Elsevier

Computers in Industry

Volumes 92–93, November 2017, Pages 161-177
Computers in Industry

A decision model and system for planning and adapting the configuration of enterprise information systems

https://doi.org/10.1016/j.compind.2017.08.004Get rights and content

Highlights

  • A decision model for planning investments on enterprise information systems is proposed.

  • EIS investment configurations are intensive in time-cost-quality, and risk decision criteria.

  • The model composes several alternatives and decision criteria for multiple decision models.

  • The software DecISIonAl was created to automate the configuration of decision scenarios.

  • These multi-stakeholder scenarios can be composed to search for optimal solutions.

Abstract

The decision to invest in an enterprise information system is usually made without taking into account the different types of subsequent decisions and without understanding the hidden implications of making them. This paper presents a decision-making model named DecISIonAl used to evaluate and manage implementation risks on ERP and CRM projects before the actual investment is made. This model was implemented into a web-based system to facilitate configuring, comparing, and selecting implementation plans by evaluating their impact in terms of cost, time, benefits, human resources capabilities, and risks. We apply our decision model to investment case studies in two enterprises. The results show a level of compliance between 80% and 83% when comparing the implications estimated by simulated scenarios and the actual investments.

Introduction

Information technology (IT) investments have become vital to deal with the changing and competitive organizational environment. Most companies invest on enterprise information systems (EIS) such as enterprise resource planning (ERP) and customer relationship management (CRM) systems. These EIS investments support several transactional, informational, and strategic capabilities expected in an IT portfolio [1].

The decision to invest on these EIS seems evident, however this decision entails implications that are not always considered by decision makers when planning the investment: high and hidden costs in implementation, time required for aligning the EIS generic data and processes with organizational-specific needs [2], specific human capabilities required in configuration, implementation and deployment risks, among others. Even when two companies face an EIS implementation within similar conditions, the way each one faces the risks of the implementation process is determinant to assure the success or failure of the project [3]. For example, some companies adopt open source platforms to tackle financial issues. Unawareness of the subsequent decisions associated with an EIS investment (e.g., target platform, implementation model, maintenance model) increases the issues and uncertainty in the implementation and management of the investment. Naturally, the subsequent decisions and implications appear once the investment has already been made and then it is too difficult and expensive to change them.

Moreover, the following challenges have been identified recently as the next generation of Enterprise Information Systems. The enterprise design (business and IT models) must be continuously adjusted according to changes in the business strategy or to changes in IT capabilities. These models must support automated processing operations for decision making on potential risks and opportunities due to enterprise change [4]. In addition, faster, better, and smarter decisions on enterprise automation require representing and analyzing a large amount of environmental and operational information on enterprise models to achieve sustainable enterprise systems [5]. Despite this challenge refer to EIS design and operation, it also applies for planning the adoption and configuration of EIS which require representing and analyzing a large amount of information for decision-making (e.g., product information, implementations risks, and decisions made by multiple stakeholders). This amount of asymmetric information increases the complexity to support decision-making by IT managers [6]. Therefore, modelling tools and systems for model analysis are required for designing and adapting EIS before they are actually changed [4]. We illustrate these challenges by presenting a detailed discussion of open issues found in literature review (see Section 2).

We aim to provide a suitable and comprehensive approach to perform an ex-ante evaluation of the implications related to an EIS investment project. This paper presents the design, implementation, and evaluation of three artefacts to plan and adapt investments in new EIS (see Section 3). First, we create a metamodel to abstract and support the implementation of model instances related to the IT investment planning domain (see Section 4.1). Second, we create a decision model instance which defines and relates 12 decision models or types (e.g., EIS modules, EIS implementation type, EIS deployment model) representing common characteristics to configure and justify investment scenarios (see Section 4.2). Each decision type contains decision alternatives or options representing the variable characteristics for a decision-making process on ERP and CRM systems. The model defines, for each decision option, a set of decision criteria or implications as a reference value evaluated for a specific dimension (i.e., costs, time, human resources, and risks) in terms of a specific business activity (e.g., maintenance, implementation, integration). This decision model also incorporates EIS business drivers, EIS-specific risks, and IT investment-generic risks as additional decision criteria to analyze the impact of an investment scenario. Finally, we develop a web-based Decision-Support System (DSS) to enact the instantiated model (see Section 5). The DSS allows multiple stakeholders to configure different EIS decision scenarios by selecting decision options for each of the decision models representing an EIS investment/implementation plan. Then, the predefined model constraints turn subsequent decisions mandatory or unavailable and the resulting implications (e.g., EIS-specific risks) are presented to decision makers to enrich the impact analysis and decision-making capabilities. The DSS also allows to automatically combine multiple EIS configurations defined for a common investment into an optimal configuration. To achieve this, we have defined in previous works [7], [8], [9] a program synthesis to specify configuration constraints (e.g., optimization functions, hard limit constraints) for general-purpose domain models and to automate their processing by using constraint programming.

An ex-ante evaluation of the implications at the planning stage of an EIS investment, allows the organization to identify and allocate needed resources, to understand and manage risks, and to manage uncertainty before an EIS implementation. We used the proposed model to analyze the EIS investments performed by two companies. The results show a level of compliance between 80% and 83% when comparing the implications estimated by simulated scenarios and the actual implications obtained when the investments were made (see Section 6). We conclude by discussing the capabilities of the decision model to be extended and reused, and also directions for future work (see Section 7).

Section snippets

Core concepts

An EIS is a system that integrates different modules required by a company to support the execution of multiple business processes in an efficient manner [1]. Two of the main subsystems that compose an EIS are ERP and CRM systems. Some of the modules considered by an ERP are the finance asset management, production, project management, sales, human resources, among others [10]. The CRM integrates different business processes and stakeholders, which are usually related to marketing, sales, and

Research methodology

The current research project follows the seven guidelines proposed by the design science research (DSR) approach [34]. For all guidelines we focused on both construction and evaluation processes.

DecISIonAl: a decision model on ERP/CRM investment

The proposed decision model is named DecISIonAl (Decisions on Information Systems and Impact Analysis) and aims to evaluate EIS decision scenarios and to compare their implementation impact. This allows organizations to plan the investment with informed decisions. Fig. 1 exemplifies the logic behind configuring EIS decision scenarios and the need of contrasting their resulting implications to select the optimal configuration: the scenario A based on costs implications or the scenario B based on

A decision support system for planning ERP/CRM investments

This section describes a DSS created to simulate ERP/CRM investment decision scenarios and to compare their associated implications. This system was implemented through a component model based on an OSGI Framework, developed using an Equinox components engine. This design aims to facilitate extensibility through the encapsulation of functionality into components.

There are two different types of stakeholders interacting with the DSS and the associated decision model. On the one hand, an expert

Case studies design

Two case studies related to companies that performed EIS investments are presented to evaluate the design of the proposed model. A case study approach was used since it is considered an important method to build and illustrate theories in research areas such as information systems, and management and organizational planning [54], [55]. In particular, this evaluation uses a comparative structure [55] by analyzing the results produced in multiple decision scenarios simulated in the model against

Conclusions and future work

Most organizations are aware of the need to invest on EIS in order to rapidly adapt to the competitive environment. However, they do not always have the resources that are required to perform and ex-ante evaluation of the implications related to an IT investment project. This paper has presented a decision model and a DSS that allow companies to identify the major decisions that should be considered when performing the analysis of an EIS investment.

The decision model defines the structure and

Acknowledgements

The authors would like to thank to Claudia López, Sebastian Lesmes, and Guillermo Beltrán who collaborated in the implementation and validation of this work.

References (58)

  • G. Colson et al.

    Editorial: Models and methods in multiple objectives decision making

    Math. Comput. Model.

    (1989)
  • K.T. Cho

    Multicriteria decision methods: an attempt to evaluate and unify

    Math. Comput. Model.

    (2003)
  • J. Broad

    System Development Life Cycle (SDLC)

  • A. Benlian et al.

    Opportunities and risks of software-as-a-service: findings from a survey of it executives

    Decis. Support Syst.

    (2011)
  • J. Kim et al.

    A model for evaluating the effectiveness of CRM using the balanced scorecard

    J. Interact. Mark.

    (2003)
  • T. Oliveira et al.

    Assessing the determinants of cloud computing adoption: an analysis of the manufacturing and services sectors

    Inform. Manage.

    (2014)
  • D.L. Olson et al.

    Enterprise information system trends

  • J.E. Scott et al.

    Managing risks in enterprise systems implementations

    Commun. ACM

    (2002)
  • J. Ni et al.

    Information technology investment decisions under asymmetric information: a modified rational expectation model

    Int. J. Inf. Technol. Decis. Mak.

    (2009)
  • L. Ochoa et al.

    Program synthesis for configuring collaborative solutions in feature models

  • L. Ochoa et al.

    Searching optimal configurations within multiple feature models

  • L. Ochoa et al.

    Using decision rules for solving conflicts in extended feature models

    2015 ACM SIGPLAN International Conference on Software Language Engineering

    (2015)
  • K. Ganesh et al.

    ERP as a business enabler

    Enterprise Resource Planning: Fundamentals of Design and Implementation, Management for Professionals

    (2014)
  • A. Payne et al.

    A strategic framework for customer relationship management

    J. Mark.

    (2005)
  • C.W. Holsapple

    DSS architecture and types

    Handbook on Decision Support Systems 1: Basic Themes, International Handbooks Information System

    (2008)
  • C. Kahraman et al.

    Selection among ERP outsourcing alternatives using a fuzzy multi-criteria decision making methodology

    Int. J. Prod. Res.

    (2010)
  • N. Karaarslan et al.

    An application for modular capability-based ERP software selection using AHP method

    Int. J. Adv. Manuf. Technol.

    (2008)
  • J. Sarkis et al.

    Evaluation of enterprise information technologies: a decision model for high-level consideration of strategic and operational issues

    IEEE Trans. Syst. Man. Cybern. C

    (2006)
  • A. Teltumbde

    A framework for evaluating ERP projects

    Int. J. Prod. Res.

    (2000)
  • Cited by (14)

    • Understanding CRM adoption stages: empirical analysis building on the TOE framework

      2019, Computers in Industry
      Citation Excerpt :

      Theoretically, CRM should be seen as a dynamic innovation that should be well managed in order to maximize operational efficiency and competitive agility for sustainability of the company [10–14]. Nevertheless, while literature often cites CRM projects or CRM performance (see, e.g. [15–17]), there is a perceived need for a holistic perspective of CRM adoption, which can help clear a vision or understanding about its adoption process at the firm level [6,13]. Considering these significant gaps in CRM literature, this study aims to understand how CRM is implemented at a company level, and how that could be influenced by determinants from the contexts of the TOE framework (i.e. technology-organisation-environment).

    • Configuration lifecycle management maturity model

      2019, Computers in Industry
      Citation Excerpt :

      Another crucial factor when discussing CLM is the feasibility analysis of a CLM system. Investing in several enterprise information systems, such as ERP, CRM, and product configuration systems, is a decision based not only on the needs of an organization and the targeted benefits, but also on the cost of implementation and maintenance, time spent, and return on investment [59,60]. In conclusion, the literature addresses the issue of handling configuration models that have a common systematic approach.

    • Constraint programming heuristics for configuring optimal products in multi product lines

      2019, Information Sciences
      Citation Excerpt :

      For instance, an automated process may consider a set of previously selected business drivers and determine the corresponding set of EIS modules to configure as shown in Fig. 1. In this example, the whole MPL has more than 1 billion valid configurations [10]. Additionally, users can define a set of configuration constraints orthogonal to the configuration process, and usually related to non-functional requirements.

    • Future developments in cyber risk assessment for the internet of things

      2018, Computers in Industry
      Citation Excerpt :

      Building upon the CyVaR, a unifying economic framework proposed the use of measurement units for cyber risk [24]. Other cyber value analysis methods have proposed to calculate the cost of different cyber-attack types [38], but they lack data for validating their models. This lack of data has motivated the development of a proof-of-concept method [23], based on data assumptions.

    • Construction of Enterprise Marketing Decisions Model Based on Decision Tree Algorithm

      2023, Proceedings - 2023 International Conference on Electronics and Devices, Computational Science, ICEDCS 2023
    View all citing articles on Scopus
    View full text