Bad governance resulted in flawed network built by inexperienced company
Telecom chief executive Paul Reynolds, who has had a terrible start to 2010, must have choked on his toast when he opened Wednesday's newspapers.
The front page of the Herald had three photographs of him under the heading: "Cellphone shambles: the bloodletting begins." Page 2 was totally devoted to the "XT Failure" and "XT woes start to hurt shares" was the heading on the front page of the Business Herald.
The front page of the Dominion Post also had a large photograph of Reynolds surrounded by the headlines: "What Went Wrong"; "Experts fly in to try to fix our phones" and "Readers Say: The System Stinks".
Negative front-page headlines are a chief executive's worst nightmare and Reynolds would not have contemplated this development when he arrived in New Zealand just over two years ago.
His first letter to shareholders, which was included in the company's 2007 annual report and accompanied by a positive photo of him, said "Telecom is ready for a strong future", partly because of "the construction of a new mobile network".
So what went wrong? Why did a major project turn into such a disaster for the company?
Mobile networks are a complex subject but in simple terms there are two systems, CDMA and GSM. This is somewhat similar to original video recorders, which had Beta and VHS systems.
In 2001, Telecom adopted the CDMA system, which was widely used by the US military and was the dominant system in the United States.
Telecom's choice was probably influenced by Verizon, formerly Bell Atlantic, which didn't sell its 20.3 per cent stake in the New Zealand telco until September 2002.
CDMA was the wrong choice because it is not compatible with GSM, which is the dominant system outside the United States. Telecom's CDMA system could not offer roaming throughout most of the world and mobile phone manufacturers were far more interested in developing new products for GSM users.
The situation was exacerbated when Telstra decided to close down its CDMA network, which occurred at the end of 2008. This meant Telecom could no longer offer roaming in Australia.
On June 8, 2007, Telecom announced it would build a new $300 million mobile network. The new WCDMA network was based on 3G technology, a combination of both GSM and CDMA, and is compatible with both systems.
On October 15, 2008, just over 12 months after Reynolds joined Telecom, the company announced it would upgrade the new network announced in June 2007.
The latest proposal was a nationwide rollout of WCDMA technology, using a frequency of 850MHz. This 3G mobile phone system, called W-850, would cost an estimated $574 million and would be compatible with both GSM and CDMA. The new service was expected to be launched in June 2009.
Reynolds was quoted as saying: "Telecom's new mobile network will be superior on every level: the best nationwide coverage, the fastest internet on your mobile, a wide range of world-leading handsets, as well as better content, music and business applications."
Telecom was determined to replicate Telstra's highly successful W-850 network, which was launched in October 2006 after an 11-month build by Ericsson.
A number of industry experts were surprised Alcatel Lucent - which was formed in December 2006 through the merger of France-based Alcatel and the US Lucent Technologies - was chosen to build the network.
The Paris-based company is a fixed-line expert with limited mobile network experience, particularly as far as 850MHz is concerned.
Geoff Zame, of Craig Investment Partners, wrote in October 2008 that Ericsson, Nokia Siemens and Huawei were the mobile network experts and Alcatel Lucent had limited experience in this area.
He believed the latter's earlier work on Telecom's mobile network, announced in June 2007, had "been problematic and plagued by delays".
Zame concluded: "The success of Telstra's W-850 deployment with Ericsson demonstrates the importance of vendor selection in limiting execution risk and maintaining an ongoing competitive advantage in the marketplace. We remain uncertain about the merits of Telecom's reliance on (principally) one vendor and we believe Alcatel Lucent's market position in W-850 increases execution risk." Telecom had a long-term relationship with Alcatel Lucent; this included the appointment of Ron Spithill, an Alcatel employee for 43 years, to its board in November 2006.
W-850, now called the XT Network, was launched in a blaze of publicity in May 2009. Telecom was extremely optimistic about the new mobile system, calling it a superior network in terms of service, coverage, roaming and innovation.
However, the original plan indicated $173 million, or 30 per cent, of the total capital expenditure would be incurred in the financial year beginning July 2009 and according to the media launch there would be additional "fibre roll out during the June 2010 year".
In other words, Telecom was aggressively launching the XT Network before it was fully completed. The marketing campaign has been far more successful than the reliability of the network as 467,000 customers had connected to XT by the end of December 2009, just seven months after its launch. Most of these transferred from the company's old CDMA network, which remains in operation.
The problem with the XT Network is an issue of poor corporate governance, at the board and management level.
Telecom has been far too close to Alcatel Lucent. Ericsson, Nokia Siemens or Huawei, all of which have considerable experience building 3G mobile networks, would have been less risky appointments.
It is also not best practice to appoint a former employee of one of the company's long-term providers to the board, because that individual will have the most board expertise as far as network construction is concerned and is unlikely to criticise his former employer.
The other issue is that XT Network was probably launched too early, before it was fully tested and completed. One of its biggest flaws appears to be the absence of a reliable backup in the lower part of the South Island.
Reynolds has had a stronger marketing focus and this is reflected in his results presentations to investment managers and analysts. He marches back and forth across a stage talking about excellent customer service and increased customer satisfaction, whereas Theresa Gattung, the previous chief executive, was more inclined to talk in technical telco terms.
This emphasis on marketing is timely but not if it has occurred at the expense of network reliability. New technology is notoriously high-risk, particularly if it has been built by a company with limited experience. Telecom would have been better advised to wait until the network was completed and tested before it had such an aggressive launch.
Ironically, Telecom's high-powered advertising campaign was probably too successful because it attracted 467,000 customers in the first seven months and these high numbers must have been a contributor to the outages.
Telecom has appointed an independent party to investigate the debacle but shareholders should be demanding answers to a number of basic questions, including: did the company have an open and contestable tender process for the construction of the XT Network?
If so, why was Alcatel Lucent appointed? Was the appointment based on price or some other issue?
Did Telecom comprehensively test the XT Network before it was launched?
Why was the XT Network aggressively marketed before it was completed?
Why is there no reliable backup when the system fails?
Telecom is now involved in a massive public relations campaign to restore its reputation and maintain its mobile customer base. This is important, as well as who is going to be held accountable for the debacle - particularly as Telstra's W-850 system, which was built by Ericsson, has been highly successful, whereas Telecom's W-850, built by the relatively inexperienced Alcatel Lucent, stumbles from crisis to crisis.
* Disclosure of interest; Brian Gaynor is an executive director of Milford Asset Management.
