What a test is for

The question a buyer actually has is whether the QR becomes a working Viettel data line, not whether a marketing page can show a large number. Our check follows the same steps we tell you to follow. Install on ordinary Wi-Fi, from the email QR, without a special appliance. Label the line. Leave the home SIM in place. In Vietnam, select the new line for mobile data and set data roaming off. Wait for registration. Open a map. Open a ride-hailing session. Turn on hotspot and load a page on a second device. Trigger an SMS to the home number and see that dual-SIM settings did not disable it.

What we refuse to publish

A single speed test in one cafe is not a city score, and a made-up score is worse. When this site shows megabits, the sentence includes the source, the report name, and the date. Opensignal’s August 2026 Vietnam report, nPerf’s 2026 barometer, and Ookla’s country median are the three we use. Ookla’s median is not described as Viettel. If a place has no published local breakdown, the page says the figures are national and leaves the local test undone rather than filling a blank.

Where judgment calls remain

Registration can take several minutes in an arrivals hall. We treat “online within about ten minutes, roaming off” as a pass. We treat “no service because the phone is a mainland China model without an EID” as a compatibility miss, not a bad batch of QRs. We treat a quiet pass on the Ha Giang loop as coverage, which we describe, not as a defect in the profile. We treat a day that dies after heavy hotspot use as the 5GB rule working. The plan is supposed to stop. A test that ignores the cap will call a working product broken at 6 p.m.

We do not currently publish a dated log of handset models, dates, and districts. That log is the next honest artifact, and it is not invented here to look complete. Reviews are in the same state: the review block stays hidden until real notes exist. If you want to send a real observation, include the city, the phone, and what you were doing. We will not rewrite it into a star rating.

TODO: publish a real test log (date, phone model, city or road, pass/fail on install, Grab, hotspot, and home-SIM SMS). No speed figures until a saved measurement exists.