itw21i

Michael Sainsbury

Telstra's back end blues

20 April 2000

Australian Information Week

Submitted for Best investigative category

Telstra is finding itself without the technology to backup its flash dot com deal maker status, because its multi-billion dollar DMO (Data Mode of Operation) network project sits stalled on the start line.

The company recently pulled off its most risky, yet potentially rewarding, deal -- a complex $5 billion plus pan-Asian data and mobile partnership -- with the prospective new owners of Hong Kong Telecom, Richard Li's Pacific Century CyberWorks.

The deal underscored the style of Ted Pretty and his team - new economy, quick, and (apparently) clean.

But the excitement on the faces of Telstra brass at the announcement of the carrier's surprise deal, is at odds with the progress of its landmark project to create a cost efficient, high speed Internet-based network.

Launched on July 1 last year, the DMO project was to create a new national TCP/IP (Internet-based) network that would cost the company more than $5 billion over the coming years.

But DMO's roll out has stalled and the carrier's tardiness in delivering on its much-vaunted "next generation" network is already hurting the bottom line. Telstra recently lost around $1 billion worth of key corporate accounts (the Commonwealth Bank and BHP) because it lacked next generation network products, such as Internet-based virtual private networks.

Last July Telstra CEO Ziggy Switkowski said: "Our Data Mode of Operation is a blueprint for a radically different marketplace based on data and the Internet. We have sought to optimise our choice of technologies to provide the best service at the lowest cost. Telstra is driving Australia's technology future."

Since then, Telstra's still heaving bureaucracy and a growing gulf between its engineering and product divisions, has many in the company worried whether it can move its network development along fast enough to stop revenues sliding further.

While Pretty and his team sit at the glory end of the business, delivering on their promises relies on a solid back-end of products and services. A network must be IP ready, flexible and able to offer cheap, slick online tools to the ventures that will carry those dreams to market. Telstra also needs a network to connect the companies through which the carrier and its partners will try and flog products and services.

Slow Gestation
The DMO promised all this, but nine months into implementation, the first phase of the project is still nowhere near completion. Is this only the tip of the iceberg of problems inside DMO?

Telstra is still undecided on the final supplier for two further tenders on the next two phases of the project - service management and the all-important next generation telephony network.

It appears Telstra is not working at Internet speed, but at the public service pace that harks back to the cosy, government monopoly days of the old Telecom Australia.

When quizzed on the timing for the first phase roll out, DMO project manager Greg Fidler, in classic public service style, could say only: "there will be outcomes this calendar year."

Telecommunications analyst Paul Budde believes Telstra has failed to adequately upgrade its network despite costly plans within the DMO initiative.
"A lot of focus within Telstra has been diverted to Ted Pretty's dot com group," said Budde. "But Telstra's core business is the network, not dot com. By taking their eye off the core network business, they are now in a position where they have fallen behind. That's given smaller telcos an inroad in to markets that Telstra is failing to service, particularly in rural areas.
"Falling behind in this way is costing Telstra plenty in potential income. The right broadband network has potential earnings of $635 million from residential and business users in Australia," he said.

Despite Gerry Moriarty, Telstra's network technology group general manager telling InformationWeek "a couple of major milestone's with Nortel's implementation were passed recently," it is understood Nortel Networks continues to struggle with the implementation of Telstra's core IP network.

Nortel itself was very coy when questioned on the project's progress - deferring to its paymaster, Telstra.

And while Fidler, Nortel's Telstra account manager Kevin Deardsley and DMO supremo Moriarty all swear the project is on track there is scant physical evidence of anything having changed.

"Yes we are on track," Moriarty said. The real question is whether that track is in tune with the rest of the organisation.

Telegraph Speed
Sources close to the company told InformationWeek the Convergence Group (which now includes the old Business unit) is getting increasingly frustrated with the measured, 'old world' pace of its engineering colleagues. Product managers must try to pass off switched data products to customers, while Telstra's rivals quickly bring an array of 'new world', IP-based products to market.

The DMO had an old economy genesis. It was first mooted in April 1998, by Telstra Research Laboratories boss Hugh Bradlow and network chief Moriarty. Some Telstra insiders suggest DMO was originally cooked up as the platform for Moriarty's (unsuccessful) run at the Telstra CEO slot.

It was a slow start, the project was not put to tender until October 1998, closing in December 1998. But it then took Telstra six months to make the final decision. Sources close to the big equipment vendors vying for the contracts said Telstra played a classic tender game - moving from player to player to drive down the price. As a result, price appears to have foolishly ridden roughshod over technical and implementation concerns.

The vendors were anxious to play along, having invested heavily in their Australian operations in the hope of latching on to the fat revenue streams flowing from Telstra's $4 billion a year capital works budget. To lose access to that money could cripple the local unit of an Alcatel or Ericsson.

When the decision on DMO phase one finally came there was widespread surprise that the carrier went with not one but four different equipment vendors, making a difficult engineering task that much more difficult.

True to form, Telstra went for the comfort zone. Trusted suppliers Nortel and Alcatel grabbed the cream of the contract. Nortel for the core network, including the Internet Protocol (IP) network, Frame Relay, the ATM data network, plus the Telstra Access Server technologies and the project's systems integration.

Alcatel got the next generation Network Management gig. Lucent Technologies was thrown the bone of the IP dial-up gateway, but this was probably because Telstra already had gear in place from Ascend, a company purchased by Lucent in 1999.

When You're Ready
At the time of the first DMO tender announcement, Moriarty said "the technology Telstra selected will not only meet the high growth in data traffic on the network from existing products now, it will also provide new capabilities to support products that are on the horizon or not even dreamt of yet."

However, when any of those products will see the light of day is anyone's guess.

Alcatel network appliances GM David Stephenson said Alcatel had begun to put the network management systems in place. This will take place in four stages over the year
"These are timed to match other parts of the DMO project over the year," he said. "At each stage we need to integrate with each specific version of specific boxes that are installed into the network."

This adds more fuel to the talk that Nortel is struggling with the network core, on which the rest of the project relies. "The project definitely has an ongoing element," said Stephenson.

But while technology implementation difficulties and delays are expected in these times of breakneck technical innovation, the extraordinary delays in Telstra signing the DMO contract and mulling over the tenders come from another era.

One of the key planks in the Telstra/PCCW deal is the development of business services. These are mooted to be in the areas of network facilities management, data hosting, enterprise solutions, application service provider (ASP) initiatives and (flavour of the month) business-to-business e-commerce.

These services and products depend heavily on the implementation of what Moriarty describes as the third phase of DMO. "This is the contract for service management, the acquisition of a full suite of new generation IT systems. We are talking about the most advanced IT concepts here," explains Moriarty.

He confirmed there is a short list of two companies (understood to be KMPG and Anderson Consulting) and while he insists a decision will be made "within a couple of weeks", the call for tenders was made around October last year.

Six months (or more) is hardly Internet speed. And once a supplier is settled, Telstra has to finalise the contract - a process that took 3-6 months for the first phase project. Yet Moriarty is certain "we will have products in the marketplace by the fourth quarter."

The fourth, and biggest, phase of the network project is for next generation telephony systems. The short list of three is said to include Ericsson, Nortel and Alcatel. The winner will convert the vast remainder of Telstra's network to IP, a project technical guru Hugh Bradlow says could take up to 10 years. "It will happen gradually but there will be an inflection point where it is too expensive to maintain the legacy system," Telstra's own Bradlow said.

Trendy Tactics
There are some encouraging signs the carrier is moving towards more modern contracting models. The smart, new European telco practice in vendor selection is to move quickly to parallel trials with two suppliers, get the technology working, then made a speedy choice on a partner.

Telstra appear to be mirroring this in its plans to change the GSM mobile network to a packet-based system using general radio packet system (GPRS) technology.

Following some initial testing, Ericsson and Nortel will be given "a city each" to prove their technology. Moriarty said Telstra had set very tight targets for the mobile network upgrade and hoped to have commercial services available by June/July.

But there is no sign that Telstra is ready to accelerate its DMO project using similar techniques.

The company will be under pressure to perform in its new PCCW partnership. At just 36, PPCW boss Richard Li is a young businessman in a hurry and he may apply serious pressure to crank up the speed of Telstra's local network modification a notch or three.

As Switkowski stated at the announcement of the PPCW deal: "We sought to partner with a company that had new economy skills and style and who shared our view that speed to market and commitment to growth were paramount."

Regardless of the words that speed to market cannot happen without a network that can handle it.

Michael Sainsbury

Position

Orgainisation

Phone