Optimal Strategy for Proactive Service Delivery Management Using Inter-KPI Influence Relationships

Abstract

Service interactions now account for major source of revenue and employment in many modern economies, and yet service operations management remains extremely complex. To lower risks, every Service Delivery (SD) environment needs to define its own key performance indicators (KPIs) to evaluate the present state of operations and its business outcomes. Due to the over-use of performance measurement systems, a large number of KPIs have been defined, but their influence on each other is unknown. It is thus important to adopt datadriven approaches to demystify the service delivery KPIs interrelationships and establish the critical ones that have a stronger influence on the business outcomes. Given a set of operational KPIs and SD outcomes, we focus on the problem of (a) extracting interrelationships and impact delays among KPIs and outcomes, and building a regression-based KPI influence model to estimate the SD outcomes as functions of KPIs. (b) Based on the model we propose a schedule of action plans to transform the current service delivery system state. (c) We also build a visualization tool that enables validation of extracted KPIs influence model, and perform what-if analysis.

Empowering global software development with business intelligence

“…By modeling the relationships between indicators we provide awareness about the impact of each decision made. This step should be performed in collaboration with experts, or using mining techniques [40,41], as the relationships between indicators may need to be estimated.…”

“…The influence map is represented by means of a new chart called Influence heat-matrix, let managers acknowledge the existing influences and navigate between the KPI details view. For showing the relationship in a chart we join a co-occurrence chart and a heatmap providing a custom visualization, similar to [41]. This view helps managers to assess the impact of their decisions on other KPIs.…”

Improvements Towards the Identification and Quantification of Relationships Between Key Performance Indicators

“…Each KPI can have multiple attributes of direction, strength, and polarity. For example, improving some KPIs might worsen others, and the impact of a KPI might be delayed and only become apparent at a later stage [1,2]. To address these challenges, the importance of quantifying inter-relationships between KPIs has been studied.…”

“…Rodríguez, Saiz and Bas [3] presented initial research on the topic, and later followed it up with their quantitative relationships at the performance measurement system (QRPMS) methodology [4]. More recent work explores KPI inter-relationships in service delivery [1,2], the manufacturing industry [5], and cellular network systems [6]. With the ease of data collection brought about by technological developments and the significant contribution that inter-KPI relationships may have on organisational decision-making, it is evident that methodologies such as QRPMS are important in managing organisational performance.…”