SofNet Day Two. A Jungle of Capabilities and a Services Desert

Summarizing day two of SofNet.

Panel session, “How Can 21CN Help Make Customers Successful?”  was a who’s who on UK telecom executives with John Cunliffe, Ericsson; John Frieslaar, Huawei; Geoff Hall, Nortel; John-Paul Hemingway, Ciena; Phil Dance, BT; David Soldani, NSN; Andy Stevenson, Fujitsu; and Phil Tilley, ALU.  With so many suppliers in such a public forum, we were not going to hear anything interesting.  A good question was “What are the top 3 enterprise services enabled by 21CN.”  The answer was a general description of network IT Services and cloud-computing; which are available without 21CN.  This unfortunately became a theme for day 2 of the event, lots of discussions on architectures, potentials and capabilities, but a dearth of discussion on what the innovators (web2.0 companies) want and what are the specific services.


  • Helmut Leopold, Telekom Austria, discussed their SDP experiences. Good review on the challenges facing a small operator, especially on processes which are “90% of the headache” in any network transformation or new service launch.  As discussed in yesterday’s weblog entry operational efficiency is the best understood part of the SDP business case: Helmut mentioned new product launch effort reduction of 70%, and TTL (Time To Launch) reduced from 9 to 5 months. An important point he raised was the SDP enabled greater transparency to understanding profit management, however, a critical issue is the BOSS (Business and Operational Support Systems) organization tends to mask such data to maintain control, hence he recommended the need for “aggressive centralization.”
  • Sally David, BT.  Discussed next generation wholesale business models.  Covered a number of examples of wholesale business models such as outsourcing voice (to Virgin media), managed access (mobile backhaul), QoS (Quality of Service), and security.  Good focus on the capabilities operators can offer to corporate customers and service providers.

Panel Session: Leveraging SOA to Achieve Flowthrough for IP-Services

  • Brian Buggy, Amdocs.  SOA is just a tool, need to define the model that needs to be common across all services whether mobile, cable, broadband or internet.  Upon receipt of an order it’s then decomposed, designed, implemented and tested.  Critical step in the decomposition process is the mapping of customer facing services to resource facing services.  Highlighted the gap that exists in the TMF SID (Tele Management Forum Shared Information/Data) on the level of decomposition to enable re-use of resource components across services, else operators end up rebuilding silos within a SOA framework.
  • Francesco Caruso, Telcordia.  Described how the core capabilities of SOA: semantic (common language), transport (common orchestration) and legacy wrapper can be applied across the product lifecycle (CRM, order management, charging/billing, provisioning/activation and service assurance).  This reminded me of the the approach taken by AT&T U-verse, which is a great case study on using SOA within a mixed (NG and legacy) environment.
  • Biru Chattopadhayay, Tech Mahindra.  SOA is a little like SDP in definition, it depends upon the user.  Challenge that there is no agreement on what comprises an SOA-compliant implementation, so SOA is not generating harmonization given the variety of implementations, platforms claim SOA compliance, when in practice significant integration is still required.  SOA is not a magic bullet, still requires good design and central management.
  • Bob Titus, Netcracker.  Today results of SOA implementations on cost saving, flexibility, improved TTL have not met those seen in other industries.  Bob highlighted the model-driven architecture gap, as per Brian Buggy’s presentation.
  • My take-away:  What should by now be common system across most operators remain bespoke.  I’d argue the problem is not technology, as businesses in other industries have achieved significant commonality in their systems thanks to SOA.  Rather the operator’s BOSS organization guarding silos of information to maintain their organization.

Panel Session: The Reality of Convergence: Operator Experiences

  • Thibaut Roussel, Orange.  Summarized Orange NeXT (New Experience of Telecom) strategy and localization issues across its OpCos (Operating Companies).  Unik (Orange’s FMC service, c.f. BT Fusion): 500k customers in Feb ’08, 25% of customers use at home and in the office (SME is a target segment), main selling point for customers is unlimited calling  that has resulted in an additional 67 call minutes per month.  Business Everywhere (remote intranet access): 910k users (600k in France) as of Feb ’08, 24% annual growth, targeting MNC and plans to expand to SMB and SoHo segments.
  • Paulo Tavazzani, Fastweb is an all-IP triple-play network based in Italy.  Deployed an integrated CPE (Customer Premise Equipment) for converged voice, video, and data for all devices, including mobile, in the home.  No plans for IMS, soft-switch is good enough, will only implement IMS when business case justifies.  Fastweb are the only convergent MVNO in Italy, hence focus on converged services to differentiate from other MVNOs and try to keep calls on their network as much as possible to manage costs.
  • Take-away: Good examples of service innovation that focuses upon the importance of providing significant customer value to stimulate adoption, even though the technologies are far from mature.

Overall Summary

  • The telecom industry is unfortunately still too ‘fat and happy.’  For example Verizon announced first quarter results yesterday where net income rose 10% to $1.64 billion from $1.5 billion, and sales rose 5.5 percent to $23.8 billion.  Isn’t the US in a recession?
  • Most of the discussion at the conference is on capabilities and potentials; hence the SofNet ideals remain trapped between the CTO (who sees the strategic threat) and the CMO (who says, “Show me the money!”).
  • Recommend each operator identify for their specific situation 5 or 6 services they can easily enable and just do it to work through organizational issues in opening the network, because its still another 2 or 3 years before they’ll have openness cracked and the strategic threat will have materialized into their financial results by then.