If you’ve worked anywhere near travel eSIMs (or you’ve just bought a few), you’ve noticed the pattern: “Unlimited” is the headline, and the fine print is… a novel.

But I don’t want to do the usual rant about throttling.

The more interesting question is: why does “unlimited” exist at all in a product where every gigabyte has a real cost somewhere in the chain?

Because “unlimited” is rarely about giving you infinite data. It’s about controlling risk. For the provider, it’s margin insurance.

And the mechanics of that insurance are mostly invisible unless you look at how mobile networks actually keep themselves upright: traffic shaping, fair usage thresholds, and wholesale cost exposure.

So let’s unpack the hidden economics.

The network does not sell “data.” It sells priority.

Here’s the first mental switch that makes the whole “unlimited” debate make sense.

On a mobile network, “data” is not a single uniform commodity. What users experience is a mix of:

• radio conditions (are you near a tower?)
• congestion (how many people are trying to use the same cell right now?)
• policy (what the network is allowed to give you, and when)

That last part is where “unlimited” quietly becomes a priority product.

In practice, many networks manage heavy usage by changing how your traffic is treated during congestion. One very plain-language example: EE’s small business traffic management documentation describes changing a user’s QCI (QoS Class Identifier) after a fair usage policy breach, which results in deprioritisation on congested cells.

That is the economic trick: you can keep the word “unlimited” while controlling when expensive capacity is allocated to the heaviest users.

Traffic shaping: the invisible steering wheel

“Traffic shaping” sounds sinister, but it’s basically how networks avoid melting.

Operators use traffic management techniques to maintain performance and handle congestion, especially as more traffic is encrypted and harder to classify at the application level. GSMA has published operator-focused material on smarter traffic management for modern protocols and encrypted data.

Now, in Europe, there’s an important constraint: net neutrality rules mean operators cannot just throttle whatever they feel like for commercial reasons. BEREC’s guidelines on the Open Internet Regulation are the reference point regulators use to judge what counts as reasonable traffic management versus prohibited discrimination.

So what does shaping look like in the real world (without getting too nerdy)?

Common “unlimited” shaping patterns

• Deprioritisation during congestion: you still have service, but you get served after higher-priority users.
• Policy-based speed tiers after a threshold: full-speed up to X, then a lower speed.
• Tethering rules: hotspot traffic often has separate allowances or stricter policies.

Notice what’s missing: a promise of infinite high-speed capacity at all times. That would be economically irrational.

Fair usage thresholds: the real product is “how much high-speed is included”

Fair usage is not just a legal disclaimer. It’s the core of the pricing model.

The simplest way to think about it is:

Unlimited plans are typically “unlimited access,” not “unlimited high-speed.”

And thresholds are how providers can publish a single attractive price while protecting themselves from the small percentage of users who would otherwise turn that plan into a loss-making product.

If you want a concrete, non-theoretical example again: EE explicitly references a fair usage threshold for unlimited plans (and actions taken when it’s breached).

So yes, “unlimited” is real in the sense that your service continues. But economically, the plan is calibrated around a statistical expectation of usage, with guardrails.

Wholesale cost exposure: “unlimited” is terrifying when you don’t own the network

Now we get to the part most consumers never see.

A large share of travel eSIM brands are not mobile network operators. They are closer to MVNOs, aggregators, or digital distributors.

That means they often buy data wholesale (directly or indirectly) and resell it.

If you are a reseller, “unlimited” is a risk product because your cost can scale with consumption, while your revenue is fixed.

To make that survivable, you need one (or more) of these:

The reseller survival kit

• Wholesale contracts that cap exposure (rare, expensive, or limited to certain footprints)
• Policy controls (thresholds, speed steps, deprioritisation)
• Network selection logic (steer users toward cheaper routes when possible)
• Product design (day passes, “unlimited” with stated high-speed quotas, region-based caps)

And yes, wholesale costs can still be meaningful even in regulated environments. The European Commission has documented wholesale roaming price trends in the EEA, including average wholesale roaming prices for data (EUR/GB) and the trajectory of regulated caps.

Even if those numbers look “low” on paper, the key point for a travel eSIM business is this:

If your cost is per-GB and your retail price is flat, unlimited is only safe when you control usage or your upstream agreement does.

Why “unlimited” is often margin insurance

So what is “margin insurance” in this context?

It’s a product designed to protect the provider from the two worst scenarios:

1. Unbounded usage from heavy users
2. Unpredictable cost spikes from wholesale routes (especially across multiple countries and networks)

Unlimited plans reduce customer price sensitivity (good for conversion), reduce comparison-shopping (good for retention), and create a predictable revenue line.

But to make the economics work, they need hidden stabilisers:

• thresholds that convert “infinite” into “statistically bounded”
• traffic management that reduces peak cost
• policy controls that stop one user from eating the business model

That’s not a moral judgement. It’s just the math.

How does this compare across the market right now

This is where the market is getting interesting, because “unlimited” is splitting into different philosophies.

Model A: Marketing unlimited

You see this where the headline is unlimited, and the experience is governed by vague FUP language. It sells well. It also creates the most confusion, because the customer can’t easily predict what “unlimited” means for their use case.

Model B: Structured unlimited

This is where providers increasingly spell out high-speed allowances and what happens after. It’s still “unlimited,” but the product is honestly a two-stage service: fast, then reduced speed or deprioritised.

This model tends to win trust with frequent travellers because it’s more predictable, and predictability is the real premium in travel connectivity.

Model C: “Unlimited” alternatives that avoid the word

Day-pass products, monthly quotas with rollover logic, or multi-tier regional bundles often outperform “unlimited” for a big chunk of users because they match real travel patterns better. They also reduce the provider’s tail-risk.

Under the hood, all three models are responding to the same reality: traffic and wholesale costs are not infinite, so the product can’t be either.

Conclusion fair usage policy unlimited eSIM

The travel eSIM market has been debating whether unlimited is “real.” The more useful question is: what kind of unlimited are you buying, and what risk is the provider trying to hedge?

Net neutrality rules in Europe shape what operators can do, but they do not remove the underlying scarcity problem. Networks still manage congestion, and resellers still face wholesale exposure.

So expect the next wave of “unlimited” to become less of a magic word and more of a declared structure: clear high-speed thresholds, explicit post-threshold behaviour, and product designs that admit the truth politely.

Because in 2026, the winning “unlimited” offer is not the one that promises infinity.

It’s the one that makes your connectivity predictable, while still being economically survivable for the provider.