What the NTC needs to really monitor Internet speeds in the Philippines

Various public consultations and meetings have been held in the past few months to specifically tackle the issue of internet speeds of Philippine Internet service providers, or ISPs.

Recently, the NTC has been able to procure a device that will do the measurements and we heard the device costs millions.

They’ve also made initial tests in the presence of several major ISPs so that a standard procedure is set in place and stakeholders are able to make the proper recommendations for consideration before  final measurements are rolled out in the coming months.

During the initial test, almost all ISPs passed the speed test in terms of download speeds. However, there are other factors that should have also been shared to the public, like latency and uplink speed.

The results are as expected since all concerned ISPs were informed ahead of time and probably made arrangements and configurations to optimize their connection for the test.

In future tests, however, NTC has stated that it will be doing the measurements in random locations and without the prior knowledge of concerned parties. This way, the results will be more reflective of the typical subscriber experience.

Multiple data sets will also be taken at different times of the day and multiple days of the month. All these data points will then be collated and averaged before they are published on the NTC website.

The results will also be compared against advertised speeds claimed by the respective ISPs. The general expectation is that measured speeds should be above 80 per cent of the advertised speed for a given subscriber plan.

If the result of an ISP does not make or is not within the expected level of service, the subscriber has the option to file a complaint with the Department of Trade and Industry for false advertising. The ISP can then be made aware and penalized for not keeping to its commitment. How effective this process is remains to be seen and we hope the DTI will act on it expeditiously.

The only problem we see at the moment is that the multi-million peso device that the NTC has procured to make the measurements is not fully compatible with the terminals of the ISPs. This means that for each of the random measurements, the NTC needs to contact the network and make the proper configuration to run the measurements.

If that is the case, then the ISP will have knowledge of the location and subscriber line that is being measured and may be able to make some adjustments to affect the results. We hope that the NTC will find a solution to this before they actually begin the random checks.

Likewise, since the measurement tools are very expensive, only a few units can be procured and used at a time. This could reduce the number of data points required to arrive at more accurate results.

We believe there is a more efficient and cost-effective way to make a realistic measurement of Internet speeds that can be harnessed through crowdsourcing.

An online tool similar to Ookla’s SpeedTest can be set up with standard paramenters and distributed to the public to access and submit test results. The test environment may need to be set up in such a way that it eliminates other extraneous factors and each result may need to be authenticated so that the data gathered more or less reflects the actual user experience.

This way, the methodology can be replicated, easily scalable and does not require expensive measuring devices. It will be a tool that both the ISP and the subscriber can agree on and use a basis to audit the advertised speeds of their respective ISPs.