BT: Separation is an IT headache

British Telecom may be the most talked-about case study when it comes to structural separation of a telco, but our view may be more colourful than real life experience, a telecoms forum heard today.

“It always amuses me to hear different perspectives on the UK experience,” said Mark Shurmer, regulatory director at Openreach BT, the functionally-separate wholesale part of BT.

“I have to say it’s a very different world that I’ve heard described today that the one I actually live in.”

Shurmer spent a significant portion of his presentation to the Telecommunications Regulatory Reform Forum in Sydney setting the record straight on BT’s experience of structural separation.

He cautioned delegates to focus at the benefits achieved over the whole separation period, not just “selected points in time.”

“We’ve heard a lot about the benefits and costs of functional separation,” Shurmer said in what appeared to be a reference to Telstra’s presentation from earlier in the day.

He believed separation had resulted in a more effective wholesale regulatory regime, greater transparency for customers - retail and wholesale, and more competitive downstream markets.

Shurmer also claimed annual fault rates were down 15 percent per year and repair lead times were 70 percent better than before BT Openreach was established.

But Telstra could find a voice of support in Shurmer confirming that “achieving equivalence was more costly and complex” that BT had expected.

“This really surprised us,” he said.

He appeared to caution against requiring a structural separation of IT systems under any separation legislation. Telstra said earlier today such a requirement would force it to break five years of transformation projects at a cost of between $500 million and $1 billion or more.

“Systems separation is very costly and complex, particularly if it’s full separation,” Shurmer said.

“You have to break the system - as distinct from logical separation where you build gateways and barriers [in them].

“I’d argue you get most of the benefits through logical separation and that getting full separation in legacy systems just isn’t worth it. For legacy systems, you have to temper [separation] with pragmatism.”