Summary

  • KaliteNet puts broadband tests in users' hands. A test describes a connection under particular conditions, not the experience of every customer of that provider.
  • Its publications page now promises annual reports from January 2027. The earlier tender's tentative six-month measurement campaigns are a different timetable, not evidence of a missed report.

The most useful moment for comparing internet providers is before signing up. A speed test usually arrives afterwards: it tells you something about the connection you already have. Mauritius's KaliteNet is meant to bridge that gap, turning individual measurements into information that can help other people choose.

The service is no longer just a proposal. On October 19, its public portal offered maps with provider selectors and views for download speeds and network coverage. Selecting Emtel Mobile brought up coloured marks over Mauritius. That is something a reader can inspect. It is not, by itself, a finding that Emtel is faster, slower or available to every household beneath a coloured patch.

The next promise is easier to put on a calendar. The publications page says annual reports on service quality and user experience will begin in January 2027. No report download was listed there when checked. That observation does not establish that no report exists elsewhere, or that one is overdue.

The number needs some company

A download result without its circumstances is easy to overread. A package with a speed cap and a faster package are different purchases. A home Wi-Fi connection and a mobile connection are different test settings. Repeating a test on one handset adds observations, not new subscribers.

These were not objections invented after launch. ICTA's May 2023 consultation response records submissions from Mauritius Telecom Ltd. jointly with Cellplus Mobile Communications Ltd., Emtel Ltd., Ookla LLC and an unnamed Respondent A. They raised uneven participation across places and times, package limits, repeat tests and privacy, among other questions. Those were design concerns, not findings of defects in today's KaliteNet.

ICTA's answer kept the user in view: assess the end-to-end experience, identify package limitations and compare providers on a comparable basis. Other measurement methods could supplement the crowd. The point was not to discard an inconveniently slow result. It was to explain what that result could fairly be compared with.

A long preparation, and a specific next step

The January 2023 proposal envisaged a one-year pilot. The 2024 tender instead specified a three-month pilot and three-year operational phase, with two-month national campaigns every six months, timing still to be determined. That campaign design must not be confused with the current annual-publication announcement.

The tender covered fixed and mobile service, manual and automated tests, statistical methods, server choice and fair comparisons. It also required results from dedicated measurements to remain separate from crowdsourced results. These are instructions to bidders, not certificates that every feature was delivered.

The contract award went to nPerf SAS on September 11, 2024, for €347,700. The ministry records the launch on May 11, 2026. Four days later, Cabinet described tests of download and upload speeds, response time, browsing and video, with collected results intended for regional charts. ICTA now offers the free application on the web and through Apple, Google and Huawei stores. Free software does not necessarily mean free mobile data for running it.

For someone choosing a service, four things still need distinct labels: the package being sold, the performance measured, the places where service is available and what users experience. A map of observations can help with that choice without answering all four questions. A blank area is not proof of no service, any more than a busy patch speaks for everybody living there.

January's promised reports will be worth reading for how clearly they make those distinctions. The reward is practical: a better basis for choosing a connection, not simply another impressive-looking speed number.

Editorial perspective: Lu Heng on BTW's purpose, not independent evidence about Mauritius.