News / Voice
Bundled VoIP gains in 2005 traffic study
On 2 February 2006, a Sandvine traffic study reported that provider-branded internet telephony packages had become more prominent than third-party VoIP in Europe. It measured traffic across a sample of broadband households rather than counting every voice customer.
The reported shares
For the final quarter of 2005, the study put branded European VoIP at 51.2 percent of calls and Skype at 45 percent of minutes. It described similar momentum for bundled services in North America. These figures are the study’s historical measurements, not a current market ranking.
What bundling changed
The report’s explanation was straightforward: newcomers to broadband often took the phone service offered with their connection. That gave providers a ready distribution channel. It did not show that existing Skype users had stopped using the service, only that the mix of observed traffic had shifted.
Traffic share and user preference are different
Sandvine’s figures compare branded provider VoIP and Skype through observed calls and minutes, not through a census of all people who might use either service. A shift in traffic share can arise because new broadband customers activate a bundled voice option, even if established Skype users do not abandon it. The result describes the composition of traffic seen by the study, not a simple popularity league table.
Bundling changes the starting decision
When voice service is offered with a broadband connection, a new subscriber can begin with the provider’s option without separately selecting and installing another service. That distribution mechanism is what the report identifies. It reduces the number of choices required at sign-up, while an independent service must still be deliberately adopted. Convenience at acquisition may explain growth without revealing whether users later value the voice service more highly.
Why the measures need care
The study’s European comparison used 51.2 percent of calls for branded VoIP and 45 percent of minutes for Skype. Calls and minutes answer related but not identical questions: one can change without the other when call length differs. The figures are historical observations from late 2005, not a current ranking. A traffic report needs its unit, sample and distribution context before it becomes a verdict on a competitor. It is especially important here because the two percentages use different denominators, so neither number alone can describe the full behaviour of all broadband voice users.
For example, many short bundled calls could affect the call count differently from fewer long Skype conversations, without any individual changing their preferred application.