Scoring Methodology

Top eSIM Review scores providers on a 0-10 scale across five weighted categories. Scores come from hands-on testing, not surveys or manufacturer claims. Each quarterly cycle covers 24 destinations across six continents.

Testing uses three devices simultaneously: iPhone 15 Pro, Pixel 8, and Samsung Galaxy S24. Running all three at once exposes OS-level differences in eSIM handling that single-device tests miss. An activation that takes 90 seconds on iOS may take four minutes on Android, and that gap matters for scoring.

Five speed tests run per day across seven consecutive days per destination. Support is evaluated through 75 mystery-shopper tickets per cycle, 15 per provider, filed from cities each provider has not seen before. Scores recalculate quarterly, with out-of-cycle updates when providers materially change a product.

Supplementary field testing on a Motorola Edge 50 Pro captures activation timing benchmarks on mid-range Android hardware. Budget-tier Android devices handle eSIM provisioning differently from flagships — the Edge 50 Pro data is not included in primary scoring but informs our Ease of Use notes where activation gaps appear.

Every provider is tested on three devices in parallel during each destination visit. Using a single device would let OS-specific bugs inflate or deflate a provider’s score unfairly.

Each destination runs a 7-day testing window. Speed tests fire five times per day on each device: 07:00, 11:00, 14:00, 18:00, and 22:00 local time. That schedule captures morning commute load, midday baseline, afternoon peak, evening peak, and late-night floor.

Tests run in three cities per country. City selection follows a fixed template: the primary transport hub (usually the capital or the busiest airport city), one secondary city with different infrastructure vintage, and one smaller destination where network density is lower. This spread prevents urban bias from inflating scores.

Each speed test records download speed, upload speed, and latency. We use the 7-day median, not the peak session, to report performance. Providers that deliver consistent mid-tier speeds score better than providers with one strong session and four weak ones.

Support testing runs concurrently. The team files 15 mystery-shopper tickets per provider from new accounts, using new device fingerprints, from cities the provider has not seen before. Ticket scenarios are fixed each cycle to allow quarter-over-quarter comparison.

Each quarterly cycle covers 24 destinations across six continents. Destinations are selected to represent distinct infrastructure environments: high-density urban markets (Tokyo, London, New York), developing-network markets (Lagos, Nairobi, Bogotá), and island or geographically isolated markets (Bali, Tenerife, Maldives). This mix ensures that a provider’s score reflects real-world traveler conditions rather than only the routes where every provider performs well.

A global score does not tell travelers what they need to know. A provider that excels in Western Europe can fall apart in Southeast Asia. The country adjustment algorithm modifies each provider’s base score using four destination-specific inputs.

The four inputs are: active carrier count in that country, 5G network presence, VPN policy (whether the provider blocks VPN traffic on local carrier agreements), and pricing floor relative to the global median for that market tier. Each input adjusts the base score upward or downward by up to 0.5 points before the final country page score publishes.

Support scores come from 75 mystery-shopper tickets filed each quarterly cycle, 15 per provider. Tickets originate from accounts and device fingerprints with no connection to our domains. Providers cannot identify test contacts from legitimate customers.

Each ticket falls into one of five scenarios: activation help (device can’t connect after QR install), billing dispute (charge on statement doesn’t match plan price), coverage inquiry (asking which carrier is used in a specific city), refund request (citing a data delivery failure), and device compatibility question (asking whether a plan works on a specific model not listed in the FAQ). These five scenarios appear in the same proportion every cycle so scores are directly comparable quarter-over-quarter.

Two metrics drive the support score: median first-response time and first-contact resolution rate. First-contact resolution means the ticket was fully resolved in a single exchange without a follow-up message required from the test contact. A provider that responds in 90 seconds but requires three exchanges to resolve a billing dispute scores lower on support than a provider that responds in 12 minutes and resolves on first contact.

Weights are fixed for each annual cycle and disclosed publicly. The 2026 weights reflect our Q1 reader survey (n=1,840) on what matters most at purchase decision.

Coverage carries the highest weight at 30% because a provider that does not work in your destination is worth zero. Pricing, Speed, and App share 20% each because per-GB cost, network throughput, and activation experience are the three most measurable buyer signals after coverage. Support at 10% reflects that most travelers never contact support during a trip.

Methodology v3 (April 2026) consolidated the scoring to five categories, combining Features and Ease of Use into the App category and dropping Value as a standalone dimension. No provider moved more than 0.2 points on the methodology change alone.

Top eSIM Review accepts no provider payment for placement, no complimentary plans, and no sponsored travel. Every test purchase uses a personal credit card from a city the provider has not seen before.

Affiliate commissions exist on outbound provider links. These commissions are identical across all five providers and cannot influence rank order. If a provider offered a higher commission to improve its ranking, we would decline and disclose the approach publicly.

Reviewers are prohibited from accepting gifts, pre-publication previews, or direct provider communications about scoring. Providers see scores at the same moment readers do.

We do not accept sponsored content, paid placement, or “partner spotlights.” Every provider on this site earned its position through the same testing protocol. There is no mechanism to purchase visibility.

If you find a discrepancy between our scores and your experience, contact corrections@topesimreview.com. We log every correction publicly and re-test when a complaint is reproducible.

Scores update on a fixed quarterly cadence: January, April, July, and October. The April 2026 cycle is now complete. The July 2026 cycle begins testing in late June.