Chapter 21 - Broadband Breakthrough - The Battle to Unbundle

From NetHistory

We’re being out marketed, out smarted and outgunned in the marketplace…(we lack) the killer instinct; we are too tame, too lame, and too timid to call ourselves a challenger (we need to become) pre-meditated, cold blooded killers of our competition. TelstraClear CEO Alan Freeth’s Christmas message to senior staff, 21 December 2006

For true interactivity, powerful real-time gaming experiences, videoconferencing and professional-level VoIP or voice-based services you need high-speed, robust Internet access and QoS, something only Telecom could provide prior to the wholesaling of raw Internet bandwidth.

Successive governments, either intimidated by our biggest company or blinkered by the contradiction of ‘hands-off’ regulation, in effect protected Telecom’s right to maintain the bandwidth bottleneck. Because of this light-handed approach Telecom had continued to monopolise the local loop or ‘last mile’ of copper connections into businesses and homes, keeping competition at arm’s length.

Reinvestment was at a low ebb; in fact Telecom’s focus seemed to turn offshore with the 1999 acquisition of AAPT when its home turf was in desperate need of attention. It had dictated connection speeds and imposed data caps, preventing anyone other than itself from delivering Internet performance greater than 2Mbit/sec for downloading and 128kbit/sec for sending. It stated on many occasions it would only increase speeds when the market demanded, but even under a deafening roar of protest, continued to drip-feed services.

It had been a long-held belief in the industry that New Zealand’s $5.3 billion telecommunications market, up from $3.6 billion in 1997, should be experiencing more robust growth. In 2006 Telecom was making about $2.4 billion from local service and calling revenues; only a few entrepreneurial players, including service and equipment providers, got to share the rest of the pie. Even its competitors, TelstraClear being the largest, ended up as its customers.

Unbundling had been spoken about for several years but was first raised as a serious option in late 2003, kicking Telecom’s lobbying machine into overdrive. The full details of Telecom’s political activism, however, had remained private until December 2005, when The Press made an Official Information Act request, revealing the contents of a letter written by Telecom CEO Theresa Gattung to Minister of Communications Paul Swain in May 2004. She had warned him that 30 cents could be potentially wiped off Telecom’s share price if unbundling was required. The letter reminded Swain that Telecom was the country’s biggest company, comprising more than 20 percent of the sharemarket, and that the government superannuation fund was invested in it.

Gattung threatened Telecom would not invest in ‘next generation’ network upgrades for residential customers if TelstraClear got access to its network, urging Swain to accept the Telecommunication’s Commissioner’s watered-down proposal. The heavying worked – despite Swain and the MED supporting unbundling, this was overruled by Cabinet and the ‘market-led solution’ was chosen, essentially buying Telecom another year. TUANZ chief executive Ernie Newman said: “The suggestion ministers would favour one company in this way over the national interest almost suggests a banana republic.”1 Business columnist Gareth Morgan was stunned at the revelation:

The same month as Telecom’s closed-doors correspondence was disclosed it was revealed that Rosemary Howard, TelstraClear’s CEO, had also been lobbying. She had written to Jim Anderton, MED, in April 2004 pointing out that the United States had lodged a complaint against Mexico for not unbundling and failing to meet telecommunications obligations under the General Agreement on Trade in Services (GATS) and the World Trade Organisation (WTO), which had ruled in favour of the United States. “I consider that New Zealand is also not meeting its GATS obligations. Accordingly, New Zealand is risking being found guilty of a breach, which would cause significant damage to our reputation.”3

Howard also quoted a 2001 letter from the assistant for US trade representative for industry and trade to the New Zealand secretary of commerce, which said the United States believed local loop unbundling would be in line with New Zealand’s commitments under the WTO’s Basic Telecommunications Agreement. She said any free trade agreement between the United States and New Zealand would require unbundling. Australia’s Trade Minister, Mark Vaile, did his bit for TelstraClear by writing to Jim Sutton, Minister of Trade Negotiations, to say that a decision not to unbundle the local loop would be contrary to the goal of Closer Economic Relations (CER) such as free trade in services. New Zealand trade officials advised their ministers that not unbundling would not breach WTO, GATS, or CER commitments.4

RAT CRASHES NET The ongoing demand for better access to broadband wasn’t just rhetoric being spouted by Telecom’s competitors wanting a greater share of the pie; it was a cry from businesses who found the cost of moving increasingly rich content around the country or the world a major impediment to growth. Despite the huge success of the Lord of the Rings trilogy, New Zealand’s screen production industry was watching millions of dollars in business head offshore because it couldn’t get access to high-speed data networks. The NZ Screen Council had asked the government for access to its new ‘advanced network’ because commercial carriers including Telecom and TelstraClear couldn’t or wouldn’t deliver the required services, resulting in lucrative contracts going to countries better equipped for the digital future. Film makers in Los Angeles, for example, were keen to use digital effects houses in New Zealand but the network capability wasn’t available to enable the exchange of such data; 30 seconds of film could easily equate to 720Mb. Both Telecom and TelstraClear were charging by the megabyte and Telecom’s cost was $95 a Gigabyte, which was described by one potential customer as ‘outrageous.’5

Cost of access to bandwidth for everyday broadband right through to high-end use was a major issue if the country hoped to compete internationally. However the state of the Telecom network, and our reliance on it as essential infrastructure, was brought sharply into focus at the end of June 2005, when, as NZ Herald technology writer Peter Nowak explained, the ‘unthinkable’ happened:

Ironically Telecommunications Minister David Cunliffe praised Telecom for its quick action in getting things sorted and Telecom was quick to defend itself, saying it used world best practice in burying and protecting its cables. It said it couldn’t do anything about the rats, chastised Powerco for not checking where the cables were first, and promised that ongoing upgrades would provide five extra switching points to minimise the likelihood of such an outage ever happening again. Still, questions were asked about why Telecom had been under-investing in its network and why it had taken a crisis like this to bring such vulnerabilities to the surface?

WE PROMISE TO BE GOOD Telecom had inflated its broadband figures by including 128kbit/sec speeds when the global consensus was a minimum of 256kbit/sec, and in most cases 2Mbit/sec. It took until mid-2004 for it to revise the minimum broadband specification. Despite industry pressure to unbundle, newly appointed Telecommunications Commissioner Douglas Webb, under instructions from Cabinet, had decided to give Telecom another chance. Instead the Commerce Commission delivered an ultimatum that it must connect 250,000 new broadband customers by the end of 2005 or face regulation. A third of those customers (83,333) had to come from wholesaling to other ISPs.7 At the time of the undertaking only 15 percent of Telecom’s broadband customers were coming through wholesale.

It remained obstinately opposed to any suggestion of unbundling, even as the government-imposed deadline loomed. The Commerce Commission’s long-awaited unbundled bitstream service (UBS) determination, based on a complaint laid by TelstraClear, roundly rejected Telecom’s concerns about the dangers of delivering ‘unconstrained broadband’ to wholesale customers. In October 2005 the Commission said Telecom should provide UBS customers with DSL that had ‘a downstream peak information rate (PIR) at the maximum technical capacity possible.’ That rate was determined at around 7.6Mbit/sec based on the DSL equipment Telecom used. This was a departure from an earlier recommendation of 3.5Mbit/sec.

Telecom warned that providing such an unconstrained service would pose a risk to existing customers through increased signal noise in the cable sheaths, despite having delivered such a service to its own customers since 1999. The Commission was confident the new specifications were unlikely to create further risks. It also said there was no justification for Telecom charging different wholesale rates for business and residential UBS, as there was no material difference between what it was offering TelstraClear and what it used itself. It said Telecom must therefore offer TelstraClear a single, uniform wholesale rate for both business and residential UBS. While it had dropped its controversial ‘churn’ fee, charged when customers switched to another provider from $110 to $36.42, the Commission said a figure closer to $8 would be more reasonable. It expected Telecom to have the new regulated UBS ready before the end of 2005.

Telecom’s manager for government relations and industry affairs, Bruce Parkes, restated his concerns that unrestrained broadband would significantly impact on its ability to service customers a long distance from exchanges. He said up to 72,000 customers in rural and urban areas would be affected but obviously the Commission had decided those risk were outweighed by the need to increase competition.8 The promised delivery remained difficult. Telecom filed for a judicial review of the Telecommunications Commissioner’s draft wholesale ruling, concurrently announcing it would not be meeting the voluntary target of 250,000 wholesale broadband subscribers promised in 2004. It lamely argued it understood the goal to have been 50,000 wholesale connections, or a third of its total growth over the period. It had missed the target by 20,000.

Rather than prolong the debate TelstraClear, eager to have access to Telecom’s network at average retail prices less a suitable margin, caved in to a 3.5Mbit/sec compromise in December 2005. The new deal was immediately passed on to Telecom high-end account holders who were upgraded from 2Mbit/sec to 3.5Mbit/sec download speeds. The controversial upload speed remained at 128kbit/sec unless you were a premium account holder paying $80–$100 a month, in which case you would get 512kbit/sec upload speeds.9

In January 2006 Telecom and TelstraClear agreed on interconnection rates for phone access to each other’s landline networks. Telecom agreed to provide limited broadband services to TelstraClear and pay it a one-off $17.5 million to settle several issues, mainly backdating the agreements on wholesale discounts and interconnection. Both agreed to drop multiple legal actions that had been used to delay and obstruct the other. “The worst part of the deal is that Telecom’s odious and arbitrary crippling of the upstream speed of broadband connections remains in place,” opined Listener columnist Russell Brown in his review of the situation.10

The industry was incensed. Having done its backroom deal, Telecom presented remaining ISPs with a take-it-or-leave-it offer on wholesale broadband rates. Ihug and Slingshot applied to the Commerce Commission for equal access to broadband at higher speeds with Slingshot’s CEO Annette Presley describing the wholesale offering as ‘bullshit.’ “It costs more, the speeds are slower, and they can’t guarantee quality of service. How can we put that to our customers?”11

GAMEKEEPER AND POACHER Prime Minister Helen Clark had warned in a February 2006 parliamentary speech that New Zealand’s speed of broadband uptake was unsatisfactory and improving the situation would be one of the government’s top three priorities. She promised urgent legislative initiatives to deliver ‘faster Internet access and at more competitive prices.’12 No one could say Telecom wasn’t warned. Its endless tactics to keep prices high and speeds low, and its rapid response to defend its patch by placing obstacles in the way of competitors were legend.

As part of its 2006 election promises, the government placed increasing pressure on Telecom to improve its record in providing faster wholesale services to ISPs, aiming to shift the average cost of an account to a dollar a day or $30 a month rather than the market average of $40. The price point was considered an important psychological barrier for uptake.13 It was the government’s stated intention to get the country into the top quarter of the OECD Broadband Statistics listing; since 2003 it had languished at 22nd place out of 30 nations.

Our level of fast Internet access – 256kbit/sec plus – had lifted only marginally from 4.5 percent to 6.9 percent of users in the year to June 2005. To reach the government’s goal by 2010, about 80 percent of residences or 600,000 customers would need to be connected to broadband – the equivalent of a 300 percent growth spurt over the next five years. “These statistics pose a challenge to the industry and the government. Telcos, ISPs and the government need to lift their game if we’re ever going to gain the benefits that widespread access to high quality, affordable broadband will bring,” lamented InternetNZ president Colin Jackson.

The number of broadband subscribers throughout the OECD continued to increase in 2005 from 136 million in June to 158 million by December with growth holding steady at 15 percent, compared to New Zealand’s dire 8 percent. An April 2006 analysis of 87 providers in the 30 OECD countries found bundled video, voice and Internet access were available from 48 providers in 23 countries by September 2005. So-called ‘quad-play’ (voice, Internet, movies, and mobile voice) was available in 10 OECD countries. The challenge for developed nations had been to create a legal framework for LLU without getting in the way of innovation or disadvantaging competitors. In 2006 New Zealand remained one of the few OECD nations, alongside Mexico and Switzerland, yet to legislate for unbundling.

Telecom had continued to play gamekeeper and poacher, all but ignoring ministerial threats to play fair or face legislative changes. Then Telecom CEO Theresa Gattung called David Cunliffe’s bluff when he threatened regulation. Addressing business analysts in Sydney in March 2006, she said the government was far too smart to ‘do anything dumb’ like unbundling; suggesting the broadband issue was just a ‘manufactured grievance’ created by competitors.

In a last-ditch attempt to stave off regulation, Gattung had written to David Cunliffe in April offering to accelerate the company’s investment in broadband infrastructure by spending ‘hundreds of millions of dollars’ rolling out fibre-optic cable to almost all small towns by 2010. The offer was rejected and in a backlash Telecom said it would scale back its ambitions by extending its fibre network in the five main centres of Wellington, Auckland, Christchurch, Hamilton, and Dunedin only.15 When you fail to take your foot off a high-pressure hose, it’s inevitable there will be a leak. That leak, delivered by rogue parliamentary messenger Michael Ryan into the hands of a senior Telecom employee on 3 May 2006, proved the government was indeed serious. A budget announcement was planned to bring New Zealand in line with 26 other OECD nations by legislating for LLU along with other proposed changes to telecommunications regulation.

Once the news got out, it shaved at least a billion dollars off Telecom’s share index and while those shares quickly bounced back, Telecom was forced to rethink what it meant to operate in the unconstrained market promised to the country for nearly 20 years. By the time the audio tape of Gattung’s unfortunate outburst reached the media the government had already acted. Pressured from all sides to open up faster and more affordable Internet access, it administered the first of a series of enemas to Telecom in May 2006 in the hope of relieving New Zealand’s communications constipation. LLU would proceed and legislation was drafted with strong measures to ensure Telecom followed through.

Within weeks the Telecommunications Amendment Bill was sent to Parliament with Cunliffe warning that further measures might be on the way, possibly even a forced split of Telecom into separate retail and lines companies, Meantime Telecom would be required to open up its books and technology roadmap, and separate out its wholesale and retail divisions so the Commerce Commission could monitor its compliance. Telecom would be required by law to allow its rivals to install their equipment at its exchanges. The government also insisted the existing 128kbit/sec cap on upload speeds would have to go.

The bill extended the powers of the Commerce Commission to regulate the telecommunications sector with a new civil enforcement regime, with fines up to $1 million for a breach of the accounting separation requirements and $300,000 in any other cases. Cunliffe also asked the select committee to seek submissions on operational and physical separation options for Telecom.

The government would also renegotiate the Kiwi Share (TSO) as an incentive to improve broadband access in rural areas. It hinted there might also be action to prevent Telecom starting price wars to keep its customers. Strong intervention in the telecommunications market would, we were led to believe, literally kick start a new era of competition. However as several people pointed out, the devil was in the details and it was imperative for the industry and broadband customers that the technical and commercial framework of the legislation was rock solid and unambiguous.

OUTAGE EXPOSES ATTITUDE Telecom’s market value of $12 billion in 2005 dropped sharply to $8.8 billion in May 2006 following the unbundling announcement, but it remained in a good position for a rapid recovery. It announced $806 million tax paid profit for the 2005–2006 year. It had invested $585 million back into its network and systems and expected to spend $610 million in 2006–2007. This was part of the ongoing $1.4 billion investment in the NGN platform, designed for more efficient delivery of voice, data, and all forms of multimedia content.

Telecom was determined to remain highly geared for long-term profitability regardless of any legislative changes. It even appeared repentant; or at the very least was revising its public persona, as the nice, friendly, helpful telephone company. In an address to TUANZ Gattung insisted Telecom was committed to a fair game, and not going to be obstructive, attempt to turn back time, mount rear-guard action or hide behind legalistic actions. It would act swiftly, she said, to come up with invigorating wholesale arrangements, and true to Telecom’s glossy charter would provide a consistent service delivery experience for everyone concerned, launch new intermediate products and offer greater transparency and communication.

Most ISPs, and others keen to invest in the new network environment, were watching one thing, whether Telecom would affirm this new attitude with action. Within days of the government’s LLU announcement Telecom’s Xtra network faced a series of intermittent Internet and email outages. This was blamed on ‘a fault with a power supply,’ then ‘a faulty load balancer’ and other unstated issues. Despite claims all customers were back on line, many were still having problems over the following week. A customer service spokesperson explained outages were a fact of life on the Internet, which was itself ‘a best efforts’ service. “ISPs can’t guarantee a continuous uninterrupted service,” said customer services manager, Kelly Moore. That statement alone raised a few eyebrows.

By the end of May, Telecom’s gracious offer to refund affected customers with the equivalent of four days’ access – an estimated $3.25 per customer – was exposed as another ‘bah humbug’ effort. It sent an email to its 600,000 Xtra broadband and dial-up customers apologising for the outage and asking them to prove they were eligible for the refund. A spokesman estimated about 90,000 customers might qualify, but couldn’t resist complaining that this would potentially cost it around $300,000. There was no mention of the thousands of dollars in lost business and the major inconvenience some customers were still facing weeks after the so-called four-day outage.

Then ISPs received notification from Telecom that it would be enforcing ‘an average aggregate data limit per customer’ and charging a premium if overall upload traffic exceeded a specified monthly data cap. While this provision had been in the small print since 2004 it hadn’t been enforced until now. Ihug, which had been offering free upload speeds on most of its accounts, would have to add upload costs to customers’ download data caps from 1 July. Incensed at the move, it asked the government to exclude such charges in the new regulations.

THE MEGABYTE SQUEEZE In keeping the market to itself Telecom had not only made it uneconomical for competitors to wholesale services from its network, it had kept such tight data caps on accounts that ISPs and their customers had to pay a huge premium if they were to stream or download significant amounts of music, film, or rich data. In 2005 most ISPs couldn’t get 2Mbit/sec download speeds from Telecom and were restricted to upload speeds of 128kbit/sec. By May 2006 many Telecom Xtra customers had their accounts upgraded to 3.5Mbit/sec speeds. There were an estimated 300,000 broadband subscribers, around 8 percent of the population, including 100,000 customers outside of Telecom’s Xtra network, many now operating at 2Mbit/sec with the hope of much greater things on the horizon.17

At the end of May, CallPlus and Ihug, motivated by the fact TelstraClear had struck what amounted to an in-house deal, were now were leading the charge for (UBS access. They believed the Commerce Commission should now deliver on its original promise of 7.5Mbit/sec, opening the way for faster sub-$30 broadband accounts. By June they had their wish. The Commerce Commission granted CallPlus and Ihug wholesale access to the fastest broadband that Telecom could provide. However this was at a price slightly higher than TelstraClear. The Commission also made a ruling to uphold Vodafone’s application to interconnect with Telecom’s network. It ruled Vodafone had the right to use a mobile as an option for a landline service, that neither party could charge each other to receive local calls and Telecom could not charge its customers a higher price to call a Vodafone phone.

ISPs, industry groups and government officials were eager to chart the way forward and hammer out the specifics of the new environment. InternetNZ, ISPANZ, TUANZ, Orcon, Ihug, CallPlus, and others, deeply concerned about the outcome of the LLU decision, were engaged in a round of meetings with each other, with ministers, bureaucrats, the MED, Commerce Commission, and the Telecommunications Commissioner. The industry had never been so organised, motivated, vocal, and determined to have its way. There was general agreement that the future framework for competition needed to be foolproof and futureproof. Certainly the right balance had to be struck and this was being clearly articulated in submissions on the Telecommunications Bill. If legislation was too prescriptive and literal it might limit access to smarter next generation technologies, and scare off potential investors. If it was too loose it might need to be revisited within three years and result in the kind of legal wrangles over detail that squashed competition in the first place.

As ISPANZ vice president Scott Bartlett said, the real battle had just begun. While ISPs now had LLU in theory, they needed to sort out access charges, backhaul charges, rack design, tenancy costs and third party access; and that was just the big ticket stuff. The legislation needed lots of detail, particularly in empowering the Commerce Commission to make rulings, otherwise it could take several years to sort out the mess. “Making LLU work for everyone is important, the more people we have singing from the same song sheet the better.”

Bartlett voiced widespread concern about the price for access to the copper tails into homes and business. “If it’s not economically viable no one will benefit.” And he didn’t believe there was room for every ISP to start putting their DSLAMS on the Telecom network. “We’re still a small market with relatively low uptake of broadband and I don’t think there’s room for four or five new networks. If we start getting independent networks throughout the country it would be such a waste.” ISPs needed to work together, perhaps using ‘port credits’ as a way to ease the burden of investment and broaden coverage. If someone built a network in Christchurch and another company built in Auckland they could offer port credits on each other’s networks.

TELECOM READS TEA LEAVES The dilemma remained whether unbundling alone would be enough to energise the market into a new era of rapid broadband growth. With a foot now in the door there was talk the market needed to be opened up further and overseen by a regulator who was more than a paper tiger. In fact InternetNZ executive director Keith Davidson wanted to see the incumbent make solid progress on voluntary structural separation. In other words he wanted Telecom to put forward its own plans for restructuring into wholesale and retail divisions and to begin negotiations with UBS wholesalers as soon as possible.

The Internet lobby group wanted an immediate improvement in broadband packages including faster upload speeds, neutral peering for Internet traffic between service providers, and the elimination of churn fees and installation delays. In short it wanted to see a transformation from Internet-delivered, Telecom-configured services and products to an era of open access end-to-end connectivity.

Telecom quickly ‘read the tea leaves’24 and publicly stated its commitment to supporting the new regulatory environment. Rather than waiting for the government to determine the conditions of any split, it made a pre-emptive move, offering a form of operational separation in the hope of influencing the impending regulatory environment. Just as the draft legislation was tabled in Parliament, Telecom announced that its wholesale operation would run as a physically separated business. Information flows between wholesale and retail divisions would be strictly controlled and there would be a series of public, legally binding undertakings on service delivery, and transparency, monitored by an independent oversight group.

The concept was not unfamiliar to Telecom, which had entered into ‘voluntary separation’ when it was first deregulated in 1989. Two years later it reconsolidated after its sale to Ameritech and Bell Atlantic. This time it favoured the British Telecom model, customised for the New Zealand environment. “There is a difficult balance to strike here. We need to both respond positively to the demands of the new environment and serve the interests of shareholders. We recognise shareholders need to earn appropriate returns over large-scale investment decisions. We believe the separation model we are currently introducing will meet the needs and expectations of key stakeholders and level the playing field for all participants – while at the same time supporting the right incentives for further infrastructure investment in the industry,” said a statement in Telecom’s 2006 Annual Report.

Since the government announcement, Telecom had forced a friendlier face but its competitors were understandably wary about how the LLU process would play out, in particular how the incumbent who fought so hard to keep them at bay would treat their demands for equal access to the loop. The potential obstacles were legion. At the very least competitors would need access to Telecom exchanges to co-locate their DSLAMs. Would there suddenly be a claim that there was no room in Telecom cabinets for co-location of equipment? If alternative cabinets were required would Telecom oppose this through the Resource Management Act (RMA)? Would local authorities allow the RMA to become an obstacle? Would competitors need a separate door to Telecom exchanges? Could they have access without a Telecom person being present? Did each exchange need a partition between competing provider’s equipment? Would ISPs need to provide their own power and air conditioning? If the latest DSLAMs cost only $45 per port would it be fair for Telecom to charge $150 installation fees? Would Telecom avoid doing anything to help its competitors until legislation forced its hand or would it begin offering more attractive deals in the interim to prevent customer churn before LLU ground zero, sometime in 2008?

There was also the question of Telecom’s connection ratio and the quality of service it might be able to offer. In other words how many users could share a single broadband link? British Telecom had a ratio of 50:1 for home users and 20:1 for business lines. It was alleged in some quarters that Telecom’s ratio could be as high as 140:1.

Another obstacle facing new competitors was a significant shortage of technicians, engineers and customer service people who had skills in installation and maintenance and network repairs. Telecom had exclusive contracts with Downer and Transfield and TelstraClear had a deal with Cabletalk to look after its networks. The other major player, GDC, had gone belly up earlier in 2006. Perhaps third parties would have to sub-contract Telecom’s contractors to install their equipment on the network? In the meantime, there hadn’t been a lot of industry training and many of our best engineers and technicians had been lured overseas by lucrative contracts, further reducing the skills pool. No one could know with any certainty what the new environment would look like until the real turf wars began, access was opened to the last mile and full disclosure provided to Telecom’s new wholesale customer-competitors.25

NEW NETWORKS IN WINGS So who was going to benefit from unbundling? Certainly TelstraClear, Ihug, Slingshot and CallPlus, Orcon, Maxnet, Quicksilver, Iconz, and the dozens of smaller ISPs who were eager to deliver new speeds and services to their customers. Because of the high cost of creating national coverage there were widespread plans to co-operate by ‘port sharing’ or ‘credit swapping’ on each other’s equipment. At least three new nationwide high-speed DSL2+ networks were likely to emerge once the devilish details of LLU were worked through and embedded in legislation. One consortium of investors was planning an independent wholesale network once it could install its equipment at Telecom’s exchanges. Other bandwidth wholesalers with independent city and regional networks; Broadcast Communications (BCL), TelstraClear, CityNet, Vector, Inspire, FXnet, and others, were also looking at alliances to expand beyond city boundaries.26

Orcon had committed to build a co-located network in the main centres which would wholesale to other service providers at a fair price and planned to install equipment in rural exchanges including New Plymouth, Whangarei, and Masterton. “This is such an awesome opportunity to build an independent network for the whole country that runs at 24Mbit/sec with ADSL2 or VDSL. Why would you spend all that money just to do a couple of exchanges in one area?” asked operations manager Scott Bartlett. He reminded ISPs the real cost was not hardware but in operating the network. “We are talking millions; we couldn’t do any of our plans for less than $5 million and that’s a fraction of the number it’ll end up being. It would be naïve to think that even $20 million would build a nationwide network.”

Ihug was still determined to invest at least $20 million over two years to reach 100,000 homes through piggy backing on the local loop, despite its Australian parent iiNet having suspended share trading due to temporary difficulties. With ‘sensible LLU pricing’ “There could be further investment,” said Ihug chief executive Mark Rushworth. A fair deal would be $12–$16 per line, per customer, per month. The OECD average was $18 but for those who had unbundling for a number of years it was closer to $12–$16. Of course the more lines you took, the greater the room for negotiation. He suggested Telecom would keep DSL wholesale prices quite high to get a reasonable return on its DSL2+ equipment. However Rushworth warned if small ISPs decided to undercut the market just to get customers, Telecom and others were likely to drop their prices in response, making LLU payback longer and more difficult. “You have to be careful or you’ll end up in a death spiral; you need a reasonable return for your investment in equipment. That’s why everyone’s watching what these prices will be.”

He wasn’t so concerned about speed, believing the environment to be similar to Perth, Australia where iiNet had co-located DSL2+ DSLAMs on the Telstra network and customers were getting 22–23Mbit/sec speeds, depending on distance from the exchange and the quality of copper. ‘Heat maps’ of customer density and penetration in Perth revealed 90 percent of those using its DSL2 service got greater than 8Mbit/sec, close to 50 percent were above 12Mbit/sec and 30 percent were getting 20Mbit/sec plus. That was sufficient to deliver most triple play services; video streaming using a MPEG4 codec, for example, only required 4Mbit/sec speeds. While Ihug would focus on voice and rapid access it would also be looking closely at video-on demand to keep the customers ‘sticky.’ It was interested in content and business offerings based on guaranteed QoS but Telecom could currently offer these.

Ihug saw mobile or ‘quad play’ as more attractive in the interim. “Mobile will become increasingly important for customer retention. You’ll see Telecom come out with bundles where broadband is cheaper if you also have your mobile with them,” said Rushworth. That approach had proven a more sound business proposition through BT Fusion in the UK and Free Internet in France. Eventually a multi-mode handset might be delivered as part of the broadband service, operating with IP on the home line as well as wi-fi and GSM. “Any IP calls are free and they can roam on their home wi-fi network or the public wi-fi network or hand off to Vodafone beyond that,” said Rushworth.

TelstraClear would remain the biggest player in the new environment, and most likely the quickest to move. It would relish the opportunity to deliver its own DSL service enabling it to pick up on the aborted 2001 plan to deliver high speed services across the country. “We’d like to be treated the same way Telecom treats its other wholesale customers – we are their largest customer so they should be actively trying to sell to us. To date that hasn’t happened,” said TelstraClear spokesman Matthew Bolland. LLU was fundamental for TelstraClear which remained committed to the residential market and had built a new residential contact centre in Kapiti to service that sector. “We have spent $1.5 billion, have 1400 staff and have been fighting for ten years to get access so don’t think we’re going to sit by and not make the most of this opportunity.”

Across the Tasman its parent Telstra has been forced to open up its coveted network to all comers since 2000, and so far around 150 rivals had jumped at the chance. They mostly had their own cabinets within Telstra’s exchanges with racks of DSLAMs27 or had deals with others to provide that access. Bolland said some of the potential nastiness between competitors was prevented by the fact that most of the contractors who installed and serviced the equipment were neutral. Standards were in place to ensure no one did anything to disrupt any other party’s service. However Telstra remained a reluctant party to unbundling. It was still grumping away with the deregulation blues, watching its market share and share prices erode while it played catch-up with new network technology, dumping thousands of workers to become more competitive. It had already milked the ‘sorry we lost the keys to the exchange’ suite of delay tactics, and earlier in 2006 increased its rental on the copper lines to the likes of Optus, Primus and iiNet by $8 per line per month to $30.

Bolland admitted it had been a pretty ugly environment for investment in New Zealand, and along with many others, he was waiting for the rules of the new game to firm up before announcing any financial commitment. “When you’ve been fighting every inch of the way as we have, you are very aware the only reason these things are happening is because the government has intervened.”

THE FIBRE DILEMMA The new, tougher regulatory environment, outlined in the Telecommunications Amendment Bill, introduced into the House in June 2006, was expected to rapidly bring an end to repressive data caps that curbed Internet use, and hasten the arrival of pricing models and speeds more aligned with the marketing promises of the always-on digital lifestyle. One thing was certain: business and home customers, investors, ISPs, and equipment vendors who’d been held back from tasting smart multimedia services now available in many parts of the world could at last see something resembling a rich 3D media pot at the end of this rainbow.

New networks that would bypass old technology and go direct to DSL2+ DSLAMS and modems capable of delivering up to 24Mbit/sec download speeds were at last on the agenda. Hopeful competitors were promising Internet surfing would power up to 10–24Mbit/sec, finally bringing the country in line with the streamlined services enjoyed elsewhere.

Telecom’s roughly 580 exchanges had in recent years been reduced to around 440 and with the advent of its NGN overhaul, it was believed this number would be further reduced. What you would get instead were more roadside cabinets with fibre-optic cable back to the main exchanges shortening the distance copper had to travel to homes and businesses. The last mile or local loop typically used copper wires to carry phone and Internet services but there were now concerns Telecom, forced to concede access to its copper lines, would ramp up its fibre roll out, taking its cabinets closer to the curb and increasingly bypass its main exchanges. Unbundling fibre was not part of the proposed legislation and where it proliferated, LLU competitors could literally be cut out of the loop.

Nortel New Zealand managing director Rob Spray said many telcos around the world were taking fibre direct to small street cabinets on the curb and making the copper loops so short it made unbundling the loop physically difficult. A major regulatory argument was already brewing across the Tasman and in other parts of the world, with new network investors fearing their investment in DSL would be lost if fibre got any closer to their customers. “They’re saying unbundling the local loop from the exchange is yesterday’s argument. They’re taking fibre to the curb, to the business and eventually to homes and you don’t unbundle that unless you start unbundling wavelengths,” said Spray. In Australia those investing in fibre wanted a safe harbour for the next ten years, and a promise competitors wouldn’t be given access. They were threatening to cut back their investment unless the regulator acted in their favour.

However Ihug chief executive Mark Rushworth was sceptical about suggestions Telecom was trying to bypass the copper loop with its NGN. “What is happening in some built-up areas and new subdivisions is not necessarily an indication of what will happen out in the suburbs of most cities and towns. A lot of those next generation cabinets will still be situated in existing exchanges – it just doesn’t make economic sense unless Telecom uses this tactic as a defensive blocking strategy.” Rushworth’s market intelligence was that even if the current 440 or so Telecom exchanges were reduced to 200 over the next year or so, Telecom’s NGN would still give 70 percent coverage for local loop investors. However people shouldn’t expect instant changes: the benefits were likely to filter down, with most activity initially in high-density areas where the business case; read ‘the quickest profits,’ made most sense.

Meanwhile Telecom attempted to reassure the market there was still plenty of life left in its copper network. A team of Telecom executives was evaluating a faster roll-out of interactive next generation broadband services based on ADSL2+ which it had been trialling since mid-2005 in conjunction with its technology partner Alcatel. The stated benefits were better traffic handling, further reach and faster speed, up to 15Mbit/s download and 1Mbit/s up. Distance from the cabinet was all important, but most homes would get the advertised performance. Telecom was also keeping an eye on very high-bitrate DSL (VDSL), which had the potential to offer speeds of up to 52Mbit/sec.28

PIPES OPENED UP Telecom had been ‘squeezing the sponge,’ as one commentator so aptly put it, but the sponge and competitor and customer patience with its anti-competitive attitude had run dry. Under new chairman Wayne Boyd the new-look Telecom had a major challenge to win back public trust and confidence. It needed to act rapidly to improve its service levels and help desk wait time. One report showed there were more customer complaints about faults in 2006 than any time in the previous three years and it was taking two weeks or more to sort them out. Many complaints resulted from winter dampness when old or decaying copper, particularly in suburban and outlying areas, once more proved the cable was well past its use-by date.

Expectation of true broadband connectivity had taken a giant leap forward with imminent unbundling legislation forcing Telecom to open up the throttle for bigger bandwidth to its customers and competitors. An ISP survey released by Statistics NZ in March 2006 showed dial-up subscribers were in decline and broadband subscribers had jumped by a third over the previous six months. There were 57 ISPs, 8 fewer than in the previous period. Internet subscribers totalled 1.3 million and just over a million were residential users with 70 percent still on dial-up connections. Within a year many regional and smaller providers, would be absorbed by those higher up the feeding chain, in a rush for customers, coverage, specialist skills, and innovative technology.

Wholesale broadband to ISPs other than Xtra had been choked back to 2Mbit/sec maximum speed, then edged out to 3.5Mbit/sec as legal and competitive pressure mounted after the TelstraClear determination. Then suddenly it was a free-for-all. Telecom, as required by the Commerce Commission, announced that from October 2006 all ISPs would get unconstrained broadband. By the end of the month most ISPs had announced new plans for faster Internet access, starting as low as $20–$30 a month, and at the higher end data caps and speed bumps almost disappeared. Xtra led the way, giving its broadband subscribers access to the maximum download speeds their copper telephone lines could cope with.

Those fortunate enough to live in a neighbourhood close to an exchange, with high grade copper lines and low broadband use, could potentially achieve speeds of up to 7.5Mbit/sec. In most areas, however, it was more likely to average 2–3Mbit/sec. Existing customers on 256kbit/sec and 2Mbit/sec monthly plans were automatically boosted to the new full speed. The best prices were available to customers who had a phone account with Telecom or Ihug, which had recently entered the home phone market; otherwise there was a $10 premium.

All but gone were the monthly data caps that either resulted in a premium being charged for every megabyte over the limit, or Internet accounts being choked back to dial-up speeds. The only restrictions were based around a fair use policy; if someone was downloading hundreds of gigabytes of data and disrupting network use during peak hours, a level of traffic management applied.29

Telecom marketed its new broadband plans as ‘blazing’ and ‘super fast’ but the fact was this was like a scene from Back to the Future, only the names had been changed to prevent embarrassment. Back in 1999 Xtra’s debut JetStream full speed account promised to deliver whatever speed the line was capable of. Depending on line conditions and distance from the exchange pioneering users could get up to 8Mbit/sec downstream and 256kbit/sec upstream for $90, although there were severe data caps of 500Mb which meant using more than your monthly quota could be expensive.

However Telecom had stopped offering full speed accounts in 2003, supplanting them with a range of staggered accounts from entry-level 128kbit/sec and 256kbit/sec up to a maximum of 2Mbit/sec then more recently 3.5Mbit/sec. Those on grandfathered ‘full speed accounts’ were now paying $60 a month with the data cap extended to 600Mb and given the option to shift across the 2Mbit/sec accounts with a 128kbit/sec upstream cap and a data cap of 1Gb all for the same price. Many users were furious; they were being asked to step down from the luxurious high speeds they had been sold as pioneering DSL users with very little alternative for those who wanted to expand their use than to take the lesser road.

Telecom general manager of consumer marketing, Kevin Bowler, said it removed unconstrained speeds because customers were not buying them in large enough numbers. It decided to bring them back because customers were increasingly using the Internet to download larger files, such as videos, music, gaming, and other interactive services. InternetNZ executive director Keith Davidson had another view, claiming Telecom removed unconstrained speeds for most customers because it was concerned about ‘underinvestment in the network’ and its ability to cope with large numbers of customers on high-speed plans. “This was a way of manipulating the network to get the traffic into manageable segments rather than investing in the network itself.”30

Xtra’s new monthly rates gave an indication of what was ahead for all ISPs. A basic plan at $30 a month attracted a 200Mb data cap but the speed restrictions had gone. The $30 Go plan retained a 1Gb data cap, slowing to dial-up once the limit was reached. The Go Large offering at $50 did away with the data cap and cranked up the speed to line capabilities. Plans for those who wanted to send a lot of data, boosted the 128kbit/sec upstream speed limit to 256kbit/sec–512kbit/sec but brought back data caps (2Gb – 30Gb), ranging from $50 to $150.

HOW COME FAST IS SLOWER? Wholesaling had only made up 10 percent or less of Telecom’s business but with Telecom Wholesale having Telecom Retail as its biggest customer, and the rush of activity predicted through UBS, LLU, and their variants, that was about the change. Telecom Wholesale general manager Matt Crockett insisted Telecom Retail would be treated the same as everyone else in the market for NGN products and services. Strong measures were in place to ensure equivalence, with monthly tracking and reporting around commercial terms, provisioning and lead times. “In the past the bulk of wholesaling was on fixed line and we had the bulk of that to ourselves but now TelstraClear and others have their own networks. WiMax is coming, Vodafone is getting into the wholesale space, and we have to make sure we have better products and service. We can’t take anything for granted.”

Telecom Wholesale already offered the full range of voice, data, interconnection, and networking services for residential or business end users and more than 100 service providers. Major wholesale products included UBS, a layer 2 connectivity service enabling customers to deliver broadband to their end users, and the newly launched QoS-based business grade Unbundled Network Service (UNS) to compete with Telecom’s xDSL One Office. There was also the resale of Telecom retail products, at around 16 percent discount. Telecom Wholesale would handle interconnection between different service provider networks and international services through Telecom New Zealand International.

Crockett said his job was to provide a healthy portfolio of choice for differing appetites. A number of providers wanted to run totally independent end-to-end networks but for others there would be equivalent services that didn’t require that kind of network investment. Telecom was in consultation over next generation Naked DSL to achieve higher QoS. There would also be ‘white label’ DSL or ISP-in-a-box solutions that could be re-branded by businesses outside the traditional telco arena, including The Warehouse or Dick Smith. Crockett knew he had a lot of work to do to grow the business aggressively. “I can only do that by helping my customers to succeed but we can’t keep doing that dropping our pants on pricing.”39

Then, just as everyone thought unconstrained access would finally give everyone more for less, there came a backlash. Telecom’s wholesale customers Ihug and CallPlus claimed they had to spend millions fixing problems on their networks since the broadband gates were opened. Thousands of customers lost connections, sometimes for hours at a time. Both ISPs said the problems were occurring with Telecom’s wholesale or UBS offering and had become progressively worse since June.

After the move to 3.5Mbit/sec plans, Telecom commissioned Alcatel to audit the state of the network. It estimated that 23 percent of customers were not achieving maximum peak download speeds, because of a range of variables including interference, the condition of the copper and distance from the exchange. After the upgrade to full unconstrained access Alcatel reported at least 5 percent of customers were actually receiving a reduction in Internet performance on what they’d had previously. Complaints about slower connections were rife, and many in the industry said the slowing of Internet speeds was far greater than that being admitted.

CallPlus and Ihug had not expected Telecom’s kneejerk reaction when they took their case to the Commerce Commission demanding equal access to unconstrained bandwidth. Telecom had opened up the throttle for all its Xtra customers, effectively forcing other ISPs to offer a similar service. “They went beyond what they were required to do, voluntarily moving all their retail customers to full speed, but not their wholesale customers who were kept on the old plan.” That meant Ihug and other ISPs were in a loss-making situation when it came to delivering the new all-your-line-can-handle Internet accounts. “We don’t make money out of broadband at all unless we bundle it with other things. We have to buy that at above retail cost and we lose money on every single customer we sell to,” said Ihug general manager of regulatory affairs, David Diprose.

Most of the complaints about slower speeds, said Diprose, were not because of the speed of access but congestion, which was seriously slowing down the Telecom network. With everybody now trying to use full-speed it was like the traffic jam at peak hours in Auckland. “It just doesn’t provide enough backhaul capacity to cope with the growth in use.” Telecom insisted it was adding capacity to its network to cope with growth. “We have made some backhaul adjustments and are in the process of revealing those. Everyone gets the average and for most people the average is better and for some its worse. That’s a feature of the Internet,” said Matt Crockett.40

Ihug chief executive Mark Rushworth said his company had spent $500,000 enabling its network to support the large numbers of customers who lost connections. It had increased call centre staff by 20 percent and compensated some customers with credit notes. Ihug was fortunate that it had the financial support of its parent group Vodafone to invest in the network problems, but the long-term damage to its brand and reputation could not be calculated. It even asked Telecom for compensation but wasn’t holding its breath. CallPlus chief executive Martin Wylie said the company spent ‘millions of dollars’ to manage the customers in the call centre and on fixing network problems. “The industry is trying to absorb it, which is putting more pressure on companies that are already finding the going very tough. We are making a loss on every broadband customer and the technology that Telecom provides is not performing, which forces us to put in more investment and put more staff on.”

Telecom’s Matt Crockett said the company first alerted customers in mid-September and within five weeks had fixed 95 percent of the problems. “We struggled to get to the root cause, but we did isolate it and fix it.” Telecom accepted that it needed to do more work on its wholesale service and would continue to work with wholesale customers to avoid problems in the future. “There are going to be issues around a fast-moving market and complex technology like broadband. None of us are happy with the experience for our end users in the past few months.”41

A week later Telecom chief operating officer Mark Ratcliffe insisted ‘relatively few’ of the problems tracked back to an issue with the Telecom network, insisting Ihug and Slingshot were ‘not geared up’ for the large migration of customers onto new broadband plans. “As well as a massive migration to broadband, we are fronting on a global basis another spam war. About 95 percent of the email traffic we carry through the Xtra network is spam.”42

THE DEVILISH DETAILS Once the direction was set by the government in May 2006, Telecom and the Telecommunications Carriers’ Forum including InternetNZ formed a working group with the ISP industry to look at methods of achieving unbundled local loop and naked DSL ahead of legislation. They worked on the technical and operational standards for LLU as a basis for either regulated or commercial agreements. Many of the InternetNZ proposals were taken up, including a review of spectrum allocation to improve community wireless access, and the need for operational separation of Telecom. A suggestion was also made to develop a fibre-to-the-home roadmap, said InternetNZ.43

Maurice Williamson, the opposition’s IT and communications spokesman and a member of the select committee that reviewed the submissions on the Telecommunications Amendment Bill, insisted some historical context be considered. He and others in the National Party along with Business New Zealand, the Business Roundtable, the Shareholders’ Association, and Federated Farmers opposed the government’s unbundling announcement as a confiscation of Telecom’s property rights. His role as an evangelist for all things technical hadn’t lessened since he was the country’s first IT minister and his ability to spin a great yarn seemed to have improved with age:

“Clear Communications’ main owner British Telecom came down a couple of times and badmouthed them for not doing this and that, saying how outrageous the environment here was. So I took the buggers on one day. I had a big clipboard and I went through all of the things that I thought were just appalling. I said, ‘The dominant carriers refuse to allow interconnection at proper rates, they’re using technology as a non-tariff barrier to stop you getting entry; you can’t use signalling at carrier level to move caller line ID across the switches,’ and on it went. Then the guys at British Telecom said, ‘Oh minister, it’s such a treat to do business with you, a politician who is so well briefed.’ And I said, ‘Well I don’t think you’re going to say that after I finish because this is all the material that I’ve got from England about your company’s behaviour in the UK.’ They were just shattered. So I said: ‘Don’t come down here to the antipodes trying to tell us all the crap about opening up a deregulated market when back home you’re the worst bloody example of monopoly abuse.’”

Williamson said he was reminded of that double standard when discussions on LLU came up at the select committee hearings. “TelstraClear started making a big submission on the need to break up the monopoly to give them unbundled access. I got first question and asked Alan Freeth, the chief executive, ‘Mr Freeth, have you read the Telstra submission?’ And then everyone went absolutely silent, like, ‘Williamson’s made a mistake here.’ He said ‘Mr Williamson, we are Telstra.’ I said, ‘No, Mr Freeth, I see on your submission here that you are TelstraClear. I ask you again, have you read the Telstra submission?’ He said, ‘Oh, they haven’t made a submission on this bill.’ I agreed, ‘No, they haven’t, you’re dead right, but they made one on a very similar piece of legislation in Australia. Have you read that?’ He hadn’t, so I read some of it to him. ‘The government’s unbundling local loops was an outrage, the government’s interfering in property rights was a disgrace…’ And I just kept reading these quotes. I said, ‘You agree with these. Tell me when I go through them whether you agree with them or not.’ And he was just glaring at me.”

The next submission in the series was from Telecom and Williamson insisted on equal treatment when evaluating the presentation put forward by new chairman Wayne Boyd and CEO Theresa Gattung. “When Theresa finished, I got the first question again and said, ‘Theresa, could I ask you … have you read the AAPT submission?’ She didn’t hear me take Telstra to pieces because they were out of the room. ‘AAPT? Mr. Williamson, those are not…’ I said, they made a submission in Australia and I’ve got a mate in the Australian Parliament who sent me copies of all their papers. Can I read you what they say? ‘The dominant carrier is holding on to an abusive monopoly position so the government must break it up, must make it available, must give access to all new players so they can get realistic access at the right price’ and on it went. There was just silence in the room so I said, ‘You obviously agree with all this?’ She said, ‘No, that’s AAPT.’ ‘Well hang on, don’t you own AAPT? Isn’t it your company? Well I’m sorry but I really had a gutful for over a decade now of telcos when they’re here saying how dreadful it is and demanding it all, and when they’re overseas saying how dreadful when the same thing happens to them there.’”

Williamson said he has nowhere near the regrets some people imagine he might have about not getting heavier with Telecom in the early days of deregulation. “It wasn’t that I didn’t want to. I thought Telecom were bastards and I told them so, and I tried to keep a sword dangling as tightly over their throats as I could. But in the end there were certain things and some of them involved property rights. I’m on the board of a private company here in New Zealand and every time we’re investing overseas, my very first question before we go into any of the detail of an investment is, ‘How is this country for respecting private property rights?’ And I think the local loop unbundling stuff that’s happened, and the breaking up of the private companies by force will have a significant impact on foreign investment coming here, whether it’s to invest in the electricity network, or even in the broadband investment side of things. I have no problem about compensating private players; we have to do it all the time when we take land off farmers for roads. We try to negotiate before a compulsory acquisition but we well and truly compensate for the land loss.”

Williamson was concerned Internet development may be one of the main things that suffered through lack of investment. “There’s a huge amount of investment needed in our total network. Our exchanges are still bloody crap, the copper to a lot of rural and remote New Zealand is such rotten copper with so many breaks and nodes that it needs a major upgrade. The cyber backbone and backhaul stuff needs dramatic upgrades and the nodes in the suburbs are going to require a huge level of investment. I hope I’m proven wrong, but really I worry whether Telecom is prepared to do it, if it has to give away what it’s invested in to anybody else that’s playing in the game.

“There’s obviously a lot of cherry-picking and a lot of low-hanging fruit in the CBD where you could easily make money from rolling out fibre and Internet infrastructure but it needs to go out to everyone. In fact it’s more important to get to rural and remote New Zealand so they don’t have to drive 100km to town to register their vehicles and then home again. You have to think, what is that taking off their GDP for the day in terms of productivity and the fuel burned which adds to Kyoto breaches and the like when it could all have been done over the Internet?”

INFRASTRUCTURE INCAPABLE Submissions to the Telecommunications Bill were still being worked through in September when someone with intimate and long-term knowledge about the Telecom network dropped a bombshell. Former Telecom chief technical officer Murray Milner, representing the IPENZ, believed Telecom’s copper cable was a major obstacle to the country moving into the top quarter of broadband OECD figures. In presenting IPENZ submission on unbundling, Milner said $1.5 billion needed to be spent replacing copper cables with fibre-optic cables if New Zealand was to match the infrastructure now available in countries such as Iceland, Japan, and Denmark. That kind of spending could reduce the length of the average phone line from 2km to 800 metres and result in 90 percent of New Zealanders having access to 5Mbit/sec speeds, the goal set by the Digital Strategy.

While the institute, which had 10,000 members, supported the regulatory update and believed unbundling was ‘long overdue,’ prices for access to Telecom’s network would have to be set at a ‘fair level’ to ensure there was sufficient incentive for Telecom to make the necessary investment in fibre. The only alternative might be for the government to dig into its own pockets. Milner dismissed the possibility that mobile broadband products such as Vodafone’s 3G broadband service and Telecom’s T3G network could provide mass-market alternatives to fixed-line broadband in the near term. He said that would require 100,000 cellphone towers to provide 5Mbit/sec connections to 90 percent of New Zealanders and cost about $10 billion, roughly the same as fibre to the home of all New Zealanders.

The availability of contract labour meant telcos could probably invest only $100–$200 million a year on replacing cables. In order to justify the $1.5 billion investment, half of New Zealand households would have to spend $100–$150 a month on triple-play services, such as broadband, phone calls, and Internet TV. That was likely only if the dominance of Sky’s satellite-based pay-TV service was broken and sports events and movies were delivered to consumers in interactive form over broadband, said Milner.44

Meanwhile InternetNZ executive director Keith Davidson was disappointed Telecom was continuing to claim its operational separation proposals were better than those implemented by British Telecom. The Internet lobby group had previously detailed the stark differences to the select committee and was frustrated Telecom chairman Wayne Boyd hadn’t made himself better informed on the issues. “Telecom’s operational separation plan is clearly inadequate because it does not include separation of the core network, where the ‘enduring bottleneck’ exists,” said Davidson. British Telecom’s model created a separate division for the network assets, providing equal access to all ISPs including BT wholesale services group and retail operations.

Telecom’s plan only provided ‘Equivalence of Inputs’ for three services. “Many ISPs have raised the issue of Xtra having favourable terms, and the recent Telecom retail pricing announcements have only highlighted the issues of Telecom treating its own ISP differently,” said Davidson. “Legislation for an operational separation plan is essential.” InternetNZ had attempted to provide a convenient mechanism for including this provision in its submissions.45

ISPANZ also issued a last-minute warning for its members and the public not to be fooled by Telecom’s claims. It rejected the attempt by Telecom to push its ‘inferior operational separation plan,’ and seriously questioned its commitment to industry co-operation and the spirit of the government’s policy direction. It railed against Xtra’s latest retail plans that created a price squeeze on ISPs and then discovered that despite requests to adjust ‘the inadequacies of its operational separation plan’ no amendments had been made.

ISPANZ director Graham Walmsley said the organisation of 30 ISPs wanted operational separation but not in the form Telecom was proposing. “Telecom seem to think that if they make the claim often enough people will believe them. Telecom clearly does not ‘get it’ and this is of major concern. We are seeking a level playing field where we can compete in a competitive market with equal access to the monopoly infrastructure. That infrastructure must be separated into another division – accountable and incentivised separately, as British Telecom has done.” ISPANZ said in the Telecommunications Amendment Bill the government must have the power to mandate separation without further legislation and that the threat of further structural separation must remain.46

The draft bill only provided for a modest accounting separation regime for Telecom. InternetNZ had made stronger written and oral submissions, similar to the statutory framework in Australia. This was picked up in the final draft. Its proposals were centred on the public good of the Internet, its users, and the economy. “The real challenge will be implementation . . . The government will need to get the separation plan right and the Commerce Commission needs to be given the money and people to make sure the plan can be enforced,” said InternetNZ in its December 2006 newsletter.47

The Telecommunications Amendment Act (No 2) was passed into law on 18 December 2006 with urgency. The revised legislation upgraded the Telecommunications Act 2001. When the bill went into select committee it proposed separate accounts for Telecom’s business arms, but after six months of consideration it came through with a much tougher regime. Telecom would be forced to undergo a three-way operational split, with retail, wholesale, and network arms. The split would take place after public consultation, and both David Cunliffe and the Commerce Commission would be given powers to ensure that it happened in a timely fashion with wide-ranging powers to monitor and enforce access to the local loop and regulate other aspects of the telecommunications network. Cunliffe said history had been made and the ammended lawl would bring faster and better broadband service.

The speedy passage was in stark contrast to the glacial movement of previous attempts at telecommunication regulation where successive governments had been persuaded to remain with a relatively light-handed approach. Labour’s former Communications Minister Paul Swain had fought to unbundle the local loop, but was overruled by Cabinet. Under National, Telecom persuaded then Telecommunications Minister Maurice Williamson that regulation was unnecessary, but Williamson later made it clear Telecom had not delivered on its promises to him. The new regulatory regime looked for the first time like a ‘hands on the shoulder’ style, as opposed to the wet postage stamp approach of the past. There was the potential for up to $10 million in fines if Telecom was slack in following through on operational separation into wholesale, network access and retail business units.

The comprehensive regulatory package, modelled on similar laws around the world included unbundling, at the local loop, at Telecom’s exchanges and its distribution cabinets. It also provided access to so-called ‘naked DSL’ for Telecom’s wholesale providers, removing restrictions placed on real-time services and upstream speeds for UBS which the Commerce Commission introduced in 2003. The Telecommunications Commissioner would also be given new powers under the Act.48

David Cunliffe expected operational separation of Telecom would be completed by the middle of 2007. His office, alongside the MED, would manage Telecom’s split and finalise the requirements of an initial plan for separation by February. He would then invite public submissions, which could be further amended. He had the power to approve or decline the plan. “I think it would be an optimist who thought it would be perfect first time, but there is so much detail to be gone through,” said Cunliffe. He was looking forward with ‘renewed vigour’ to the development of the IT sector, a rural strategy, rolling out further digital strategy funds and new broadband wireless services.49

GRINCH STEALS CHRISTMAS While all the focus seemed to be on Telecom and its exploits, the changes in regulation and the grand plan for a new no-holds-barred environment, it soon became apparent that all wasn’t well across the road at TelstraClear. It should have been a time of celebration, of Christmas parties and thinking happy thoughts about sun, surf, and sand. However TelstraClear CEO Dr Alan Freeth was in no mind to be playing jolly old Saint Nick. What he pulled out of his sack shocked not only his staff but, when it was leaked, left the industry scratching its collective head.

In a 2000-word, metaphor-packed message he explained the company was ‘on a trajectory to disaster.’ It was being ‘out marketed, out smarted and outgunned in the marketplace,’ it ‘lacked the killer instinct.’ ‘We are too tame, too lame, and too timid to call ourselves a challenger,’ he ranted. Instead of Winding the competition and making it bleed like Uma Thurman did in Quentin Tarantino’s Kill Bill, TelstraClear was acting more like Captain Feathersword in the children’s show The Wiggles or Disney’s Bambi. Freeth insisted TelstraClear needed to become ‘pre-meditated, cold-blooded killers of our competition.’ Instead of a summer of sunshine and barbeques, up ahead was ‘a winter of despair’ based on current forecasts. An expected $14.8 million profit could become a $7 million loss unless staff worked harder in 2007.

Freeth said Telstra’s chairman had accused him of running a company that was ‘out of control.’ Customer service was indifferent at best and ‘rubbish’ at worst, resulting in a gradual loss of customers. Plan numbers were not being achieved, there were billing problems and product development needed to be sorted. The SME unit had a backlog of 300 orders which needed quicker installations. Cost containment was high on the list with the suggestion the headcount was too high and staff were travelling too much, some acting like ‘petulant teenagers’ and some focused on their pet projects regardless of what was happening around them. Some of these would be shut down. Questioned about his message, by journalist Juha Saarinen, Freeth said it was ‘a clarion call to arms’ given to about 50 senior managers and posted on TelstraClear’s intranet; even the projected loss might not have been accurate, it was intended to motivate staff. He claimed staff reaction to his Christmas tidings was ‘incredibly positive,’ and thanks to impending regulation, 2007 was full of opportunities for TelstraClear to move into full challenge position.

Chapter 21 Sidebars

IN THE BROADBAND BASEMENT New Zealand remained a digital ditherer, down in the bottom quarter of the OECD’s broadband penetration statistics. By the June 2005 OECD broadband report we’d lifted our game only marginally to 4.5 percent or 191,695 users for the year to December 2004. The OECD average was 10.3 users per 100. By the 2006 report there we were again in 22nd place for the fourth year running, a dire position confirmed in an independent survey by InternetNZ.

Only 11.7 percent of New Zealanders, about 479,000, were connected to the Internet at high speed, slightly higher than the 8 percent penetration of December 2005. The OECD average was now 15.5 connections per 100 inhabitants.

While New Zealand’s Internet access cost was considered reasonable, restrictive data caps had limited the effectiveness of high speed access. We were only saved from being the compost at the bottom of the heap by the move to free up wholesale 3.5Mbit/sec speeds. A total of 2586 residential and business broadband packages from 388 ISPs in 26 countries were analysed in the InternetNZ survey, with data from early May 2006 confirming we were indeed ranked 22nd. We had more data caps than any other country and less choice for broadband access than Australia, Ireland, and Britain. Our low upload speeds also came in for criticism.

As an ICT nation we were ranked by the annual IDC Information Society Index at number 17 of 53 countries in 2005, but when our low levels of broadband were taken into account we slipped back to 28. Australia was in seventh place, the United States at 13, and United Kingdom at 19. The Information Society Index combined a country’s ranking across computing, telecommunications, Internet, and social aspects. The index was based on the number of Web users, e-commerce maturity, home and mobile Web users. Again the not-so-magic number 22.

Earlier in May, the 2006 World Competitiveness Scoreboard of 61 national and regional economies saw New Zealand fall from place 16 to, you guessed it: 22nd, with the footnote ‘must improve broadband.’ All of this made a sham of the government’s goal to have us in the top quarter of the OECD figures by 2007 and Digital Strategy promises to have 5Mbit/sec DSL available across the country by the same deadline, possibly soaring to 50Mbit/sec by 2010. David Cunliffe agreed we were at real risk of being left behind. “No broadband means no play. Do not pass ‘go.’ Do not collect a first world income. Do not expect the next generation of the best and brightest to live in New Zealand.”16

CYBER TILLS RINGING Consumer e-commerce took a while to get established as a mainstream option in New Zealand, largely because retailers had been scared off from major investment through the failure of pioneering on-line superstores. By 2005 the shadow was lifting and while local shoppers still loved offshore bargains delivery costs were a turn off. At home on-line banking had boosted confidence, technology had begun to mature and all the signs were in place for a local e-shopping boom.

Overseas trends revealed 2005 as a watershed year for on-line sales. In Britain, 24 million consumers spent NZ$13 billion on-line in the 10 weeks before Christmas. In the more mature US market, people bought on-line in record numbers. By December 2005 on-line retail sales had reached $US17 billion, up 24 percent on 2004.18 By the end of 2006 New Zealand retail spending was estimated at about $60 billion but only about 0.3 percent of this was being transacted on-line compared with around 10 percent for Britain and 5–7 percent in the United States.19

Most major business now had some form of Web access and there were a growing number of sites able to provide a full retail experience but there were few mega-sites offering to bring the entire shopping experience together. Telecom tried to fill that space by launching its Ferrit on-line shopping portal in mid-July 2005 after finally acquiring the domain from a former host of pornography. Despite a marketing budget of $15 million, the site’s debut was unspectacular and according to statistics, Web surfers hardly paid it much attention in the early days, with around 3000 hits a day in January 2007 compared with TradeMe’s 284,000.

The goal in phase one of Ferrit’s life was modest: expanding the number of retailers and products searchable on the site. By February 170 merchants had joined and several hundred more had shown an interest. Sales queries were automatically referred to a retailer’s web site, but ultimately purchases would be directly made through Ferrit.20

That step up to a full-scale e-commerce operator came in time for what Ferrit hoped would be the Christmas rush, with widespread television advertising pushing the revamped site. Around 90 retailers were using Ferrit’s new shopping cart system to buy directly from retailers and pay by credit card. Previously the site simply directed customers to retailers’ own web sites. Ferrit was nowhere near break even and while many believed New Zealand was on the verge of an on-line retail boom, many were still wondering whether it had the formula right. Nielsen NetRatings estimated Ferrit had around 172,000 unique users in November but market analysts believed it should be attracting many more visitors, based on the marketing investment.21

New Zealand’s biggest on-line success story by miles was Internet auction site TradeMe and it was no surprise when Fairfax Holdings fronted up with $700 million in cash to acquire it in March 2006. An additional sweetener of $50 million would be paid if the company met certain earnings targets over the next two years.

Sam Morgan founded TradeMe in 1999 at the age of 23. It rapidly became the country’s most visited web site, with 1.2 million members who were expected to run 35 million auctions in 2006.22 ACNielsen said 1.265 million people had made purchases on-line to December 2006, spending around $1.5 billion. Of those 423,000 or 33 percent were regular on-line shoppers. This compared with 239,000 regular on-line shoppers in 2001. New Zealand’s on-line shopping figures were up at Christmas 2006 in line with world trends. Nielsen NetRatings senior analyst Jennifer Reddington said about quarter of New Zealanders had purchased something on-line during 2005, and the number of on-line shoppers had increased consistently over the previous two years. “People are getting more comfortable conducting commerce on-line, and anything the retailers can do to boost people’s confidence is obviously going to help drive the market.”

TradeMe reported a stronger increase in activity than usual over the Christmas period, particularly in January where people were re-gifting unwanted Christmas presents. Telecom had launched Ferrit at the end of November 2005 and expected between 50,000 and 60,000 visitors over the Christmas period. Instead it had more than 100,000 visitors in the lead-up to Christmas. About 170 retailers had signed up for the shopping site, the top selling categories being electronics, apparel, whiteware, DVDs, CDs, and clothing. US figures showed an increase of 25 percent over 2005 with shoppers spending US$25 billion (NZ$36 billion) between the end of October and December 16.23

PEER PRESSURE INTERNET COP The concept of peering is essentially a meeting of equals on neutral territory to achieve a common good. In relation to the Internet the term describes peer-to-peer networking based on agreed points for exchanging data to ensure traffic gets to its destination the most efficient way, rather than heading off to the United States or the other end of the country when its real destination is across the street.

It is the nature of the Internet that such circuitous routes are not only possible but frequently occur without the right agreements and infrastructure in place between carriers and ISPs. Peering has traditionally involved carriers and ISPs paying a general fee to locate their own equipment at a central point, and exchanging data between their networks without charging each other.

It is also common practice for content providers to peer; for example, at the Wellington Internet Exchange (WIX) providers including news media site Stuff, TradeMe, and Radio New Zealand connect to peering ISPs. Peering works in very well with Anycast technology to dramatically improve the distribution of content across geographic areas. This means multiple locations have the same IP address, so the content can be drawn from the closest source. For example, Radio New Zealand uses Anycast to efficiently deliver its programming from regional servers located at peering exchanges.46

By peering, ISPs agree to provide routing information about their network to other exchange participants to optimise traffic flows. Currently there are exchanges operated by APE, Palmerston North (PNIX), WIX, the 3 Cities Internet exchange (3CIX), NZ IPv6 Internet Exchange, Dunedin Peering Exchange (DPE), Christchurch (CHIX), and one planned in Southland (SIX).47

INTERNET INTERCHANGE The first pseudo Internet Exchange where ISPs could peer and exchange traffic was on the ground floor of the Computer Science Department of Waikato University. NZIX was the end point for the NZGate international access gateway, which was moved across to Telecom in 1996 but remained independently operated. The equipment was owned by the various exchange participants. It ran a range of protocols including older ones that had been required to support the Tuianet operations of the universities.

The first serious talk about the need for an external two-way Internet exchange came after a server in Los Angeles went down causing a major crash of the international gateway at Waikato in 1995 because there was nowhere else for the traffic to go. Additional DNS were added to the backbone and Telecom-owned network provider Netway promised to increase its own redundancy. Traffic was often routed offshore before it reached its destination in the next city or went through Waikato when it was meant for the business directly across the road. In other words a lot of traffic was taking an unnecessarily long journey to get to its destination, a term known as tromboning in the Internet community.

Exchanges kept local traffic within a city where bandwidth was cheaper and provided redundant paths so there was no single point of failure. Former ISOCNZ chairman Jim Higgins said a new DNS would ensure there would always be somewhere to resolve addressing. The DNS, ‘the heart of the Internet’ turns Internet names into numbers so the various servers know where to send mail. He was confident the proposal to establish an Internet exchange in Auckland would result in ‘something sensible happening,’ although he warned there would be a cost involved which could be passed on to the end-user.48

There had been backdoor paths between ISPs in Auckland for some time and several larger companies had alternative international routes. A consortium of Internet Prolink, the Iconz and KCBBS had established virtual links to keep local traffic local. Internet Prolink director Craig Anderson was concerned virtual exchanges would become too expensive and unmanageable. “Everyone would prefer a real exchange, managed by an independent body in a basement somewhere in downtown Auckland. That way each ISP would only need one link.”49

Initially telecommunications market rivals Clear, Telstra, and Netway Communications were looking at ways to link each other’s networks in the hope of providing more efficient routing of Internet traffic. They had regular meetings to try to agree on some kind of ‘bridge’ – either direct links or stand-alone exchanges. Telecom saw the move as an Auckland solution, while the other two parties saw it as a trial that could end up in a nationwide approach to routing between Internet providers.

APE IN SKYTOWER The WIX opened in 1998 and the APE located on Auckland’s Skytower came into play in 1999. Once these peering points were in play the bilateral agreements between Clear, Telecom and others started to multiply. NZIX at Waikato was overtaken by the Auckland and Wellington public exchanges around mid-2001 and fully decommissioned at 10.45 a.m. on 26 August 2002. It was dismantled an hour later without leaving any physical evidence of existence. The last router was a Cisco 4000 series router, owned by Netway Services.50

CityLink in Wellington had set the peering pace. The pioneering gigabit capacity metropolitan network, established by Capital Holdings, a joint venture between the WCC and 20 shareholders including Clear, Saturn, and INL, had established an open fibre network around the Wellington CBD late in 1995. Writer and technologist Richard Hulse explained the evolution of what became WIX:

In 1998, CityLink started work on deploying a BGP/4-based Internet exchange on top of the public Ethernet. One of the key design criteria was a low cost of entry. So the CityLink team took a step back and looked at the exchange problem from a logical point of view: What were the core requirements of an exchange, and how could they use the existing fibre infrastructure to create an environment that would encourage as many people as possible to join?

A key component of CityLink’s approach involved placing routers at a customer’s point of connection to the fibre infrastructure rather than in a central location. Because the exchange was distributed, CityLink needed a low-cost mechanism to allow people to peer, and so they started using Zebra on Linux as an alternative routing platform to Cisco … In addition to devolving hardware to client’s premises, the routing tables for the whole exchange were managed by way of route servers. The new exchange was dubbed WIX, the Wellington Internet Exchange. There was a fixed monthly charge for connection, but traffic over the fibre network was free. Because the network often runs right past a potential user’s door, it is easy for anyone to connect and peer. And this is exactly what has happened – even end users who could never peer under a traditional model now could. For all exchange users, access to the global Internet still necessitates the purchase of “transit” access to the global Internet routing table – from at least one ISP.

The open peering approach made a huge impact on traffic flows and latency. When the Internet first started in New Zealand, the country’s single exchange point, at Waikato University, was also the international gateway for the whole country. Two businesses in Wellington wanting to exchange data would send it to their respective ISPs, who sent it upstream to Waikato. The path time for this typically ran 50–200 milliseconds (ms). When ISPs began to exchange data directly in Wellington, this rate dropped to 20ms. End-user peering reduced this by another factor of ten, to 2ms.

In addition to shorter and faster network paths, no charges were levied for traffic sent over CityLink by way of the exchange. ISPs never see the traffic, because the exchange directs it through the shortest path to the requesting party. A number of printing companies were peering at WIX and exchanged huge publishing files at no cost. One local newspaper ran an FTP server on WIX to which pre-press companies could upload files. Some graphics houses also ran media stores for their clients, and these could be browsed at no cost, as though they were part of the client’s own LAN.

The distributed exchange environment has been likened to a market square, where anyone could trade their wares at no cost. Not everyone chooses to peer openly though, and the exchange supports all types of participation. A number of ISPs exchange traffic only with established customers and ignore any other traffic. Some simply use WIX as a neutral point to exchange data with one other party. By 2005, 130 entities are peering at WIX, with close to 1000 using the CityLink fibre for private purposes or public Internet connectivity.51

TELSTRACLEAR EXITS XS Things ticked along quite nicely for a time with all parties acting rather chummy about the mutual sharing arrangements but then peering came back into focus in a big way when TelstraClear decided in November 2004 that it wanted to start charging to exchange traffic with other parties. Both TelstraSaturn and Clear Communications had openly participated in the shared arrangement and in fact were instrumental in setting them up. Now it seemed TelstraClear was unhappy about the underlying costs of interconnection, even intimating that the smaller ISPs had been getting a bit of a free ride. It discontinued peering from WIX and APE.

Maxnet and others wondered what all the fuss was about. It paid for a connection into the Auckland exchange and CityLink took responsibility for the actual handover of traffic. It wasn’t as if Maxnet had more traffic passing through TelstraClear than was coming back its way; the flow was about even. It concluded the big carrier must have been driven by financial motives. The fallout was immediate with high hit rate web sites including TradeMe receiving triple their normal bandwidth bills because customers had to access the site in a more circuitous route than previously. Some end users were also noticing a jump in their monthly accounts.52

TelstraClear’s de-peering was closely followed by Telecom’s, which also detached itself from arrangements with WIX, causing an uproar in the ISP community. TradeMe was forced to cut a deal directly with the big carriers to access their ISP customers and a number of other larger content providers followed suit.

Some New Zealand providers were already ‘keeping local traffic local’ at peering exchanges in Auckland, Wellington, Palmerston North, and Dunedin. However the two major telcos had for three years favoured commercial agreements, where other providers paid them to send traffic into their networks. The de-peering situation had led to New Zealand–bound traffic ‘tromboning’ to Australia and the United States.

ISPANZ president at the time, David Diprose, said local data should be exchanged locally and exchanging local data internationally was nonsense. “Lack of local or regional data exchanges and the resultant ‘tromboning’ of data overseas is unacceptable.” Peering was key to the future of Internet connectivity in New Zealand and enhanced the structure of the Internet ‘through resilience, diversity and economic efficiency.’ If the telcos recognised the value of local data exchange, this would resolve some of their own national and international capacity issues, he claimed.

That provoked chastisement from TelstraClear. Spokesperson Mat Bolland said the two largest telecommunications companies invested heavily in New Zealand and if ISPs were not happy with the service they received they should look elsewhere. The price for broadband in the market was much cheaper for customers than it used to be. “Not-peering keeps it simple for us and simple for the consumer.”53 Telecom then jumped into the deep end of the debate, obviously keen to distance itself from TelstraClear’s comments. Telecom Wholesale spokesperson Melanie Marshall said Telecom was reconsidering local peering. Telecom Wholesale has been investigating local Internet interconnection against the background of LLU the arrival of new networks including the MUSH roll-outs and the possibility of more regionalisation of service delivery. “While supporting the idea that local traffic should not unnecessarily be routed outside the local area for distant handover to another network … parties should be free to choose which service attribute best meets their business objectives and budget.”54

ROUNDTRIP COSTS ESCALATE Hi-tech entrepreneur and software developer Rod Drury was concerned the actions by the major players put our economy at risk. In his discussion paper ‘Securing Our Digital Trade Routes,’ he expressed his frustration at the lack of action on broadband access and suggested the government step in and create a state-owned fibre network open to all parties and charged on a cost-plus basis. Peering would be managed by a carrier-neutral body.

The lack of peering in New Zealand since 2004, he said, was costing the country dearly. For example, TradeMe’s Internet charges had increased tenfold in 2005 when Telecom and TelstraClear de-peered, forcing much local traffic to move via Australia. At the same time as Drury’s paper, the ISPANZ called for an industry-based solution, and suggested government buyers of telecommunications services mandate their providers to become involved in peering. Drury said Radio New Zealand was providing streaming and hosting from both New Zealand and the United States. Even though 70 percent of the US traffic was bound back to New Zealand it was still cheaper than paying for domestic transit. The result: ‘an unnecessary import component to a strong content export story.’56

In a comment piece in Computerworld in March David Diprose kept pushing the debate along. If it didn’t make sense for courier packages destined for Wellington to travel via Auckland, Australia, or even the United States why would it be any different for Internet packets?

That’s what had been happening to the traffic of many of the countries’ ISPs since 2004 unless they paid TelstraClear and Telecom a premium to keep their traffic local. “Of course all this happens in the background so the customer may have no idea why their Internet experience is slower than it needs to be, and no idea they may be unnecessarily using international bandwidth, or that the Internet itself is less robust and lower capacity than it could be.”57

InternetNZ organised a team of external experts headed by independent IT consultant and ex-Telecom CIO Dr Murray Milner to look at issues of peering and local interconnection to help devise industry approaches and solutions. Peering was firmly put back on the agenda by the David Cunliffe at the New Zealand Friends of O’Reilly (Foo) Camp meeting.58 It was this meeting that was the catalyst for ISPANZ to take the issue further.

Telecom was now proposing a multi-points of interconnect project, which could at best be described as local peering, and consisted of the local handover of Internet traffic at network aggregation points across the country. ISPs would share the circuit cost with Telecom at the nearest local point of interconnect, and traffic would stay local. ISPANZ spokesperson Jamie Baddeley said the debate was far from over. “The proposal is not ideal, but it could be a lot worse … Telecom has come out and said they want a real consultative process on local interconnect … but they do need to listen to our concerns.”59

The Telecom proposal to set up 29 regional peering points ultimately got the thumbs-up, after being fleshed-out by Telecom’s wholesale division. Under the proposal, Telecom would peer or exchange traffic with all comers. Any ISP or content provider that saw a business case for peering with Telecom, and was willing to bear the cost and effort to do so, could interconnect with its network. The driver for local peering was mainly the improvement of performance for wholesale providers. Telecom wanted to avoid traffic tromboning over great distances.

Baddeley thought 29 points was a lot. “That means 120,000 possible customers approximately per point, which may not be the best number. Then again, having fewer points could see some smaller regional ISPs miss out on peering as they can’t reach Telecom easily.” He said Telecom’s willingness to meet at existing shared peering points was good news for providers, even though Telecom may not ‘peer’ at the actual Internet exchanges (IX), such as those in Auckland and Wellington. Of greater concern, said ISPANZ, was that only Xtra DSL customers would be reachable through the peering points. Telecom’s business customers were not covered and ISPANZ wanted that changed. “There must be reciprocity at all levels.”60

However peering did mean Internet traffic could be shared between service providers over higher bandwidths in local areas – up to a gigabit per second – assuming the ISPs had networks ‘up to the task,’ said Baddeley. A more efficient local network and vibrant Internet community could lead to export success for local content and software developers. “If we’ve got a good fabric for communication we develop interesting applications. If they’re interesting and worthwhile to New Zealand businesses then there is a good chance they could be of interest to the rest of the world,” he said.

Paul Hayes, Telecom Wholesale’s head of product management, said as well as better performance, local peering was part of a push to create a level playing field among New Zealand ISPs. Telecom would focus initially on peering Internet traffic from its retail brand Xtra with other ISPs. “Ultimately what we’re doing is laying down the kind of infrastructure that allows service providers to develop their services.” The 29 peering points tie in with Telecom’s plans for its NGN which will provide high-speed Internet and a VoIP phone service to replace the current phone system.61

PIONEER JOINS NEW TEAM CityLink had been a world pioneer in establishing open fibre networks, and independent community participation in civic affairs, as well as setting the pace for peering. When listed mobile radio company TeamTalk and a partnership of existing management acquired CityLink in November 2006 for $22.76 million, some thought the party might be over.

TeamTalk acquired a 76 percent shareholding in CityLink for $12.68 million and existing management acquired the remaining 38 percent balance of ownership. CityLink had been 98.7 percent owned by Lower Hutt businessman Ron Woodrow, who quit the company. TeamTalk had been looking around for several years for a suitable partner. Managing director David Ware said Citylink fulfilled that goal. “We were looking for a profitable company with a proven business model, a good management team, a New Zealand focus and a strategy for growth.” He said the acquisition would capitalise on a ‘mindshift’ among IT and communications managers as they looked beyond the major telcos for cheaper and more advanced services.62

TeamTalk would contribute capital and administrative expertise to help CityLink expand, possibly even setting up metropolitan networks in smaller cities. CityLink already had some coverage in Auckland and had recently expanded into Christchurch. Both companies had wi-fi networks and there were broad opportunities with the government making available funding to expand metropolitan networks and looking at last mile opportunities for both the GSN and the Karen research and education network.

A new company, provisionally called Neuco, would be established to explore the opportunities, headed by CityLink managing director Neil de Wit, with equity of $2.5 million and around $5 million from TeamTalk.63 CityLink’s revenue had grown around 20 percent a year for the past four years. “The ownership change will not lead any formula changes as long I’m at the helm, nor should it. Operationally there is no change. I now have a new board focused at delivering even more good stuff,” said de Wit.

“Telecom will bilaterally peer and use the current Internet exchange fabric where and when they choose. Additional Internet exchanges (IX) will slowly be rolled out by us as needed and local loop unbundling may drive that faster in the next two-three years.” He said Telecom’s motives were straightforward. “With LLU, their DSL network needs to have a sense of local otherwise Telecom will have to transport bits and bytes all over the country. Local peering is a must or they kill their own network capacity once ‘real’ bandwidth is in demand.” Currently the Telecom DSL network hauls all domestic and local traffic to Auckland and re-routes from there. He believed TelstraClear would eventually follow suit.

ISP EVOLUTION ESCALATES The pressure was on. Those who jumped in the deep end, following Xtra into the full speed Internet pool offering the highest speed line conditions allowed, had done so at a cost. Many ISPs claimed they were paying out several dollars a month for each customer but had little option if they wanted loyalty from a jittery client base closely monitoring market shifts.

Their hopes hung on imminent legislation, which they believed would provide a much better wholesaling deal. Otherwise they were up the creek without a paddle. The only way to make money was to offer voice, video content, mobile, and business grade services which would become hotly contested once the network opened up. Only those with deep pockets, strong branding, and enough value to add to their services would still be standing when the dust settled. Partnerships with outsourcing companies and equipment providers were essential. Certainly the winds of change had seen Ericsson, Alcatel, Nortel, Siemens and Cisco field increased inquiries for major carrier class and CPE.31

“Everything in the industry is for sale. It’s all up for grabs,” said Woosh chief financial officer, Gary Neil in a Computerworld article in July 2006. Telecom, TelstraClear, Vodafone, Ihug, Orcon, Slingshot, CallPlus, Maxnet, and Iconz and even the vendor and systems giants behind-the-scenes;Juniper, Cisco, Alcatel, Lucent, and Ericsson,32 were all up in the air.

INDUSTRY EXPOSED If ever there was a case for ISPs to become the fully fledged telcos they’d always wanted to be, this was the time. Changes to legislation, Telecom faced with a regulatory split, rumours about Ihug’s future ownership, ISPs being acquired, new DSL speeds and competitive pressure were delivering a challenge that could not be ignored.

Woosh had gone out on a limb when it moved to a wireless business model. Now it paid the price for the risk and was looking to bolster its viability by acquiring an ISP and Vodafone was dipping its toes into the world of ‘fixed-mobile convergence.’

The shake-out had actually begun earlier in 2006. Alliances were being formed, partnerships forged, acquisitions considered, and conspiracies hatched as major and minor players visualised the opportunities ahead in a new era of competitiveness. Smart ISPs were already putting NGNs and converged billing systems through their paces and eyeing ways to add value to their broadband offerings, including voice, business-level services, data warehousing, security, e-commerce solutions, and even IPTV and video-on-demand.

It was a time of tidying up loose ends. It was the end of an era for pioneering ISP Ihug, which announced in February 2005 that it had sold its broadband satellite service to Australian rural ISP Bordernet. The service began life as StarNet in October 1997 based on a microwave link from Auckland’s Sky Tower. It offered download speeds of up to 1Mbit/sec, well ahead of Telecom’s JetStream. The upload service ran over existing telephone lines but the company admitted it was never really a winning formula. The subsequent Ultra satellite service provided to subscribers beyond the central city and in rural and other regions, however, remained the only broadband option for many years.33

Vodafone and Telecom were both pushing 3G mobile phone services, which delivered higher speeds for data and even videoconferencing. While mobile was a major global trend the local cost was still outrageously high for Internet surfing. Wireless access, however, remained an attractive local loop alternative. In October 2005 South Auckland utility company Counties Power sold Wired Country, its wired and wireless provider, to Auckland ISP Compass Communications. Compass chief executive Karem Hussona said he would extend Wired Country coverage and the range of services offered. Counties Power also sold a of its 3.5GHz band radio spectrum management rights to Telecom.34 This was the first indication Telecom was seriously looking at moving into wireless broadband. It had acquired spectrum before but never used it and its only foray into the field so far had been through its partnership with BCL.

Then Auckland wireless voice and data provider Woosh made its move, acquiring Auckland ISP Quicksilver, which claimed around 10,000 customers. Quicksilver knew it needed a cash injection, a bigger partner or both, if it was to continue to be successful in the emerging environment. Woosh had been in the game since 2001 with a fixed wire replacement solution. Now it had mobility options and was offering voice. It had been struggling to make headway; after three years and $130 million invested it had only 15,000 customers and had even begun selling dial-up accounts to supplement its wireless service. It was also moving to a WiMax-based platform but some of the spectrum it hoped to use had been clawed back by the government.35 Its speeds had increased from 1.6Mbit/sec to 3Mbit/sec and another technology upgrade promised 14Mbit/sec speeds by the end of 2007. With Quicksilver on board it could now play in all the patches. The company had an aggressive expansion plan and was keen to place copper access equipment at Telecom exchanges.

DEFER, DENY, DELAY CallPlus was sufficiently capitalised for the near future, planning to invest up to $200 million over five years on building out its WiMax network. Ideally though, its roll-out of nationwide phone and Internet services with IPTV as a likely third component would involve a mix with DSL. “It is quite appealing to go with WiMax and get out of Telecom’s face with genuine alternative speed and reliability in the local loop, totally bypassing all the negotiating and complaining and permissions,” said CallPlus director Malcolm Dick. Like everyone else though, he was waiting to see what kind of pricing was involved. “I would hope Telecom would come to us and discuss ways of doing a deal on the copper. Why wouldn’t they try and entice us to use the copper loop? At least they get some business from us this way but with WiMax they get zero.”

He suggested Telecom should have a management change, mainly because it was not well trained or experienced in doing ‘win win’ deals. “They’ve spent 15 years playing an amusement park game where they run around waiting for the little rat to pop up and then try and hit it with the hammer. It must be really hard having to turn around and start feeding them instead.” Dick remained cautious, recalling his experience trying to interconnect a competing telco in Australia where Telstra publicly described his operation as ‘a germ invading the nooks and crannies of their tariff.’ He said incumbents typically used the three Ds: defer, deny, delay.

Ihug CEO Mark Rushworth was prepared to spend $200 million rolling out an alternative network that he could wholesale to other smaller ISPs. Like Orcon, he was keen on port sharing to quickly get the widest coverage. Rushworth, however, believed only two or three players could afford to create their own networks. He said the credit swapping model became the de facto way of doing things in Australia when Telstra refused to allow more than one provider into its exchanges at a time. “Optus would go into one exchange and someone else would put five cabinets in another and sell on that capacity. It certainly forced the industry to work together.” He was hopeful CallPlus, TelstraClear, and Ihug could operate similarly. “It makes a the roll-out a lot quicker – we’re certainly keen to work that way so we don’t end up jumping on each other’s toes.”36

Ihug was leaving Australia to its new Perth-based owners iiNet, “overhauling its product line, re-energising the brand and reducing operating costs.” As New Zealand’s largest wholesaler of DSL broadband, it had announced its intention to deploy its own network once access to Telecom’s exchanges was permitted. In fact the country’s third largest ISP, with around 120,000 customers, was ripe for the picking with iiNet suffering some setbacks in profitability. It was placed only behind top dogs Xtra with 500,000 customers and TelstraClear with around 200,000.37

Orcon was even cheeky enough to send out a press release saying it would love to buy the company. An hour later another press release qualified that it ‘wasn’t actually in discussions at this time.’ Then on 9 October 2006 Vodafone stepped up and offered its hand. Vodafone ad battled from way behind in the mobile phone business to run a nose to nose race with Telecom, at times edging way ahead. Now it needed to add value in the converging marketplace and saw itself as the ideal suitor. Vodafone paid iiNet NZ$41 million for the ISP with CEO Russell Stanners saying the deal further strengthened its position in the market. “It’s a perfect fit. Right now, we are the leaders in mobile, however we only have 20 percent share of the telecommunications market. When combined with Ihug’s strength in fixed line broadband and calling, we can develop and deliver even more compelling propositions for our customers.” Vodafone would remain a mobile-centric business and Ihug would continue to operate as a standalone company.

In February 2006 Orcon had taken on board Juniper’s platform to meet increasing demand for scaleable high speed DSL software-based services giving it application-specific and per-customer QoS capabilities, including the ability to deliver multimedia broadband services. Then in August it announced it had also teamed up with hardware provider Siemens in a $30-million five-year plan to go straight to high speed VDSL2 for IPTV and related services. It planned to deliver voice services and IPTV with more than 50 channels, which it claimed it would be able to launch in late 2007.

Meanwhile Maxnet, Compass, Iconz and other second-tier ISPs were biting at the bit to show what could be done with better access to bandwidth. Each had their own specialty to differentiate from mere bandwidth resellers, whether it was e-commerce hosting, secure back-up, disaster recovery, software development, integration, or other ICT skills for the modern business. Alternative bandwidth wholesalers were also lifting their game to be ready for the new environment.

The Consumers’ Institute’s 2006 Internet survey which collated the views of 10,000 Internet users showed increasing dissatisfaction with Internet services. Only 66 percent said they were happy with their ISP, down from 82 percent in 2005. Xtra had come in for the strongest criticism – only 55 percent of its customers were satisfied, down from 78 percent in 2005. In particular its Go Large plan, which offered unconstrained download speeds and removed the monthly data cap for a flat-rate fee of $50 a month for Telecom phone customers, had been criticised for patchy download speeds.

The result was a high defection rate away from Telecom to smaller ISPs with a stronger focus on customer service and more flexible plans. The competition was fierce, with Inspire, Actrix and Xnet rated the best in the most recent Consumers’ Institute survey and Ihug, TelstraClear, Orcon and Slingshot in particular trying to win Xtra customers by waiving sign-up fees and offering free modems and discounts.38