Obtaining Voice-Over-IP Services: A Comparison Between Selection Processes Amongst Single Vendor Options And Multi-Vendor Options

The quality of service (QoS), availability and cost for availing VoIP services have become critical operational performance factors of many global organizations especially in the context of process outsourcing. Consequently, a number of business models and delivery models are being tried and tested across the industry sectors to get the best possible fit between the clients´ demands for quality, cost and bandwidth from the VoIP providers and the supply-side of VoIP, i.e., what the service providers promise to give in the contracts or Service Level Agreements and what they are actually able to provide. The latest trend in the outsourcing industry shows that clients are looking for more multi-vendor options for procuring any service than a single vendor option. This model reduces risk and increases the negotiation advantages for the clients. Considering this phenomenon, in this paper, we have proposed two different ways to select vendors for VoIP: one in a single-vendor option, another in a multi-vendor option. In the single-vendor option we have used data envelopment analysis. Data envelopment analysis (DEA) is not a new technique in itself at all, in fact, in this paper we have shown the various ways it has been applied in other domains including service domains like financial services and banking etc. The main strength of DEA in the context of this paper is its ability to generate an efficiency frontier, which can be used for selecting a VoIP provider or comparatively evaluating VoIP service providers´ performance data. We have explained this approach using a simple example, and outlined the implementation possibilities with more complex variables and models. For the multi-vendor option, we have shown an approach using goal programming. Using the same constraint sets as in the DEA formulation, the only difference that we have assumed here is that the client can distribute its VoIP service requirements to more than one vendor in contrast to a single- vendor situat- - ion in the DEA example. Goal programming is then used to select the optimal distribution of VoIP service requirements to a number of providers.