The missing file: where is value really lost between giants and startups?

The Hook: The day everyone showed their dashboards and no one talked about the patient

In a windowless boardroom, five executives are discussing what to do with a healthtech startup that wants to partner with the hospital.

On the screen, the executives show pristine charts: cost per bed, occupancy rate, return on assets. On the startup CEO’s laptop, another dashboard glows: daily active users, churn, acquisition cost per channel.

In two hours of meeting, no one utters the word “pain” a single time. Neither physical nor emotional. No one describes what it feels like to be the patient who waits three months for an in‑person appointment that a video call could resolve in two days.

From a behavioral psychology standpoint, that scene is a perfect crime: there are motives (growth, protecting the business), there are instruments (technology, capital, brand) and there are victims… but the main body of evidence is missing: the real perceived value for the person at the end of the chain.

This article is the autopsy of that lost value.

Genesis: How we ended up confusing “doing more” with “adding more value”

Before comparing sectors and models, it’s worth understanding the mental script repeated by both giants and startups.

Three recurring scenes

1. Traditional companies: decades optimizing internal processes, entrenched hierarchies, strict regulatory frameworks. Dominant script: “We can’t fail.” Psychological outcome: extreme risk aversion, status quo bias and an attachment to legacy systems that no longer reflect how users think or behave.

2. Startups and scaleups: flat structures, venture capital, aggressive growth metrics. Dominant script: “We can’t slow down.” Psychological outcome: action bias (doing something always seems better than waiting), unrealistic optimism and a tendency to prioritize usage metrics over impact metrics.

3. Regulators and the market: they demand safety and speed at the same time. Contradictory script: “Be innovative, but don’t make mistakes.” Result: cognitive overload in leadership teams, who respond with innovation rituals instead of deep changes.

The context you bring already outlines the structural differences:

• Business models: incumbents with diversified revenues, high fixed costs, trust- and reputation-based value propositions; focused startups with subscription or transaction-based revenue and an agility narrative.
• Technology: legacy systems and limited digitalization vs. cloud, data, AI and automation enabling fast cycles and short time‑to‑market.
• User experience: rigid, low‑personalization processes vs. omnichannel experiences, reduced friction and personalization.

But what interests a behavioral psychologist is not only what they do, it’s why they persist in patterns that destroy value even when the data says otherwise.

The Invisible Conflict: the bias giants and startups share

In almost every sector we’ll look at (finance, retail, health, mobility, education and manufacturing) the same paradox repeats:

• Incumbents have scale, resources, licenses and a loyal customer base.
• Startups have speed, technological focus and superior UX.

And yet, the potential value that could be created by combining the best of both rarely materializes.

The real suspect: the self‑relevance bias

Both giants and startups fall into the same cognitive trap: overestimating the importance of their own model and underestimating the human context.

• Banks that believe the customer “values solvency” when what they actually remember is whether they could solve an urgent problem on a Sunday night.
• Neobanks that assume a smooth app equals trust, ignoring that in high‑perceived‑risk decisions, users look for signals of stability beyond a clean interface.
• Hospitals that invest millions in buildings and equipment, while the patient’s emotional memory centers on a form they had to fill out three times in one day.
• Healthtechs obsessed with app adoption, but barely integrated into the patient’s real clinical journey.

The crime is not technological; it’s psychological. “Value” doesn’t disappear for lack of APIs, but because nobody is willing to give up their favorite narrative.

The Crime Lab: General framework for the forensic inspection

To organize this autopsy, we’ll use your three classic dimensions, but with a behavioral reading.

1. Business models: what is promised and how money is made

• Incumbents: diversified revenue, heavy cost structures, propositions based on solidity and broad coverage.
• Startups: focus on a specific problem or niche, recurring revenues (SaaS, subscription, pay‑per‑use), “grow first, monetize later” mindset.

Psychological trade‑off:

• Incumbents protect what exists (loss aversion): they see every change as a threat to their installed base.
• Startups maximize what’s potential (overconfidence bias): they downplay the risk of not finding a sustainable model.

2. Technology: how the machine is built

• Incumbents: legacy systems, technical debt, partial digitalization.
• Startups: cloud‑first, modular architecture, intensive use of data, AI and automation, short time‑to‑market.

Psychological trade‑off:

• Incumbents confuse “robust” with “immutable”.
• Startups confuse “new” with “better”.

3. User experience: what the user lives, not what the committee approves

• Incumbents: legacy processes, friction during onboarding, incomplete omnichannel, bureaucratic after‑sales service.
• Startups: UX design focused on reducing effort, data‑driven personalization, digital channels as the main gateway.

Psychological trade‑off:

• Incumbents prioritize internal consistency over external fluidity.
• Startups prioritize experimentation speed over perceived stability.

With this framework, let’s move to the crime scenes, sector by sector.

Scene 1: Financial services – Trust versus friction

Your data clearly marks the contrast: traditional banks relying on fees and interest, legacy systems and clunky UX; neobanks like N26 or Revolut with 100% digital services, smooth UX and lower fees.

From psychology’s point of view, the question is: why do so many customers keep their main account at a traditional bank despite that UX gap?

Typical state of traditional banks

• Business model: interest income, fees, cross‑selling; wide range of services.
• Technology: old core banking systems, limited or partial digitalization, but reinforced by strong security layers.
• UX: high friction in onboarding and complex operations; physical branches acting as trust anchors.

Typical approach of neobanks and fintechs

• Business model: low or zero fees, focus on payments, cards, remittances or microloans; monetization via usage or premium subscriptions.
• Technology: cloud infrastructure, open APIs, extensive data use for scoring and fraud prevention.
• UX: onboarding in minutes, real‑time notifications, clear visualizations of spending and saving.

Table 1 – Forensic scorecard: Finance

Psychological angle: the “crime” here is treating trust as static. Banks see it as a historical asset; fintechs as a by‑product of UX. Neither treats it as an emotional experience built in each critical micro‑moment (claims, frauds, errors).

Scene 2: Retail and e‑commerce – The endless aisle versus the “buy now” button

Your context already points to it: large retailers with incomplete omnichannel, struggling to integrate physical and online; startups (and born‑digital giants) betting on fast delivery and personalization.

Typical state of traditional retail

• Business model: tight margins, massive volume; revenue from direct sales, brand agreements.
• Technology: old POS systems, fragmented ERPs, e‑commerce as an “add‑on”.
• UX: decent in‑store experience, uneven online; click‑and‑collect with friction.

Typical approach of e‑commerce and quick commerce

• Business model: marketplaces, commissions, subscriptions (e.g. premium delivery), on‑demand models.
• Technology: scalable platforms, recommendation engines, data‑driven logistics optimization.
• UX: purchase in a few clicks, order tracking, personalized suggestions.

Table 2 – Forensic scorecard: Retail/E‑commerce

Psychological angle: the crime here is channel myopia. Incumbents still think in “online vs. store”; startups in “psychological moment” (impulse, urgency, repeat). Whoever better understands the customer’s emotional state in each purchase owns the relationship, not the one with more square meters.

Scene 3: Health – The hospital that talks about beds and the app that talks about clicks

The information you provide underlines this: hospitals with fragmented data systems and healthtechs focused on telemedicine and online booking.

Typical state of hospitals and clinics

• Business model: fee‑for‑service, insurance agreements, public or mixed funding.
• Technology: fragmented medical records, systems that don’t “talk” to each other, low use of data for experience management.
• UX: multiple forms, long waits, unclear communication.

Typical approach of healthtechs

• Business model: B2B2C subscriptions, pay‑per‑use, platform models between patients and professionals.
• Technology: secure cloud platforms, telemedicine tools, analytics to optimize schedules and triage.
• UX: quick booking, automatic reminders, virtual appointments.

Table 3 – Forensic scorecard: Health

Psychological angle: the crime in health is maximally depersonalizing the person. The traditional system reduces them to a clinical case; the startup to a “user”. Both forget their emotional experience in processes full of anxiety and vulnerability. Real value will keep leaking away as long as the main metric is bed occupancy or MAUs instead of adherence, understanding and feeling of control.

Scene 4: Mobility and transport – Optimized routes vs. messy lives

Even though your context didn’t elaborate on it, the broad pattern holds: traditional transport operators vs. mobility startups.

Typical state of traditional operators

• Business model: ticket sales, public subsidies, service contracts.
• Technology: old ticketing systems, low real‑time data integration.
• UX: rigid schedules, unclear information, slow incident resolution.

Typical approach of mobility startups

• Business model: platform intermediation (ride‑hailing, micromobility), dynamic pricing.
• Technology: mobile apps, geolocation, vehicle assignment algorithms.
• UX: booking in seconds, real‑time tracking, integrated payments.

Table 4 – Forensic scorecard: Mobility

Psychological angle: the crime here is assuming the user’s time is cheap. Traditional operators talk occupancy rates; startups talk ETA. Neither fully embraces that the truly scarce resource is the attention and mental energy of citizens juggling remote work, errands and fragmented lives.

Scene 5: Education – Printed diplomas vs. push notifications

Traditional universities and new edtech players replay the same basic conflict: long time horizons vs. instant gratification.

Typical state of universities and formal institutions

• Business model: tuition, fees, public funding, research.
• Technology: virtual campuses with limited functionality, manual admin processes.
• UX: rigid programs, little personalization, recurring bureaucracy in enrollment and procedures.

Typical approach of edtech startups

• Business model: subscriptions, on‑demand courses, intensive bootcamps.
• Technology: cloud platforms, streaming, progress analytics.
• UX: short content, asynchronous interaction, quick feedback.

Table 5 – Forensic scorecard: Education

Psychological angle: the crime in education is ignoring the human motivational system. Universities ignore the need for frequent reinforcement and specific feedback; edtech ignores the importance of identity, belonging and social signalling that degrees provide.

Scene 6: Manufacturing and Industry 4.0 – Machines speaking different languages

In manufacturing, the contrast centers on data use and automation.

Typical state of traditional manufacturers

• Business model: physical product sales, maintenance contracts, industrial margins.
• Technology: robust machinery, old SCADA and MES systems, partial IoT adoption.
• UX: the “user” is mostly B2B; experience centered on classic commercial relationships.

Typical approach of industrial startups and Industry 4.0

• Business model: monitoring platforms, industrial SaaS, “equipment‑as‑a‑service” models.
• Technology: IoT sensors, advanced analytics, AI for predictive maintenance.
• UX: intuitive dashboards, proactive alerts, integration with modern ERPs.

Table 6 – Forensic scorecard: Manufacturing / Industry 4.0

Psychological angle: the crime in industry is underestimating shop‑floor tech‑change aversion. Executives buy 4.0 solutions; operators passively sabotage them if they perceive threats to their role or added cognitive complexity.

Cross‑sector patterns: the same modus operandi everywhere

Comparing these scenes, clear patterns emerge.

Where startups usually win

• User experience: they better process cognitive cost for users and design to minimize it.
• Iteration speed: their structures and technology let them correct course quickly.
• Data & AI use: they turn every interaction into input to improve the product.

Where incumbents still dominate

• Regulated market access: banking, health, education licenses, transport concessions.
• Shock absorption capacity: large balance sheets, diversification, positive cash flows.
• Institutional trust: in high‑perceived‑risk decisions, users choose signals of solidity.

The role of regulation and entry barriers

• Regulation as shield: it protects incumbents, but also anchors them to outdated models.
• Regulation as grey area: many startups grow in lightly regulated spaces until their scale forces new rules.
• Cognitive barriers: even when formal barriers drop (open APIs, sandboxes), mental ones persist: fear of cannibalization, “not invented here” syndrome, founder ego.

From a behavioral perspective, the real barrier to entry isn’t in the law or the core system; it’s in leaders’ willingness to tolerate psychological discomfort while they transform their model.

Case file: credible, non‑heroic examples

Without turning this into a unicorn gallery, some patterns fit the context you shared:

• Neobanks (e.g., N26, Revolut): forced banks to improve apps, reduce visible fees and speed up account opening. Yet many still aren’t users’ main bank, reflecting the gap between UX and structural trust.
• On‑demand retail (e.g., Glovo): raised the bar on delivery time expectations. Traditional retailers reply with faster click‑and‑collect and logistics partnerships, but suffer on profitability, revealing the real cost of extreme convenience.
• Healthtech (e.g., platforms like Zocdoc or Teladoc): normalized online booking and telemedicine. Hospitals respond with patient portals, but partial integration keeps the experience fragmented.
• Edtech (e.g., bootcamps and course platforms): redefined how long it “should” take to learn something applicable. Universities create micro‑credentials and flexible executive programs, but carry structures built for four‑year degrees.
• Industry 4.0: integrators and IoT startups have pushed manufacturers to open plant data and think in terms of service, not just product.

Each case shows the same dynamic: the startup changes users’ psychological expectations; the incumbent reacts partially, changing the surface without questioning the core narrative.

Strategic twist: from “beating” the other side to reducing value leakage

If we treat this context as a criminal file, the question isn’t who wins between startups and incumbents, but how much value is lost in the friction between both worlds.

As a behavioral psychologist, the proposal isn’t a new framework, but a shift in mental focus.

1. Redefine what gets audited

Boards audit financials, legal risks and cybersecurity. They rarely audit:

• Unresolved user pain: journey moments where the dominant emotion is frustration, anxiety or distrust.
• Internal cognitive friction: decisions made to protect the organization’s mental comfort, not the customer.

A “Forensic Value Audit” should map these two dimensions before deciding any startup collaboration.

2. Change the questions the executive committee asks

Instead of asking:

• “How many innovation projects do we have?”
• “How many startups are in our accelerator program?”

The board should ask:

• “At which key points in the customer journey are we still generating more mental effort than our digital competitors?”
• “Which internal rules exist only to reduce our fear, not real risk?”

3. Design collaborations that reduce biases, not just costs

Many corporate‑startup initiatives focus on saving time or money. A smarter design would also ask:

• Which corporate biases can the startup offset? (external lens, speed, creativity)
• Which startup biases can the corporation offset? (risk discipline, long‑term view, regulatory reading)

Instead of acquiring startups only for their tech, acquire them also as an organized antidote to internal biases.

4. Move from PoCs to “reality tests”

From psychology’s perspective, PoCs (Proofs of Concept) are often rituals that confirm the organization’s pre‑existing beliefs. A more honest alternative would be Behavioral Reality Checks (PRC):

• Experiments that observe not just whether the tech works, but how actual behaviors change: decision times, abandonment rates, explicit and implicit complaints.

The Mental Blueprint for an incumbent that doesn’t want self‑deception

Here’s a simple but uncomfortable scheme any board can use.

Step 1: Crime map

• Draw the full journey for your customer in a key product or service.
• Mark in red the points where users churn, complain or seek external alternatives (including startups).
• Explicitly ask: what emotion dominates at each point?

Step 2: Internal suspect profile

For each red point, answer:

• Which internal process, policy or legacy tech sits behind this friction?
• What specific fear keeps us from changing it? (regulatory, reputational, loss of internal power, income uncertainty)

Naming the fear reduces its power. As long as risk remains vague, it always beats any concrete initiative.

Step 3: External accomplice profile (startup or scaleup)

Instead of seeking “the most innovative startup”, seek:

• Who has already proven they can change user behavior in the type of friction we have?
• To what extent does their model correct or amplify our biases?

A good match isn’t the most glamorous one, but the one that balances mutual limitations.

Step 4: Psychological, not just legal, contract

When designing the collaboration:

• Set clear expectations on decision speed, success criteria and learning sharing.
• Make sure someone in your organization is explicitly mandated to defend the user’s perspective, even against the immediate interests of both partners.

Step 5: Ongoing audit of perceived value

• Define a few deeply human metrics: perceived effort, sense of control, clarity, trust.
• Review quarterly where life has truly improved for the user, not just internal efficiency.

This blueprint doesn’t replace strategy, but it can protect you from the most dangerous illusion: believing a new innovation lab equals a new way of thinking.

Big picture: what if true disruption is about learning to let go?

Across these sector files, the public conversation looks trapped in a false dichotomy: either startups win or giants win.

From psychology’s point of view, that narrative reflects a zero‑sum bias that rarely fits reality.

• Incumbents will have to give up some narrative control: stop being “the institution” and become a platform where others add value.
• Startups will have to give up some ideological purity: accept regulatory compromises, slower rhythms in critical areas and metrics that go beyond growth.

In any forensic investigation there’s an uncomfortable moment when someone must admit the crime wasn’t committed by an external villain, but by everyday, reasonable decisions… accumulated over years.

The value lost today between giants and startups doesn’t vanish for lack of ideas or capital, but because of a shared inability to change mental habits.

The question that remains for any board is not which startup to partner with or which new technology to try, but something simpler and harder:

Which belief about who we are as an organization are we willing to abandon so we stop being accomplices in this crime?

Until that answer is clear, any innovation initiative—no matter how brilliant it looks in a pitch or a committee—will be just another scene in the open case file of missing value.

References

1. 49k.es. “¿En qué se diferencian las startups tecnológicas de las tradicionales?” (consulted via provided context).
2. Wadhwani Foundation. “En qué se diferencian las startups tecnológicas de las tradicionales” (consulted via provided context).
3. Comparative contextual analysis provided in the conversation: differences between traditional companies and startups in business models, technology use and user experience in financial services, retail/e‑commerce and health.