Orcon and Slingshot have topped New Zealand's broadband environment in the latest Epitiro performance survey.

It is the international broadband benchmarking company's first survey for 2009, and backs up Orcon's strong performance in a joint Epitiro/Herald survey last year.

The survey covered the "big five" ISPs: Orcon, Slingshot, Telecom, Vodafone and TCL.

Orcon and Slingshot shared top honours, followed by Telecom.

Epitiro boss Mike Cranna said the company has changed its ISP (internet service provider) testing methods, and the results are the combination of customer computer and lab-based testing.

The survey covered 12,000 homes, where users had download the company's iSposure software.

It found that big city users could expect significantly better download and upload performance, with rural users left out in the digital cold.

"There is a significant difference between ISP performance in those regions that contain one of New Zealand's five large cities, and those that don't," he said.

"To use line speed as an example: download speeds outside the big five city regions are a third slower than those within them. Upload speeds are half as fast.

"Orcon's performance is particularly strong in Auckland, where its unbundled services perform well, but it also tends to be a good performer in most other regions of New Zealand too," Cranna said, "Slingshot also performs well."

Orcon CEO Scott Bartlett was, predictably, happy with the results.

"We're very pleased to see that our commitment to providing the best broadband and the best service to New Zealanders continues to get recognised," he told nzherald.co.nz today.

But Cranna urges internet users to use Epitiro's software to get a full picture of their own broadband performance.

"Everybody should run tests from their own home and compare the results to other ISPs locally before making a purchase decision," he advises.

"Broadband service can vary from street to street. We encourage all users to download our freeware if you want your service experience to be included in our data."

As broadband infrastructure has improved, urban areas have been the big winners.

"2008 was a big year for broadband investment. Cabinetisation, content caching, ADSL2+ and unbundling have all had positive impacts on consumer experience of broadband services," he said, "but the impacts of these thus far have been felt more in the larger urban areas."

Test method

When run, the iSposure software measures various aspects of the ISP's performance from each individual machine, aggregating information to show which companies are providing the best overall service.

"This approach includes the effects of some factors which are beyond the ISPs' control, such as house wiring, phone line and modem quality," he said.

"If we have enough measurement points out there in the real world, these factors will average out for each ISP, and we'll get a sense of what their services are like to the user, despite these other factors.

"We do this by providing incentives for people to download our software onto their PC. This software then runs regular tests on their ISP's service."

"We measure performance for common things users do like downloading, surfing and gaming. We do this on a daily basis, which means we have millions of data points to construct a performance picture from."

The lab-based tests help assess 'perfect world' aspects of each ISP's performance.

"Broadband is tricky to measure, because the ISP doesn't always have control over all the factors that determine the quality of their service. This was one of the main drivers behind LLU (local loop unbudling).

"So when ranking ISPs, we use different methodologies to assess service performance on the one hand, and user experience of that performance on the other."