Data & methodologyPrice/GB V2, with explicit limits.
The rule, the aggregate calculations and the evidence retained in the frozen Q4 edition. Research methods are public; unknown fields stay unknown.
Price/GB methodology V2
Selection rule and actual volume
The methodology is designed to select the cheapest eligible publicly payable offer per provider × destination × tier. Price per GB = payable price ÷ actual delivered GB. The selection rule is the pipeline design; the Q4 frozen data cannot independently prove every selected offer was the cheapest available.
- 10 GB: actual_gb ≥ 10 and actual_gb < 20.
- 20 GB: actual_gb ≥ 20 and actual_gb < 50.
- 50 GB: actual_gb ≥ 50 and actual_gb ≤ 100. Secondary indicator.
- No invented duration, plan name, source URL or collection timestamp. Missing observation metadata remains UNKNOWN.
External competitor median
The external panel is Airalo, Nomad, Saily, Ubigi, Yesim, Jetpac, BNESIM, aloSIM, GigSky, Airhub. DriftEuropa and the network’s other own brands are excluded. For each destination and tier, take the median of the eligible external price-per-GB observations. At least seven competitors are required for the primary tiers and four for 50 GB.
Regional gap and market metrics
Destination gap = (DriftEuropa USD/GB ÷ competitor median USD/GB − 1) × 100. An eligible own-brand offer is required. The regional gap is the median of these unrounded destination gaps, with equal weight per destination. It is not the ratio of two global medians.
The market KPI is separately calculated as the median of the 12 destination competitor medians. It does not pool all provider offers and is not an own-brand pricing KPI. Published values are rounded for display; the downloadable numeric aggregates retain precision.
Currency and frozen baseline
The international benchmark is entirely in USD. The validated Q4 data uses 1 EUR = 1.1378 USD, recorded as the ECB reference conversion for 28 September 2026. This is a non-transactional conversion. The source workbook hash is a historical reference; the physical workbook was unavailable for direct hash reverification in this mission.
Q4 2026 is the first official V2 baseline. No V1 series is reconstructed into V2 history. Q1 2027 is the first possible comparable quarter, subject to a validated dataset.
Limitations and transparency
Q4 2026 is published from the frozen validated benchmark dataset used for this edition. The methodology is designed to select the lowest-priced eligible public offer for each provider, destination and tier. However, the frozen Q4 snapshot does not retain plan names, source URLs or duration metadata for every observation, so that offer-level selection cannot be independently re-audited from the published snapshot alone.
This limitation applies to the auditability of individual offer selection, not to the published aggregate calculations derived from the frozen Q4 dataset.
Unknown remains unknown. This edition does not demonstrate a universal validity of at least 30 days. Plan names, duration, source URLs and provider-specific collection timestamps cannot be supplied where the frozen data does not retain them.
The 12-destination panel does not represent every country, offer or service. Duration, networks, speed, support and conditions may differ. Prices can change after the snapshot date; this benchmark is not a live offer catalogue or a measure of network quality.
Q4 2026 is the first official V2 baseline. Trend status: INSUFFICIENT_HISTORY. No Q3-to-Q4 comparison or reconstructed V1/V2 trend is published. Q1 2027 is the first possible QoQ comparison, subject to a valid comparable dataset.
Other DriftEuropa research methods
How we document operators
Operator data is built from verifiable sources: live pricing from the eSIMaccess API, operator metadata (parent company, 5G launch, market position) cross-checked against official sources, and flags from open CC0 libraries. Each dataset carries its own “as of” date and is reviewed monthly against operator reports and field observation.
Field tests
Living across Southeast Asia since 2020, the founder logs real speed and coverage measurements on actual eSIMs, in actual locations. Those readings feed the mobile speed index and provide context for connectivity research. They are not part of the Price/GB competitive median.
Reviews
We do not publish invented reviews. Every customer review shown on the site is tied to a paid order and verified by cryptographic token.
Connectivity data and legacy indices
The separate connectivity dataset and historical Price Index and catalogue-history publications retain their own scopes and methods. They are not the Q4 V2 competitive benchmark. Their monthly or weekly series must not be used as V2 QoQ history.
Questions about a method: contact DriftEuropa.
Data and downloads
The public regional dataset exposes the frozen validated aggregates, provider coverage, method, provenance and Q4 limitations. All benchmark prices and price-per-GB values are in USD. CSV contains the 36 destination × tier aggregate rows; it is not an offer-level catalogue.
Frequently asked questions
How is price per GB calculated in V2?
The designed selection unit is provider × destination × tier: the cheapest eligible publicly payable offer, divided by actual delivered GB. External competitor medians exclude DriftEuropa. Regional gaps are medians of destination gaps with equal destination weight.
Can every Q4 selected offer be independently re-audited?
Q4 2026 is published from the frozen validated benchmark dataset used for this edition. The methodology is designed to select the lowest-priced eligible public offer for each provider, destination and tier. However, the frozen Q4 snapshot does not retain plan names, source URLs or duration metadata for every observation, so that offer-level selection cannot be independently re-audited from the published snapshot alone. This limitation applies to the auditability of individual offer selection, not to the published aggregate calculations derived from the frozen Q4 dataset.
Does Q4 establish a 30-day minimum validity?
No. The frozen Q4 data does not retain duration metadata for every observation. Unknown remains unknown; no universal 30-day validity claim is made.
Why is the 50 GB result secondary?
Comparable own-brand coverage is only 8 of 12 destinations at the four-competitor threshold. The −0.3% gap describes that small sub-panel and cannot be generalised to the region.
Does this publication show a quarterly price trend?
No. Q4 2026 is the first official V2 baseline and trend status is INSUFFICIENT_HISTORY. V1 monthly and catalogue histories remain separate. Q1 2027 is the first possible comparable quarter.
Who publishes the benchmark?
DriftEuropa publishes the research and is the own brand in the comparison. Own brands are excluded from competitor medians. The methodology and regional dataset are public.