Back to the study

Methodology

How the roaming-day sample was selected, how official evidence was weighted and how 420 trigger observations and 84 scenario outcomes were coded.

Download methodology as Markdown

Research question

How do major mobile carriers define, trigger, price and explain a paid roaming “day,” and can identical short periods of qualifying use produce different numbers of daily fees?

Sample

The purposive sample contains three large mobile carriers from each of four English-language markets:

  • United States: AT&T, Verizon and T-Mobile.
  • United Kingdom: EE, Vodafone UK and Three UK.
  • Canada: Rogers, Bell and TELUS.
  • Australia: Telstra, Optus and Vodafone Australia.

For each carrier, the audit selected the closest mainstream consumer postpaid product sold as a daily roaming fee, daily pass or one-day pass. Where rates vary by zone, the database identifies the selected zone and records the other prominent rate. Where the product varies by plan generation, the current rate for a newly eligible plan is used and the limitation is disclosed.

T-Mobile remains in the sample even though its one-day pass is scheduled rather than triggered by first roaming use. That difference is itself relevant to the definition of a “day.”

Research date and source hierarchy

Sources were accessed on 21 November 2026. Evidence priority was:

  1. Legal terms, tariff/price guides and critical-information summaries.
  2. Carrier support pages.
  3. Carrier product pages.
  4. Regulators.
  5. Official carrier communities only if no stronger source resolved the question.

No third-party comparison site, affiliate page, search-result snippet or AI summary is treated as evidence. The source register stores a direct URL, access/effective dates when available, a short evidence excerpt and neutral paraphrase.

Coding

Day type

Each selected product was coded as one of:

  • rolling_24_hours: a period beginning with qualifying use and ending 24 hours later.
  • rolling_24_hours_or_data_exhaustion: a period that can end after 24 hours or earlier when the included data is exhausted.
  • calendar_day_home_timezone: a period ending at a fixed boundary in the carrier’s home time zone.
  • calendar_day_visited_timezone: a period ending at local midnight in the visited destination.
  • scheduled_24_hours: a purchased period beginning on a selected effective date/time rather than first use.
  • not_disclosed: no current product-specific day boundary located.

Trigger status

Each of 35 possible events was coded:

  • yes: explicitly starts the daily product.
  • no: explicitly does not start it.
  • conditional: can start it under stated conditions.
  • not disclosed: no product-specific official rule located.
  • not applicable: the concept does not apply, such as a scheduled pass not started by usage.

Silence was never converted to “no.” Generic “data use” establishes that an actual transfer can trigger a fee, but it was not automatically converted into a claim that every named background service is separately disclosed.

Disclosure scoring

Fifteen policy topics were scored 0 absent, 1 disclosed but incomplete/ambiguous, or 2 clear and prominent. The maximum is 30. The score measures disclosure only—not price, network quality, destination coverage or overall product value.

The provisional scorecard must not be presented as a final league table until it has been independently reviewed and carriers have had a right to correct factual errors.

Modeled scenarios

Seven scenarios were modeled for all 12 carriers, creating 84 rows. Models apply published rules; they are not observations from live devices or bills.

The late-night model uses a covered destination aligned with the relevant carrier reset where possible:

  • U.S. carriers: Colombia in November.
  • UK carriers: Portugal in November.
  • Canadian carriers: the U.S. Eastern Time zone; TELUS uses its stated traveller-local rule.
  • Australian carriers: Papua New Guinea.

The traveler makes qualifying use at 11:55 p.m. and again at 12:05 a.m. local time. The destination alignment is used to isolate the effect of the billing clock; it does not imply that all destinations behave identically.

The seven-day model intentionally includes eight use moments: noon on arrival, then noon on each of the following seven dates. It tests what happens at the exact end of a nominal seven-day span and is labelled accordingly.

Monetary estimates use each database row’s selected-zone base daily price. They exclude taxes unless included by the carrier, domestic-plan charges, data top-ups, long-distance charges and maritime pay-per-use charges.

Denominators

Counts use the full 12-carrier sample unless stated otherwise. When a question is not applicable or not disclosed, that status remains in the denominator and is reported separately.

Limitations

  • Carrier policies and prices change frequently.
  • Several products vary by domestic plan, plan vintage, customer tenure and destination.
  • The audit did not log in to customer accounts or test personalized eligibility.
  • No live-device, roaming-network or bill testing was performed.
  • Search engines can expose stale official pages; current dated price guides were preferred where conflicts existed.
  • Some policy details are split across legal terms, price guides and support pages.
  • The first pass was completed by one researcher; independent review remains outstanding.
  • Carrier replies remain outstanding.
  • The sample supports claims about these 12 products, not the global telecom market.

Independence

uPhone supplied the research brief but is not included in the dataset, scoring or comparative findings. The study does not compare the daily carrier products with travel eSIM prices. Any later commercial commentary should be visibly separated from the research.