicine, even blockchain for data traceability.

Generative AI is already used in pharma to predict molecular responses and personalize treatments. But the distance between that paper and the patient goes through a minimal flow: booking an appointment and getting to the consultation without losing their mind.

• Traditional hospital: queues, calls, paper forms, long waits.
• Healthtech: clean UX apps, but often disconnected from insurers, prescribing doctors, and hospital systems.

Common results:

• The patient ends up printing out PDFs from the app to take them to the hospital, where they’re re‑digitized by hand.
• The flow breaks at the edge between digital platform and traditional system, precisely where it hurts most.

A digital twin of an organ is useless if the patient drops out at the app registration form because they’re asked for a data point their insurer never gave them.

4. Mobility/transport: dynamic pricing vs. impulsive cancellations

Business models.

• Traditional operators: B2C, fixed fares, owned fleets, asset‑intensive.
• Mobility platforms: asset‑light models, dynamic pricing, subscriptions, per‑ride commissions.

Technology.

• Operators: fleet management systems, often poorly optimized.
• Platforms: matching algorithms, real‑time route optimization, mobile apps with integrated payments.

The microscopic flow that rules here is requesting a ride in non‑ideal conditions:

• Spotty network coverage.
• A tired user, in the rain or in an urgent situation.
• Confusing surge pricing.

Platforms congratulate themselves for measuring arrival times, occupancy, price elasticity. Almost nobody measures the critical metric: how many users abandon the app on the last step because the dynamic pricing interface generates distrust.

Traditional operators, for all their rigidity, offer something rudimentary: you get in, you pay, you get off. The mental form is simpler than the slickest screen. Again, the differential is a flow: if the fare changes three times before confirmation, users hold a grudge even if they arrive faster.

5. Education: the course lost between passwords and enrollments

Business models.

• Universities and institutions: tuition fees, public funding, physical infrastructure.
• Edtech: subscriptions, à‑la‑carte courses, B2C and B2B2C, scalable and digital models.

Technology.

• Institutions: inherited LMSs, rigid platforms.
• Startups: cloud platforms, personalization, AI to adapt content.

The promise: accessible, personalized education. The real flow: enrollment and first session.

• Traditional institution: admission forms, separate payments, manual validation of documents.
• Edtech: fast sign‑up, but scattered steps: email, verification, placement tests, choice overload of courses.

In a regional labor market as segmented as Latin America and the Caribbean’s—where comparative studies show persistent structural inequalities—failing at the access flow to training isn’t a detail: it’s one of the bottlenecks that perpetuates the gap.

A course that promises to improve employability but requires five screens to choose a payment method and one more to accept policies loses exactly the person it can least afford to lose: the precarious worker with limited mobile data and scarce time.

6. Manufacturing / Industry 4.0: the work order nobody wants to touch

Business models.

• Traditional manufacturers: B2B, tight margins, long contracts, high CAPEX.
• Industry 4.0 startups: SaaS, pay‑per‑use, sensors, data platforms, adjacent consulting.

Technology.

• Incumbents: robust but rigid SCADA and ERPs.
• Startups: IoT, real‑time analytics, digital twins.

In telecom, for example, in 2023 revenues in Mexico grow 5.6% and a single player concentrates almost 59% of the market. Behind that concentration is something unglamorous: highly standardized internal procedures, flows that avoid costly errors.

In Industry 4.0, the minimal point is the digital work order:

• Who creates it?
• On which screen?
• With how many clicks can an operator modify a parameter without blowing up a production line?

Startups arrive with shiny dashboards and simulators, but if the creation or edit flow of an order is confusing, the entire plant boycotts the system. The operator prefers paper or the old Excel sheet because they know exactly where to write and what to cross out.

In critical systems, robustness isn’t optional: a poorly designed form on the shop floor is worth more than a hundred slides about AI.

Table 1 – The silent scoreboard: abandonment rate in the critical flow (operational hypothesis)

There are no universal statistics published for all sectors, but we can reason out a qualitative comparative scoreboard:

The paradox: startups optimize the interface, incumbents operate reality. The minimal touchpoint, the flow, determines who keeps the customer or the operator.

The Strategic Turn: think “micro‑friction” before “business model”

If we accept that the critical flow is the cell that decides success, strategy changes radically for both giants and startups.

1. Design business models from the flow, not the other way around

Financial services, retail, healthcare, mobility, education, and manufacturing have embraced subscription, pay‑per‑use, marketplaces, B2B2C models. But almost never are these models built from one brutally simple question:

What does a person have to do, minute by minute, for this model to work without hating us?

Examples of a changed approach:

• A fintech dreaming of freemium should start with the cancellation flow: if it’s hell, the regulator will show up before growth does.
• An edtech obsessed with generative AI should make its first metric “minutes from seeing the landing page to seeing the first useful content”, not courses sold.
• A healthtech charging subscriptions should assume erratic usage is the rule and design flows that don’t penalize absences.

2. Stop the feature obsession and measure regulatory friction

Regulatory conflict isn’t going away. Banking will remain supervised; healthcare and education will remain bound by heavy rules; manufacturing will keep dealing with safety standards.

The usual temptation:

• Startups: “we’ll own the pretty front end, the big partner can handle the back end and the regulator.”
• Incumbents: “we’ll shoulder regulation, the startup can dress up the service with a modern app.”

That division of labor creates schizophrenic flows where the user floats between two worlds. The regulatory friction rate should be a KPI:

• Number of times the user must provide the same data to different systems.
• Number of steps where they’re asked for something “just in case the regulator requires it” without an actual obligation.

3. Make the operator and internal client the “primary user”

In manufacturing, telecom, and large retailers, the user who decides the success of an innovation is not the end customer, but the employee who has to use the new system.

The Mexican telecom case, where one player concentrates almost 59% of revenues in a growing market, is explained partly by repeatable operational execution. Not by the prettiest app, but by processes that didn’t blow up every week.

Uncomfortable strategy for B2B startups:

• Before selling dashboards to executives, design the minimal work flow for the field technician or plant operator.
• Accept that in critical systems, simplicity beats visual spectacle.

4. Accept asset asymmetry as a friction advantage

The common line says: “physical assets are ballast and digital platforms are freedom.” But 2023 real estate data in Spain shows that with fewer mortgages and fewer deals, prices rise because housing supply is shrinking.

The lesson nobody wants to process: physical scarcity generates power even when UX is poor. A badly designed hospital, bureaucratic university, or clumsy transport operator can survive for decades simply because they control key assets: beds, classrooms, fleets.

If startups really want to challenge these players, apps and circular models aren’t enough: they have to attack the access flow to those assets (waiting lists, resource bookings, supply‑demand matching), where current friction is sky‑high.

Table 2 – Scorecard of who really wins the critical flow (qualitative view)

Beyond the binary: cross‑sector patterns seen from the lowest common denominator

From the “critical flow” cell, the big patterns everyone repeats look different.

Disintermediation and the platform effect

Startups brag about disintermediating: fintech vs banks, marketplaces vs retailers, edtech vs universities. But the minimal flow reveals something else: they often swap one intermediary for another rather than eliminating it.

• In crypto, for example, total market cap barely grows 0.14% in Q2 2023, while two assets capture much of the gain. The real flow today runs through centralized exchanges, KYC, and custodians: intermediation returns wearing a different jersey.

Platforms and ecosystems work when they cut access friction (single login, single payment method), but often they multiply it: multiple apps, accounts, passwords, and terms.

Technology as core vs support function

The mantra “technology must be core” is so often repeated it goes unquestioned. From the critical flow, the real question is different:

Who owns the key form: business, technology, or compliance?

• When tech alone designs it, the backend is optimized but the user’s emotions are ignored.
• When business leads, too many data points are requested “just in case.”
• When compliance dominates, the flow becomes a wall.

True technological maturity isn’t having AI, blockchain, or IoT, but shared governance over the critical flow.

Structural advantages that persist

• Scale, licenses, physical assets, and access to capital let incumbents tolerate friction that would kill a startup in months.
• Startups have less time and more pressure: their margin for error in the flow is near zero.

That’s why the most likely future isn’t one where “startups kill giants” or the reverse. It’s a hybrid field where:

• Giants buy or copy efficient flows.
• Surviving startups learn to live with legitimate friction (regulatory, safety, robustness).

The Big Picture: an operational framework for the next 5–10 years

An uncomfortable typology: classify by flow, not by marketing

Let’s think of incumbents and startups across three axes, evaluated by sector:

1. Business model innovation (low, medium, high).
2. Technological maturity (closed legacy systems vs open, secure architectures).
3. User experience quality in the critical flow (from “painful but predictable” to “smooth and trustworthy”).

Seen this way, four types emerge:

1. Hidden strengths: low model innovation, old tech, but an acceptable critical flow thanks to humans (e.g., an efficient bank branch, a small neighborhood clinic).
2. Fireworks: high model innovation, advanced tech, but a broken critical flow (e.g., a fintech with no regulatory integration, a healthtech disconnected from hospitals).
3. Titans in transition: medium innovation, heavy tech investment, flows under renovation (e.g., banks rewriting mobile onboarding, serious omnichannel retailers).
4. Radical craftsmen: high innovation, sober but well‑applied tech, near‑clinical obsession with a single flow (e.g., an edtech that perfectly nails access to first content, an Industry 4.0 startup centered on work orders).

If you’re an executive or founder, your task isn’t “be more digital” or “be more disruptive,” but move deliberately towards the radical craftsmen quadrant.

Five theses for the next decade

1. Competitive advantage will cluster around three flows per sector.
   Not the whole app or portfolio, but the three processes where most users have something at stake: onboarding, payment/complaint, and exit/cancellation.

2. Regulation will reshape UX as much as the market does.
   In financial services, healthcare, and education, UX will stop being “a pretty layer over the rules” and become an object of direct dialogue with regulators. Projects that involve authorities in co‑designing flows will survive better.

3. Incumbent–startup collaboration will be judged by shared flows, not press releases.
   Programs like BMW Startup Garage will matter only if they reach the point of redesigning work orders, customer forms, and operator screens. Alliances like Blockbuster–Netflix will keep failing as long as they add logos without touching processes.

4. Circular economy and B Corps will be credible only if they cut real friction.
   Recovering and recycling furniture, vehicles, or devices are noble promises, but what users experience is: how many steps to return something? Who picks it up, when, how? Sustainability without simple flows will remain marketing.

5. Generative AI will win or lose legitimacy at the moment it asks for the first data point.
   Whether a system recommends a drug, a course, or a product will matter less than how it asks permission to use data. Trust will be broken or cemented in the first field that appears under “I agree to personalization.”

References

1. Coingecko / Criptotendencia. “Análisis profundo del mercado de criptomonedas – Informe Q2 2023.” Marginal 0.14% growth in total market cap and 6.9% and 6.0% rises in Bitcoin and Ethereum.
2. Idealista. “Análisis del mercado inmobiliario en 2023 y previsiones para 2024.” Fewer mortgages and transactions, rising housing prices due to shrinking supply and limited effects from the Housing Law.
3. Canalys / ChannelNewsPeru. “Mercado global de smartphones muestra caída de 12% en primer trimestre de 2023.” Market share distribution among Samsung (22%), Apple (21%), Xiaomi, OPPO, and vivo.
4. The CIU / Milenio. “Sector telecomunicaciones creció 5,6% en 2023.” Revenues of 577.3 billion pesos in Mexico and 58.9% concentration in América Móvil.
5. UAB – Portal Recerca. “Para un análisis comparativo de las desigualdades sociales en el mercado de trabajo en América Latina y el Caribe.” Labor segmentation and persistent inequalities in 28 countries.
6. PwC. “Sectores donde tendrá impacto la IA.” Applications of generative AI in pharmaceuticals and other sectors.
7. Vorecol. “Cuáles son las tendencias emergentes en modelos de negocio para startups en 2023.” Circular economy, subscription models, platforms, B Corps, hybrid businesses, and focus on customer experience.
8. Emprender Fácil. “Cómo las startups están perturbando industrias tradicionales.” Use of blockchain and emerging technologies in new business models.
9. IESE Business School. “Guía para la colaboración entre corporaciones y startups.” BMW Startup Garage case and success factors in alliances.
10. EUCVC. “Case Studies in Corporate–Startup Collaboration.” Analysis of the Blockbuster–Netflix case and lessons about failure to adapt.