Off the shelf v custom software

Telcos are under tremendous pressure to reduce costs, while at the same time improving the experience of customers. Trying to achieve this requires a thorough review of supporting IT systems’ operational support systems/business support systems (OSS/BSS) and a drive to automate service-delivery processes.

This is where the modern telco’s competitive advantage lies. It certainly doesn’t only lie in network infrastructure: Nowadays, that is pretty much a commodity. Everyone has similar networks, built on the same equipment from the same suppliers and to the same standards.

The network is just the first element in a telco’s service-delivery capability. The second, critical element is the OSS/BSS.

In fixed-line operators, OSS/BSS have traditionally evolved over many years from fragments supporting different networks and services. The fragments have been “integrated” — or rather, bolted together with whatever is available — getting ever more unwieldy and opaque as exceptions and changes accumulate. The end result is often a dysfunctional “system of systems” with limited architectural coherence. In this situation, trying to reduce costs and improve service at the same time is like running a 100m sprit with your shoelaces tied.

The review of IT systems inevitability leads to a debate about whether to buy commercial off-the-shelf software or to build custom systems. I would argue, however, that if telcos want to make the right investment decisions today, that will provide meaningful software platforms for the next 15 years, they need to take a different view — one that transcends the traditional dichotomy.

The shift towards off-the-shelf software is partly a response to an out-dated custom-build approach, which saw systems growing more or less at random. The major arguments for buying off-the-shelf software are strategic. In the first place, developing software systems may not be a key competency that a telco wants to develop. Secondly, off-the-shelf systems embody the best-practice lessons of a large pool of customers and economies of scale can be leveraged. In addition, vendors consciously and carefully manage the product cycle to ensure compatibility with fundamental changes in both the IT and the telecoms industry — covering data, services and content domains. If your software forces you to adopt new business processes, that may be a trade-off that is worthwhile in the long run from a business perspective.

With one primary delivery process across mobile, IP and traditional circuit technologies and services, it is inevitable that customisations or additional components will be required to accommodate specific requirements that the one standardised process cannot offer. So for telcos, installing pre-packaged software can mean surrendering one of their sources of competitive advantage.

This is often the first step in the argument for custom-built software: “If you’re going to customise off-the-shelf software it’s going to be costly. You might as well build your own system that will be better able to support your business.”

And so the argument goes back and forth. But in reality, neither of these paths offers more than a temporary illusion of safety for those who believe that “nobody ever got fired for going with Gartner’s Magic Quadrant”.

What these debates miss is that the world has changed: open standards, architectural frameworks and business process management/service-oriented architecture platforms from a single vendor mean that it’s now possible to have the best of both options.

Telco system architecture and design standardisation championed by the global TeleManagement Forum, the Open Group etc mean that there is now a best-practice approach to building telco systems that has been co-created by the best minds in the industry. This approach is based on standardised processes and interfaces that are common to all telcos, but allow each operator to customise these processes in their own unique way — without losing the architectural coherence that is needed to carry the systems into the future.

Some would quite correctly caution that there is still a gap between this theoretical paradise and the real world. There is no silver bullet but at least there is a target architecture that has been developed for the industry, by the industry.

The future is about “off-the-shelf” components that can be pulled together on single, integrated OEM supported platforms so that very complex products can be constructed in relatively short time scales by relatively small teams of people. These services may well co-exist with and complement traditional off the shelf implementations. But one thing is for sure: The days of the rough-and-ready “make a plan” approach to custom development are over.