For years, the travel eSIM industry sold a beautifully simple promise: connectivity without complexity. Download an app. Choose a plan. Install the eSIM. Land connected.

 

When everything works, it remains one of travel technology’s better propositions. No plastic SIM card, no airport kiosk and, ideally, no unpleasant roaming bill waiting at home.

Behind that clean customer journey, however, sits a considerably less elegant question: how much of the underlying technology should an eSIM company actually control?

Should it own provisioning, connectivity, billing, policy management, analytics and customer support? Or should it remain deliberately light, combining specialist suppliers into a flexible, market-ready service?

This is often framed as a choice between a unified stack and a best-of-breed model. In practice, the industry is moving toward something more complicated. Telecom infrastructure is becoming more modular, while software companies are taking control of the customer-facing layers. The two sides are meeting somewhere in the middle.

The result is not simply an architecture debate. It is a battle over margin, product quality, speed and, ultimately, who holds the power when something goes wrong.

And the answer looks very different depending on whether the company is an MNO, an MVNO or a travel eSIM brand.

Integration without the monolith

The attraction of vertical integration is not difficult to understand.

Fewer suppliers can mean fewer points of failure. Owning more of the stack may provide better visibility into network performance, clearer accountability, greater pricing flexibility and less finger-pointing between vendors when a customer cannot connect.

For mobile network operators, much of this infrastructure already exists. They have network assets, roaming agreements, charging platforms, regulatory obligations, customer data and SIM lifecycle systems.

What many operators do not have is an architecture designed for instant travel products, app-based activation, real-time partner management or rapid experimentation.

Legacy OSS and BSS platforms were built for a different telecom era. They are reliable, but reliability does not automatically translate into speed. Launching a new destination bundle, testing a pricing model or onboarding a non-telecom partner can still involve multiple systems, departments and approval processes.

This is why initiatives such as TM Forum’s Open Digital Architecture matter. ODA is not encouraging operators to replace one giant system with another. Its broader aim is to make telecom technology more modular, cloud-based and interoperable, allowing components to communicate through standardized interfaces.

Even the unified-stack argument, then, is becoming less about owning one closed technological fortress and more about controlling how the components work together.

Charging illustrates the point. Nokia promotes converged charging as a common platform for consumer, enterprise, wholesale, prepaid, postpaid and hybrid services. That matters because an eSIM proposition is never only about downloading a profile. It also involves pricing, usage control, policy enforcement, partner revenue, settlement and customer visibility.

The operator that connects those pieces effectively can make network access behave like a modern digital product.

That is more valuable than integration for integration’s sake.

Why modular models moved faster

Best-of-breed software emerged because telecom transformation programs are expensive, complicated and rarely famous for finishing early.

A travel eSIM company cannot wait several years for every system beneath it to become perfect. It needs to launch plans, add destinations, support customers, run campaigns and adjust pricing now.

The modular answer is to select a specialist for each job: provisioning from one supplier, connectivity from another, fraud detection from a third, and separate platforms for CRM, payments, analytics or support.

This model offers speed and specialization. It also reflects the wider direction of enterprise technology, where multicloud strategies, industry-specific platforms, AI services and digital sovereignty requirements are producing increasingly mixed technology environments.

READ MORE: Cerqle Showcases eSIM Orchestration Platform at MWC

Gartner has repeatedly identified these themes as major forces in cloud adoption. At the same time, enterprises are becoming more sceptical when cloud spending produces higher costs without enough control or measurable value.

Telecom companies recognize that feeling. Everyone wants flexibility until flexibility becomes a collection of expensive dependencies that nobody fully understands.

Still, the modular model has been extraordinarily effective for travel eSIM brands.

These companies did not need to recreate an operator. They needed dependable wholesale connectivity, a competent app or website, a straightforward checkout process, responsive support, and a brand travellers were willing to trust.

That is the thin-stack model: own the customer relationship and outsource much of the telecom machinery.

1GLOBAL has described the distinction between full-stack providers and typical travel eSIM resellers in similar terms. A thin-stack company may be excellent at customer acquisition and digital experience, while upstream suppliers retain control of network agreements, provisioning and core connectivity.

That is not automatically a weakness. It is a deliberate commercial model.

Airalo Partners offers APIs, integrations, co-branding, affiliate tools and partner platforms that allow other companies to distribute eSIM products without building telecom infrastructure. eSIM Go follows a comparable approach, giving businesses access to ready-made data bundles and tools for managing their eSIM propositions.

This is what modular orchestration looks like in practice: businesses can assemble and launch a connectivity product far faster than they could build one.

The trade-off is that speed at launch may create dependency at scale.

MNOs have assets but lack urgency

For operators, the stack question is partly defensive.

MNOs possess advantages that most travel eSIM companies cannot easily reproduce: spectrum, regulated infrastructure, wholesale relationships, network data and established enterprise credibility.

What they often lack is the commercial urgency of a digital challenger.

A travel eSIM brand can test a new landing page, destination plan, airline promotion or discount code within days. An operator may need commercial approval, technical integration, legal review and coordination across several internal teams.

This helps explain why operator-led travel eSIM products can feel dependable but not always particularly distinctive.

The strongest MNO response is not to imitate Airalo, Holafly or every other consumer travel app. Operators should expose the capabilities those companies cannot easily build themselves.

That means programmable charging, transparent network information, partner-friendly APIs, easier onboarding, compliance support, real-time reporting and more flexible wholesale products.

Nokia’s 2026 5G Core SaaS deployment with Citymesh offers a useful signal. Network functions that once required major infrastructure projects are becoming more consumable, while programmable APIs are being presented as a route to faster monetization.

Even core telecom technology is beginning to arrive in a more software-like form.

For operators, the opportunity is not merely to sell more data. It is to turn network capabilities into products that other businesses can use without first becoming telecom specialists.

MVNOs must choose what matters

MVNOs face the least comfortable version of this decision.

They typically do not own the radio network, but they may control significantly more of the customer journey, billing logic and commercial proposition than a simple reseller. They sit between operators with heavy infrastructure and travel eSIM brands with lighter, more flexible models.

Trying to replicate an MNO can become expensive very quickly. Remaining little more than a branded storefront, however, leaves the business exposed to price competition and supplier decisions.

The sensible position is selective integration.

An MVNO should own the layers that directly influence customer experience, differentiation and margin. These may include charging flexibility, SIM lifecycle management, usage visibility, fraud controls, customer segmentation and partner management.

Infrastructure where scale matters more than differentiation can remain with specialist partners.

The goal is not maximum ownership. It is meaningful control.

An MVNO does not need to operate every technical component. It does need enough visibility and authority to understand what is happening, intervene when necessary and avoid becoming commercially interchangeable.

Travel eSIM won the packaging battle

Pure-play travel eSIM brands have taught the wider telecom industry a useful lesson: customers do not buy provisioning architecture.

They buy confidence.

Will the service work when they land? Is installation straightforward? Is the price understandable? Can they use hotspot data? Will someone help when the phone shows signal but no internet connection?

Travel eSIM companies succeeded by turning several complicated telecom processes into something that looked and felt like an ordinary digital purchase.

That packaging advantage should not be underestimated. Airalo’s $220 million investment round in 2025, described by CVC as supporting the company’s expansion in borderless mobile connectivity, demonstrated substantial investor confidence in software-led eSIM distribution.

But the easiest stage of travel eSIM growth is ending.

Most established providers can now offer polished applications, large destination lists and similar data bundles. Product pages frequently look alike. Coverage claims sound alike. Even pricing is becoming harder to distinguish once promotions and regional plans are taken into account.

The next competitive layer sits deeper.

Supplier redundancy, network selection, latency, routing quality, fraud prevention, entitlement management, support diagnostics and enterprise reporting will matter more. So will the ability to explain what “unlimited” actually means and identify which network a customer will use before a problem occurs.

A thin stack is efficient while upstream partners perform reliably. Its limitations become visible when the brand needs more data, more flexibility or more influence over the customer outcome than its suppliers are prepared to provide.

That does not mean every travel eSIM company should start building telecom infrastructure. For most, that would be financially reckless.

It does mean that the strongest brands will gradually take control of selected layers beneath the interface.

Embedded eSIM changes the calculation

The stack question becomes more important as eSIM moves beyond dedicated travel apps.

Banks, airlines, hotel groups, mobility platforms and financial apps are beginning to treat connectivity as a feature that can sit inside an existing customer journey.

These companies do not want to become MVNOs. A bank does not want to negotiate roaming agreements. An airline does not want to manage profile provisioning. A hotel group does not want its reception staff investigating an SM-DP+ error.

They want connectivity to work inside their product while someone else absorbs the complexity.

GSMA’s eSIM and remote provisioning specifications provide the standardization needed for that ecosystem to expand. Newer work around SGP.32 also supports IoT deployments where devices may have limited interfaces and remain in service for long periods.

This broadens the market far beyond travellers purchasing a few gigabytes for Italy.

eSIM is moving into connected vehicles, wearables, enterprise hardware, payment apps, airline journeys, accommodation platforms and IoT fleets. In these environments, the most important provider may not be the company with the best-known consumer brand.

It may be the company that can coordinate networks, provisioning, charging, compliance and support so effectively that the partner barely notices the telecom layer exists.

The real competitive advantage

The meaningful question is no longer whether a company operates a unified or modular stack.

It is whether the company understands which parts of the stack determine the customer outcome—and whether it controls enough of them to keep its promises.

A provider can own large amounts of infrastructure and still deliver a poor product. It can also rely heavily on partners and create an excellent service. Architecture alone does not guarantee quality.

READ MORE: Booking Platforms Monetized Flights and Stays. Connectivity Is Next

What matters is visibility, leverage, and the ability to act.

Can the provider identify why a customer is offline? Can it change routes or networks? Can it create a new commercial model without waiting months? Can it give a partner useful data? Can it replace a supplier without rebuilding the entire proposition?

Those questions reveal far more than a “full-stack” label ever will.

Conclusion

The eSIM market will not divide neatly into vertically integrated winners and modular losers.

It will divide between companies that understand their dependencies and companies that discover them only when something breaks.

MNOs will integrate more aggressively because they need to monetize their infrastructure faster and remove internal friction. MVNOs will need to decide which capabilities genuinely create leverage rather than collecting technology for appearance’s sake. Travel eSIM brands will continue to lead in packaging and customer acquisition, but they will need stronger control beneath the interface as the category becomes more competitive.

The next source of differentiation will not be another destination page or a slightly larger coverage number. It will be the intelligence underneath the offer: how traffic is routed, how problems are diagnosed, how suppliers are managed, how policies are applied and how easily connectivity can be embedded into another company’s service.

The smartest players will not necessarily own everything.

They will know exactly what they cannot afford not to control.

And from the customer’s perspective, the stack will disappear—which is probably the best evidence that it is working.