A travel eSIM is one of the easiest digital products to launch and one of the hardest to sell at a profit. A reseller can connect to a wholesale API, list plans for well over a hundred destinations and take the first order within weeks. So can everyone else, and most of them are reselling the same inventory. What separates the stores that last from the ones that burn their budget on paid clicks is rarely the data plan. It is where they are allowed to sell, what they promise, who brings them customers and how those customers bring the next ones.
The demand is real. Kaleido Intelligence puts traveller adoption of travel eSIM packages at 12% in 2025 and forecasts 31% by 2030, and its travel eSIM research projects retail spend rising from $3.8 billion in 2025 to nearly $12 billion by 2030. A growing market with almost no barrier to entry is exactly the kind that punishes undifferentiated sellers.
A disclosure before the checklist: Switas builds and operates Spidot, a travel eSIM store, so these are questions we have had to answer for our own product and not only for clients. Nothing below is legal or tax advice. Treat the regulatory sections as a list of questions to take to counsel in each market you sell into.
1. Check where you are allowed to sell, not only where the data works
Every eSIM store publishes a coverage list. Far fewer keep the list that matters more for marketing: the countries whose residents they may lawfully sell to. These are two separate questions with two separate regulators. The destination country decides whether the plan can be used on its networks. The buyer's country decides whether you may advertise and sell a telecom service to people standing in it.
Two cases show how quickly the second question can close a market:
- Türkiye. In July 2025 the telecom regulator BTK blocked access to the websites of at least 35 eSIM providers, under two rulings dated 10 and 17 July, according to bne IntelliNews reporting on the EngelliWeb project. The list included the largest names in the category. eSIM technology was not banned. The storefronts became unreachable from inside the country, which for a marketer has the same effect.
- India. Selling foreign SIM products to customers in India requires a no-objection certificate from the Department of Telecommunications under a policy in force since January 2022. In January 2024 two of the best-known travel eSIM apps were removed from the Indian app stores for operating without one, as MediaNama reported. The certificate comes with conditions, including a locally registered company, identity documents collected from buyers and regular reporting to the authorities.
The practical output is a two-axis market matrix. One axis is the sell-to country, the other is the destination. Each cell records whether you are authorised, what the buyer must provide and which channels are open. Ad targeting, app store availability, checkout geo-rules and affiliate agreements should all read from that matrix. Running a campaign into a country where your site can be blocked next week is spend with no residual value.
2. Budget for identity checks and tax before you set a price
Some destinations require the buyer's identity to be verified before an eSIM is activated. Airalo, for example, tells customers that an eKYC check applies to plans for certain countries and recommends completing it several days before travel. The industry has not settled how to do this well: the Mobile Ecosystem Forum's wholesale working group dedicated a session in September 2026 to how travel eSIM sellers can meet KYC requirements without damaging conversion.
For a growth team this has three consequences. The checkout needs a per-destination identity step that appears only when required. The product page has to say so before payment, because a document request after payment produces refunds and bad reviews. And the last-minute airport buyer, usually the highest-intent customer you have, cannot be served for those destinations at all, which should be reflected in how you bid on them.
Tax belongs in the same conversation. In the EU, consumer telecom services have been taxed in the customer's member state since 2015, and member states may apply use-and-enjoyment rules that change the treatment of services consumed outside the Union. On an order worth a few dollars, whether VAT applies is not a footnote. It can be the whole margin. Get the treatment confirmed per market before you publish a price list, not after the first return is due.
3. Make every claim in the ad survive the small print
"Unlimited" is the most valuable word in eSIM advertising and the most dangerous. Most unlimited travel plans are a daily high-speed allowance followed by reduced speed. Regulators have a settled view on this pattern. In a 2025 ruling on a UK mobile operator, the Advertising Standards Authority restated the test: an unlimited claim is acceptable only if exceeding a fair-use threshold brings no extra charge or suspension and any speed or usage limits are moderate, and the qualification must be presented clearly alongside the claim. The operator's plan passed the first part. One of its ads still broke the rules because the qualification was not shown clearly enough.
The ruling is British, but the principle applies under consumer protection law in most markets, and it is also good conversion practice. A buyer who discovers throttling on day two of a trip asks for a refund and writes the review that your next thousand visitors read. On Spidot we state in the plan description itself that speed may slow after heavy use, instead of leaving it to the terms page. The same discipline applies to "5G", to "works in 190+ countries" and to any speed figure: if you cannot substantiate it per destination, do not put it in the headline.
4. Know your cost floor, because you will not win on price per GB
Price competition in this category is structural, and the numbers show why. The 32nd BEREC International Roaming Benchmark Report, published in March 2026 with data from 148 European roaming providers, records an EEA average wholesale data roaming rate of €0.57 per GB in Q3 2025 against a regulated cap of €1.30. The lowest rates operators reported paying were €0.03 to €0.04 per GB. The same report notes that MVNOs generally pay above the average, close to the cap. The cap itself steps down to €1.10 in 2026 and €1.00 from 2027 under the EU Roaming Regulation.
Those are intra-EEA operator-to-operator figures, not what a reseller pays a wholesale platform. They are useful for one reason: they show a spread of more than tenfold between the best and the typical cost of the same gigabyte. A small store buying through an aggregator sits at the expensive end of that spread. The largest brands, with volume commitments and direct operator deals, sit near the cheap end. A price war against them is a war against their cost base.
Retail prices are falling too. BEREC's series for data used outside the EEA dropped from €5.12 per GB in Q4 2024 to €4.42 in Q3 2025. And the sameness is visible in public price lists. One reseller's 2026 price index, built from its own catalogue of 2,668 plans in 184 countries, shows a global average of $5.19 per GB and a dozen major destinations sharing an identical best price to the cent. It is a single vendor's data and should be read as indicative, but identical prices across unrelated countries are what a shared wholesale catalogue looks like from the outside.
5. Sell the trip, not the gigabyte
If the gigabyte is a commodity, the unit you price and present should be something else. The buying behaviour supports this. Kaleido's latest traveller survey found average travel eSIM spend per trip up 133% to $28, driven by larger bundles and longer-duration plans. Travellers are not racing to the lowest per-GB price. They are paying more to stop thinking about data.
Ways to move the comparison away from price per GB:
- Lead with trip length and usage type. "Seven days in Japan, maps and messaging" is a decision a first-time buyer can make. "3 GB" is not, because most people do not know how much data they use in a week.
- Offer plan shapes competitors' lists do not have. Daily unlimited, data that does not expire between trips, data with a local number for calls, and regional bundles each answer a different anxiety. A regional plan also raises order value without a discount.
- Remove the fear of buying the wrong thing. The two most common pre-purchase doubts are whether the phone supports eSIM and whether it is carrier-locked. A compatibility check before the payment step converts better than a refund policy after it.
- Design for the second purchase. This is a need-based product bought a few times a year. An account that keeps past eSIMs, one-tap top-ups and a reminder timed to the next booking season do more for lifetime value than a loyalty tier.
- Test prices by destination against revenue per visitor. Conversion rate alone rewards the cheapest plan. Revenue per visitor, net of payment fees and refunds, shows whether a higher-priced default plan is the better storefront.
6. Build owned demand by destination and by corridor
Search demand for this product is shaped like "eSIM plus destination". That makes a page per country the natural architecture, and it is also where most stores damage themselves. A template with the country name swapped, multiplied by 180 countries and then by 20 machine-translated languages, is the pattern Google's scaled content abuse policy describes. Each destination page needs something a template cannot produce: the actual networks used, real plan prices, local caveats such as identity checks, and installation notes for that market.
Think in corridors, not countries. A corridor is an origin and a destination together: German residents travelling to Türkiye, British residents to the United States. A corridor tells you the language, the currency, the payment method and the season. It also tells you which origin markets you may sell into under check 1. Three corridors served properly will outperform a hundred country pages served identically.
Two more notes on owned channels:
- Seasonality is strong. The price index cited above, using Google Ads keyword data, places worldwide search interest in eSIMs at its peak in September, elevated from August to October and lowest in February. Content and creative should be live before the curve, not on it.
- AI assistants are a comparison channel. "Best eSIM for two weeks in Italy" is the sort of question people now ask an assistant. Stores with clear, dated, crawlable price and coverage data are the ones that get cited. This is the same discipline as the destination page: publish specifics a model can quote.
7. Borrow distribution you cannot afford to build
The category leaders are no longer waiting for travellers to search. Connectivity is being placed inside the booking itself. In June 2026 Orange Travel and Trip.com announced a global distribution partnership that puts Orange eSIM offers for five European destinations in the Trip.com purchase flow, paid in the traveller's local currency. Fintechs such as Revolut sell eSIM data inside the banking app their customers already open while abroad.
A new store will not sign a global travel agency. It can apply the same logic one tier down, where nobody is competing yet:
- B2B and group buyers. Tour operators, corporate travel desks, conference organisers, language schools, sports clubs and relocation firms all send groups abroad and none of them want to explain roaming. Bulk purchase, delivery by email and a single invoice is a product, not a favour.
- White-label and API resale. Regional travel agencies and booking tools can sell your inventory under their own name for a revenue share. You give up margin and gain customers at zero acquisition cost.
- Affiliates and travel creators. They work, with two cautions. Coupon sites will claim credit for orders that were already yours unless attribution rules exclude last-second code lookups. And incentivised recommendations must be disclosed as such, which should be written into the partner agreement.
8. Design referral and member-get-member around how people travel
Referral suits this product for a simple reason: people rarely travel alone, and the moment one person in a group gets online at arrivals, the others ask how. The category benchmark is modest. Airalo's programme gives the referrer $3 in account credit and the friend $3 off a first purchase, with no cap on the number of referrals and the reward paid only when the friend buys. That structure contains most of the right decisions. The design choices worth making deliberately are these:
| Decision | Sensible default | Why |
|---|---|---|
| Reward type | Account credit, not cash | Credit costs nothing until it is redeemed, and redeeming it is the second purchase, which is the hardest behaviour to produce in a need-based product. |
| Reward size | Set against contribution margin per order, not revenue | Order values are small. A reward that looks trivial can exceed what the order earns after wholesale cost, payment fees and support. |
| Who is rewarded | Both sides | The friend needs a reason to use the code now. The referrer needs a reason to send it. |
| When you ask | After the first successful connection at the destination | The product has just worked and travel companions are within arm's reach. Asking at checkout asks before any value has been delivered. |
| Qualifying event | Friend's plan activated and past the refund window | Paying on purchase alone invites self-referral and refund abuse. |
| Sharing mechanism | The referrer shares their own link or code through their own channels | See the compliance note below. |
Beyond the standard two-sided scheme, travel offers member-get-member mechanics that other categories lack. A group checkout that lets one person buy for the whole party and send each eSIM by email turns the buyer into a distributor without calling it referral. A gift flow does the same for families. A separate, higher tier for people who refer at volume, such as tour guides and expat community organisers, belongs in the partner programme with its own agreement.
The compliance note. How the invitation is sent matters legally. The UK Information Commissioner's Office distinguishes two kinds of refer-a-friend scheme, as summarised here. Where the business collects the friend's contact details and sends the marketing message itself, it is hard to meet consent rules, because the friend never agreed to hear from you. Where the customer passes on a link or code themselves, the business is much less likely to be in breach. Build the second kind. Do not ask customers to type in their friends' email addresses.
Fraud and measurement. Cap rewards per device and per payment instrument, make credit non-transferable and watch for clusters of new accounts redeeming one code. Then measure whether the programme adds customers or relabels them: referral share of new customers, time to second order among referrers and, where volume allows, a holdout group that does not see the prompt.
9. Treat payments, support and activation as part of marketing
Three operational areas decide whether the acquisition spend above pays back.
- Payments and fraud. An instantly delivered prepaid digital good is a target for stolen cards. Blanket friction is the wrong answer, because the genuine customer is often standing in an airport. In a published case study, one eSIM provider reports wallets such as Apple Pay and Google Pay accounting for more than 70% of its transactions, and a 0.02% dispute rate after moving to custom risk rules that apply 3D Secure only to high-risk payments. Wallet-first checkout and risk-based authentication are conversion work.
- Support. Most contacts are installation problems, and they arrive at local arrival time, not office hours. Each one costs a meaningful share of the order's margin. Device-specific install guides, a status page per eSIM and a way to install before departure reduce contacts more cheaply than staffing does.
- Activation as the real conversion. A paid order is not a success until the eSIM is installed and has carried data. Track purchase, installation and first data session as three steps. The drop between them tells you where reviews, refunds and referral potential are being lost, and it is invisible in a standard e-commerce funnel.
The nine checks on one page
- A sell-to by destination matrix exists and drives ad targeting, app availability and checkout rules.
- Identity requirements are shown before payment, and tax treatment is confirmed per market.
- Every headline claim, "unlimited" above all, carries its qualification where the claim is made.
- You know your wholesale cost floor and have stopped benchmarking against sellers with a lower one.
- Plans are presented by trip, not by gigabyte, and tested on revenue per visitor.
- Destination pages contain market-specific facts and are organised around a small number of corridors.
- At least one distribution channel does not depend on paid clicks: B2B, white-label or affiliates.
- Referral rewards are credit-based, margin-aware, triggered after first use and shared by the customer.
- Activation rate, not order count, is the number the team reports.
None of these requires a larger media budget. Most of them reduce it. If you are launching or scaling a connectivity or travel product and want a second pair of eyes on the growth plan, the Switas Growth Management team does this work for travel and e-commerce brands. You can reach us here.
Sources
- Kaleido Intelligence, Roaming Data Hub press release and Travel eSIM research bundle
- BEREC, 32nd International Roaming Benchmark Data and Monitoring Report, BoR (26) 28, 12 March 2026
- bne IntelliNews, Turkey blocks e-SIM providers, 21 July 2025
- MediaNama, eSIM apps removed from app stores in India, January 2024
- Airalo, What is an eKYC and Refer and earn
- Mobile Ecosystem Forum, Travel eSIM and KYC working group session, 8 September 2026
- Advertising Standards Authority, ruling on EE Ltd
- Flysimio, Travel eSIM Price Index 2026
- Orange Travel and Trip.com partnership announcement, June 2026
- Foot Anstey, summary of ICO guidance on refer-a-friend schemes
- Stripe, ByteSIM customer story
