Effects of Reliability Measures on Market Share

Although unavailability and failure frequency are both important reliability measures for designing telecommunications networks, having two different reliability measures causes difficulties when their evaluation results contradict each other. To overcome these difficulties, I propose an approach to determining a single reliability measure to be used. I focus on market share, which is a very important management index, and study the effect of each reliability measure on this index. If we find that the effect of either reliability measure is small enough to be ignored, then we choose the other one as the single reliability measure. I give a statistical analysis procedure for this study and demonstrate test results that show failure frequency to be promising as the single reliability measure.