Letters from the Uncomfortable Lab: What in 2030 You Still Don’t Want to See About Giants and Startups

Letter 1 — I’m writing from the point where your Excel no longer reaches

Professional of 2030,

when you open this letter you’ll be in the middle of a crisis your innovation committee didn’t see coming. It won’t be a new unicorn. It won’t be a new regulation. It’ll be something more mundane: your customers will use your service… but they’ll check out emotionally long before they log out.

Your dashboard will show acceptable metrics: revenue growing at 4%, churn within the “tolerable” range, NPS moving between positive and negative digits like a tired ECG. And yet something will have broken: neither the giants nor the startups will really be winning.

What in 2024 you called “disruption” will have degenerated into an awkward draw: banks copying fintech interfaces, hospitals with apps nobody opens, traditional retailers with e‑commerce that looks like startups… and startups that, in order to survive, end up looking far too much like the giants they claimed they would replace.

You’ll keep receiving decks about omnichannel and platforms, but the really interesting data will be somewhere else: in what neither side —traditional industry and the startup ecosystem— wants to look at head‑on. That’s where I work.

Letter 2 — How we got to this false draw

Don’t fool yourself: you’re not living a war of heroes against dinosaurs. You got here through three simultaneous currents that intersected between 2010 and 2024:

1. Capital current: cheap money pushed startups to prioritize growth over sustainability, and incumbents to set up venture arms so as not to look obsolete.
2. Regulatory current: slow regulation protected giants in critical sectors (banking, health, regulated mobility) while leaving gaps where startups could experiment… as long as they didn’t touch the heart of the system.
3. User expectations current: users got used to smooth experiences in e‑commerce and mobility, and started demanding the same from banks, hospitals and public administrations.

The result was a theatre of convergence: giants digitizing processes without changing internal incentives, and startups sophisticating their financial models without fixing the fundamentals of the service.

I’m going to bring order to the chaos by sector. Not to tell you who’s winning, but so you can see an uncomfortable pattern: both sides compete on the visible (apps, campaigns, funding rounds) while ignoring the real battle in the invisible (cost structure, tech debt, user fatigue, regulatory trust).

Letter 3 — Fintech/Banking: when margin eats the story

3.1 Business models: revenues that look new but aren’t

By 2024 we already knew the basics: traditional banks living on interest, fees and ancillary services; fintechs betting on free accounts, cashback cards, “premium” subscriptions and transaction fees.

By 2030, you’ll see this:

• Banks will have copied subscription plans, but won’t have fixed their fixed cost structure (branches, compliance layers, legacy technology).
• Many fintechs that boasted of operating “lean” will be trapped in a chronic dependence on external capital, because the unit margin of many of their products (payments, cards, multicurrency) is too thin without massive scale.

The table almost nobody will show you in a committee is this:

Banks sacrifice agility to gain regulatory permanence. Fintechs sacrifice stability for acquisition speed. Both underestimate a variable your 2030 analysts will still be capturing poorly: account fatigue. Users can open six accounts in a year, but they only mentally trust one or two.

3.2 Technology: the myth of legacy versus the myth of the cloud as savior

In 2024 the refrain was: “banks are weighed down by legacy.” True, but incomplete. The uncomfortable question was: what happens when you partially migrate to the cloud without changing your decision‑making processes?

• Banks: core banking still on mainframe systems, layers of APIs and microservices on top, increasing use of public cloud for analytics and front ends. Barriers: zero‑risk culture, prudential regulation, reputational fear.
• Fintechs: architectures born in the cloud, heavy use of open APIs, AI/ML models for scoring and fraud detection, but with silent tech debt: architectural decisions taken in a rush in early stages.

Tech adoption is genuinely faster in fintechs, but not linear: at scale, many slow down more than a bank that gets a single core migration right.

3.3 User experience: the paradox of friction

Your 2030 users will have learned something they only intuited in 2024: too much friction is unbearable, but zero friction creates distrust.

• Traditional banks: longer processes, but perceived as more “serious” for complex products (mortgages, investments).
• Fintechs: onboarding in minutes, crystal‑clear interfaces, but a perception of fragility in crises (account blocks, outages, abrupt changes in terms).

Metrics like NPS, churn and retention will be used routinely, but few organizations will track the more revealing one: correlation between extreme simplicity and level of regulatory complaints.

Letter 4 — Health/Healthtech: brilliant tech on top of broken processes

4.1 Business models: when fee‑for‑service collides with subscription

In 2024, hospitals and insurers lived on fee‑for‑service: consultation, test, procedure, stay. Healthtech startups proposed subscriptions, digital pay‑per‑use, on‑demand teleconsultation.

Towards 2030, you’ll see this unresolved tension:

• Hospitals and clinics: still financing costly infrastructure, specialized staff, high‑CAPEX equipment. Growth through service expansion, hospital mergers, vertical integration (diagnosis + treatment + rehab).
• Healthtech startups: monetizing through B2C subscriptions, B2B2C deals with insurers and employers, pay‑per‑use models (teleconsultation, remote monitoring). More flexible variable costs, but very high patient acquisition cost.

The real clash is this: the traditional health system is optimized to bill “episodes”; many healthtechs are optimized to maintain a continuous relationship. The problem: the payer (insurer or public system) doesn’t adjust tariffs to continuous value as fast as the pitch decks promise.

4.2 Technology: scattered data, fragmented diagnostics

• Traditional sector: fragmented electronic health records, systems that don’t talk to one another, (justified) fear of medical data breaches. Slow adoption of clinical AI due to legal and regulatory risk.
• Healthtech: telemedicine, wearables, population‑level data analysis, diagnostic‑support algorithms, but almost always outside the transactional heart of the health system.

The pattern repeats: startups build experience and analytics layers around the core, but can’t get into the “central plumbing” where it’s decided what gets billed, to whom, and for how much.

4.3 User experience: easy access, broken journey

Your 2030 users will be able to book a video consultation in seconds and receive personalized wellness recommendations. But when they need complex surgery, they’ll return to the traditional ecosystem, with phone bookings, paper forms and waiting rooms.

• Healthtech startups measure NPS, retention and re‑use rates of wellness apps.
• Hospitals measure bed occupancy, waiting times and readmissions… but almost never the end‑to‑end patient experience.

This asymmetry kills any simplistic narrative that “startups improve health.” They improve pieces of the journey and create expectations the giants can’t meet without redesigning financial and clinical incentives.

Letter 5 — Retail/E‑commerce: when everyone knows how to sell but almost no one knows how to say no

5.1 Business models: margins versus speed

In 2024, classic brick‑and‑mortar retail was built on margin per product + volume, with high fixed costs (rent, staff, inventory). Digital‑native e‑commerce ran on D2C models, marketplaces, subscriptions and long tail, leveraging lower marginal costs.

By 2030 you’ll see this paradox:

• Large traditional retailers with online channels that are no longer secondary, but operating with mixed margin logics: physical stores require high gross margin, online tolerates lower margin in exchange for data and recurrence.
• E‑commerce startups that, after growth ramps driven by aggressive digital marketing and influencers, discover that profitability doesn’t arrive without adding the very frictions they once criticized: shipping fees, return limits, price hikes.

5.2 Technology: personalization that turns into noise

• Traditional retail: inherited inventory systems, old ERPs, AI pilots for demand forecasting and recommenders.
• Native e‑commerce: cloud stack, advanced analytics, big data, continuous A/B testing.

So far, nothing new. What your 2030 reports are beginning to admit is that hypertrophied personalization reduces signal:

• Too many irrelevant recommendations.
• Segmentations with apparent precision but based on poor data or increasingly limited cookies.
• Campaigns flooding every channel, eroding perceived value.

The tech advantage remains, but with diminishing returns: more AI models don’t equal more sustainable sales, but rather more operational complexity.

5.3 User experience: omnichannel or brand schizophrenia

Your 2030 customers will still want the obvious: to find, buy, return or exchange a product without thinking about the channel.

• Digitized traditional retail: struggles to integrate stock, returns and pricing systems. Typical friction: buy online, return in store… and discover the system doesn’t recognize the transaction.
• E‑commerce: flawless purchase flow, but post‑sales outsourced at rock‑bottom cost, with agents lacking contextual information.

Here’s another ignored metric: rate of invisible conflicts (problems the customer solves alone, with effort, without complaining). Startups celebrate high NPS among survey respondents; retailers console themselves with gross sales. Both neglect the silent erosion of trust.

Letter 6 — Mobility/Transport: logistical efficiency, social chaos

6.1 Business models: platforms versus assets

In 2024, airlines, public transport companies and logistics operators lived on per‑journey and related service fees, with high fixed costs (fleets, infrastructure, fuel, staff). Startups like Uber or Lyft presented themselves as sharing‑economy platforms, with assets in third‑party hands.

By 2030, you’ll see:

• Traditional operators that have introduced flat‑rate models, integrated digital passes, multimodal integration, but remain subject to public‑service regulation.
• Mobility platforms that have had to assume more responsibilities (safety, working conditions, local regulation), which brings their cost structure closer to that of incumbents.

What was sold as a “lightweight platform” is revealing itself as a partially disguised transport company, with all the labour, regulatory and safety issues… plus real‑time public judgment.

6.2 Technology: automation that doesn’t go as far as promised

• Traditional operators: experiments with IoT for predictive maintenance, advanced route‑planning systems, digital ticketing.
• Mobility startups: real‑time matching algorithms, dynamic pricing, polished apps, heavy use of geolocation data.

The myth was that technology would eliminate transport inefficiency. What you’re seeing in 2030 is that a great deal of algorithmic efficiency dissolves into physical and political bottlenecks: urban congestion, environmental regulation, social pushback.

6.3 User experience: from the promise of freedom to algorithm fatigue

Your 2030 users will know that moving around the city means:

• Platforms that offer you an “optimal” route in minutes, but change prices in seconds.
• Public services that are somewhat more digital, but with planning decisions that don’t respond to demand in real time.

User experience becomes a constant negotiation between price, time, comfort and principles (environmental footprint, driver conditions). App simplicity is no longer enough to mask structural tensions.

Letter 7 — The invisible conflict running through every sector

So far I’ve fulfilled your taxonomic brief: business models, technology, user experience in four industries. But if you stop there, you’ll repeat the mistakes of 2024’s analysts.

What unites fintech, healthtech, retail/e‑commerce and mobility is not disruption. It’s a double mismatch:

1. Mismatch between promise and cost structure.
2. Mismatch between user experience and regulatory/ethical framework.

7.1 Common patterns in business models

Startups, sector after sector, tend to introduce:

• Subscriptions where there used to be pay‑per‑use (premium accounts, health plans, mobility memberships).
• Marketplaces and platforms where there used to be bilateral relationships (banks–clients, clinics–patients, stores–buyers, taxis–passengers).
• Freemium models (basic service free, advanced features paid) to quickly gain user scale.

Incumbents, under pressure from competitors and investors, copy part of these models without adapting their internal incentives or their regulatory framework. Many end up with unstable hybrids: flat fees that don’t cover costs, marketplaces where they bear reputational risk but capture minimal margin.

7.2 Technological levers that repeat… and wear out

The disruptive technologies you’ve heard repeated as mantras since 2015 are three:

• Public cloud + open APIs: to build connected ecosystems.
• AI/ML: for scoring, personalization, process optimization.
• Mobile platforms: as the dominant user interface.

Yes, they’ve changed the game. But the marginal return of adding yet another layer of AI to a poorly designed process is approaching zero. The real bottleneck isn’t technical, it’s organizational and regulatory.

7.3 The “new minimums” of user experience

Startups have raised expectations to set new floor levels:

• Fast, self‑service onboarding.
• Real‑time status and notifications (payments, orders, trips, appointments).
• Support accessible from the same channel where the service is delivered.
• Coherent, simple design, without internal jargon.

Giants that don’t meet these minimums lose relevance. But startups that do meet them discover another limit: users don’t become loyal just because of a pretty interface when their sense of fairness, safety and consistency is threatened.

Letter 8 — Symmetrical risks: what each side prefers not to see

Here’s a table summarizing the pattern the public discourse hides:

Collaboration cases —corporate VC, accelerators, APIs and white‑label— have been sold as the perfect synthesis. The reality of 2030 is less romantic:

• Many collaborations are reputational theatre: pilots that don’t scale, POCs that live only in press releases.
• Others, fewer but important, achieve what neither can alone: giants contribute licences, regulatory solvency and customer base; startups contribute build speed, superior UX and data focus.

Tech integrations (banking APIs, health‑record connectors, white‑label mobility or payments) work when someone accepts the political cost of redesigning processes, not just connecting systems.

Letter 9 — The strategic turn you still haven’t made

This is where my lab diverges from the consultant chorus.

Your whole mental frame is focused on “how to be more like the other”:

• The bank wants to look like a fintech.
• The clinic wants to look like a health app.
• The retailer wants to look like a marketplace.
• The transport operator wants to look like a platform.

I’m proposing a more honest turn: specialize the convergence.

9.1 For traditional industry: what capabilities you need, not what logos to copy

If in 2030 you’re running a traditional organization and want to stay relevant in 5–10 years, forget the fetish of superficial “start‑up‑ification.” Your critical capabilities are:

1. Serious data governance:

   • No more decorative data lakes.
   • Lineage, quality, controlled access and teams who understand both business and statistics.

2. Modular technology architecture with hard decisions made:

   • You can’t indefinitely maintain three tech generations in parallel.
   • The political cost of migrating a core, closing branches, eliminating redundant products is high, but delaying it is worse.

3. Organizational design built around user journeys, not internal silos:

   • Teams responsible for an entire journey with an associated P&L.
   • Shared metrics across business, tech and operations.

4. Real capacity to experiment in a regulated market:

   • Regulatory sandboxes actually used, not just press‑release fodder.
   • Internal frameworks for “controlled failure” separating experimental from critical areas.

9.2 For startups: structural challenges your pitch avoids

If in 2030 you’re leading a startup that’s past the initial rounds, your threat is no longer “the giant.” Your threat is yourself as you scale.

The challenges that don’t fit in your slide deck are:

1. Maintaining excellent UX while your complexity explodes:

   • Every B2B integration, every deal with an incumbent adds layers of friction.
   • Users don’t care whether the problem comes from you or your corporate partner.

2. Avoiding the trap of tech debt and single‑vendor monoculture:

   • You built fast on a single cloud, a single core, a single critical provider.
   • At a certain scale, that dependency becomes an existential risk.

3. Shifting from a “hero team” culture to a reproducible system:

   • The founding talent doesn’t scale. You need processes without turning into useless bureaucracy.
   • The “we’re a family” mantra breaks when you have to make hard decisions in a crisis.

4. Taking regulatory and ethical responsibility before you’re forced to:

   • In fintech or healthtech, regulatory arbitrage only works for a while.
   • User trust isn’t bought with branding; it’s built with visible protection mechanisms (insurance, guarantee funds, clear incident protocols).

Letter 10 — The next cycle: convergence without clear winners

In 2024, the dominant storyline was simple: “either giants transform or they die”; “either startups grow fast or they disappear.” The data we’re seeing in 2030 suggests a different outcome: asymmetric convergence.

• You’ll see banks that are, in effect, regulated tech platforms providing white‑label services to fintechs, retailers and big tech, while keeping their brand only for high‑value segments.
• You’ll see hospitals becoming clinical infrastructure operators, while third parties provide the patient‑facing interfaces.
• You’ll see physical retailers repositioned as logistics and experiential hubs, ceding part of their direct transactional relationship to platforms.
• You’ll see mobility platforms that, pushed by regulation and economics, look more like classic transport operators, with unions, service obligations and modest margins.

No one will win outright. The real variable to watch will be who controls the critical decision point in the user journey:

• Who decides which loan to offer and on what terms?
• Who decides which specialist to assign and in what timeframe?
• Who decides which product to show first and at what price?
• Who decides which route is “optimal” and at what fare?

In many cases, it won’t be either the giant or the startup, but a middle layer of infrastructure and algorithms, sometimes in big tech’s hands, sometimes in sector consortia.

If you want agency in 2030, you’ll have to look beyond the “startups vs. traditional industry” story and ask yourself:

At what point in the system do I make decisions that change the user’s reality, and that no one else could make for me without changing the nature of the service?

That’s your strategic frontier.

Letter 11 — Final warning from the lab: what to measure differently tomorrow

I’ll close with something you can do even if your organization loves empty PowerPoints: change what you look at.

1. Stop comparing only growth rates between startups and giants. Start measuring model robustness:

   • Margin per user after marketing.
   • Model sensitivity to regulatory changes.
   • Dependence on 3–4 critical providers.

2. In UX/CX, don’t settle for NPS and churn. Add:

   • Silent conflict index (unreported problems detected through anomalies in usage).
   • Trust elasticity: how the user behaves after a serious failure (outage, error, data incident).

3. In technology, stop counting only projects and features. Track:

   • Average age of code in critical systems.
   • Real time and cost to modify an important business rule.

If you start watching these data in 2030, you’ll still be in time not to repeat 2024’s blindness: believing disruption was an aesthetic battle between pretty apps and ugly branches.

The real contest, as you now know, is over who bears the structural cost of keeping the promise made to the user.

I’ll leave it here. You’ll have to write the next letters yourself, from your own dissent.

References

1. Comparative analysis between traditional industry and the startup ecosystem in fintech/banking, health/healthtech, retail/e‑commerce and mobility/transport, focusing on business models, technology and user experience (research context provided).[1]
2. Highlighted cases: Revolut, N26 (fintech); Zocdoc, Teladoc (healthtech); Warby Parker, Glossier (e‑commerce); Uber, Lyft (mobility) as examples of scalable digital models, light in physical infrastructure and platform‑based.[2]
3. Contextual information on fintech business models (P2P lending, blockchain, big data) and the platform economy, with emphasis on network effects and personalized services.[3]
4. Digital transformation of traditional retail towards omnichannel and fast‑delivery models, exemplified by large players’ investments in e‑commerce capabilities.[4]
5. Cross‑cutting observations on user experience: intuitive interfaces, data‑based personalization, reduced friction and new market minimums established by digital startups.[5]