Case Study on why Ripping out the old IN to put in an open source NG SDP makes business sense

Even though M1 is a small mobile operator, roughly 1M customers, it remains one of the most innovative mobile operators.  M1’s business strategy is simple: to constantly deliver value to its customers by rolling out new and innovative applications and services.  In mid-2008, M1 began its search for a next generation service delivery platform (NG SDP) to provide a framework for traditional telecommunication services, next generation of value-added services, and enabling it to remove its old IN that limited its ability to innovate.

M1 selected an open source JAIN SLEE (Java™ APIs for Intelligent Networks Service Logic Execution Environment) solution provided by its long-time IT partner hSenid, called the mChoice SDP.  M1’s relationship with hSenid goes back to 2002. For nearly eight years, hSenid has helped M1 implement major projects such as the deployment of MySQL cluster, mChoice Rewards, mChoice Recharge, and mChoice Reporting – a comprehensive business intelligence system to help M1 to make strategic decisions and revenue calculations. These solutions were built on open source, resulting in significant cost savings in terms of licensing and maintenance fees.

The standard application programming interfaces (APIs) help M1 provision, control and bill for all the value-added services they provide, whether the services are developed in-house or created by third-party application.  This is a key point the APIs are not just for third parties, they’re for internal consumption as well.  M1 set out with the objective of looking for a SOA (Service Oriented Architecture) based solution, given their enterprise architecture.  I discussed SOA in this article.

For M1 its main benefits in adopting a NG SDP are:

  • Opens up service innovation, letting third parties offer services to M1’s customers or using M1’s network capabilities to their customers, which opens up 1000s of new applications and services;
  • Launch new services faster, moving from months to days in launching new services (factor of 10 improvement);
  • Protects existing investments while enabling future growth, i.e. reusing amortized equipment, e.g. SMSCs, while putting growth onto lower cost platforms (a 10/100 factor cost reduction for growth); and
  • Lower operational overhead by simplify on-boarding, contracts, etc. enabling M1’s limited people resources to launch 100s more services each year.

hSenid’s whitepaper provides more details on the M1 case study.

There’s an interesting discussion the industry needs to have on whether a SOA or a looser web-centric integration framework is the right long-term approach.  For smaller operators this distinction is mute, but the larger the operator, the closer its cost-basis need to tend towards Google – as in the limit they’re both service providers.