Strategic Analytics E-Book

Martin Kunc

• Herausgeber: John Wiley & Sons
• Kategorie: Fachliteratur
• Sprache: Englisch

Seitenzahl: 655

Veröffentlichungsjahr: 2018

Beschreibung

Defines common ground at the interface of strategy and management science and unites the topics with an original approach vital for strategy students, researchers and managers

Strategic Analytics: Integrating Management Science and Strategy combines strategy content with strategy process through the lenses of management science, masterfully defining the common ground that unites both fields. Each chapter starts with the perspective of a certain strategy problem, such as competition, but continues with an explanation of the strategy process using management science tools such as simulation. Facilitating the process of strategic decision making through the lens of management science, the author integrates topics that are usually in conflict for MBAs: strategy and quantitative methods. Strategic Analytics features multiple international real-life case studies and examples, business issues for further research and theory review questions and exercises at the end of each chapter.

Strategic Analytics starts by introducing readers to strategic management. It then goes on to cover: managerial capabilities for a complex world; politics, economy, society, technology, and environment; external environments known as exogenous factors (PESTE) and endogenous factors (industry); industry dynamics; industry evolution; competitive advantage; dynamic resource management; organisational design; performance measurement system; the life cycle of organisations from start-ups; maturity for maintaining profitability and growth; and finally, regeneration. 

• Developed from the author's own Strategy Analytics course at Warwick Business School, personal experience as consultant, and in consultation with other leading scholars
• Uses management science to facilitate the process of strategic decision making
• Chapters structured with chapter objectives, summaries, short case studies, tables, student exercises, references and management science models
• Accompanied by a supporting website

Aimed at both academics and practitioners, Strategic Analytics is an ideal text for postgraduates and advanced undergraduate students of business and management.

Leseprobe

Table of Contents

Cover

About the Companion website

Chapter 1: Introduction to Strategic Analytics

1.1 What is Analytics?

1.2 What is Management Science?

1.3 What is Information Technology: New Challenges?

1.4 What is Strategic Management?

1.5 Strategy Analytics: Integrating Management Science with Strategic Management

References

Chapter 2: Dynamic Managerial Capabilities

for a Complex World Under Big Data

2.1 Dynamic Managerial Capabilities

2.2 Integrating Management Science and Strategic Management: Managers as Modelers

2.3 End of Chapter

References

Further Reading

Chapter 3 External Environment

: Political, Economic, Societal, Technological and Environmental Factors

3.1 The PESTE Analysis

3.2 Integrating Management Science in the Strategic Management Process

3.3 End of Chapter

References

Further Reading

Chapter 4: Industry Dynamics

4.1 Defining the Industry

4.2 Porter's Five Forces and Industry Dynamics

4.3 Integrating Management Science into Strategic Management

4.4 End of Chapter

References

Further Reading

Chapter 5: Industry Evolution

5.1 Dynamic Behavioral Model of Industry Evolution

5.2 Integrating Management Science into Strategic Management

5.3 End of Chapter

References

Further Reading

Chapter 6: Competitive Advantage: Static Analysis

6.1 The Direction of a Company: Vision and Mission

6.2 Defining Value and Market Segmentation

6.3 Mapping the Activities to Deliver Value

6.4 Type of Business Strategies

6.5 Integrating Management Science into Strategic Management

6.6 End of Chapter

References

Further Reading

Chapter 7: Dynamic Resource Management

7.1 Resources and Capabilities

7.2 Resource Management

7.3 Integrating Management Science into Strategic Management

7.4 End of Chapter

References

Further Reading

Chapter 8: Organizational Design

8.1 Organizational Components

8.2 Integrating Management Science into Strategic Management

8.3 End of Chapter

References

Further Reading

Chapter 9: Performance Measurement System

9.1 Measuring Financial Performance

9.2 Strategic Controls

9.3 Integrating Management Science into Strategic Management

9.4 End of Chapter

References

Further Reading

Chapter 10: Start‐ups

10.1 The Components of a Business Plan for a Start‐up

10.2 Financial Management

10.3 Integrating Management Science into Strategic Management

10.4 End of Chapter

References

Further Reading

Chapter 11: Maturity

11.1 Strategies for Mature Organizations

11.2 Integrating Management Science into Strategic Management

11.3 End of Chapter

References

Further Reading

Chapter 12: Regeneration

12.1 Strategies for Regenerating Organizations

12.2 Integrating Management Science into Strategic Management

12.3 End of Chapter

References

Further Reading

Index

End User License Agreement

List of Tables

Chapter 1

Table 1.1 A selection of Analytics

tools.

Table 1.2 Approaches to analyze big data.

Table 1.3 Qualitative methods in Management Science.

Table 1.4 Quantitative methods to calculate one attribute in Management Science.

Table 1.5 Quantitative methods to replicate in Management Science.

Table 1.6 Quantitative methods to optimize in Management Science.

Table 1.7 Description of the strategic management process components.

Table 1.8 Four situations between strategy and business analytics.

Table 1.9

Complexity facing Strategy Analytics

Practitioners.

Table 1.10

Relationship between the problems faced in the strategic management

process and the management science

modeling process.

Chapter 2

Table 2.1 Praxis and practices for three practitioners in strategizing.

Chapter 3

Table 3.1 PESTE analysis.

Table 3.2 Methods to evaluate the impact of external factors.

Table 3.3 Scale of relative importance to evaluate PESTE factors.

Table 3.4 Pairwise comparison matrix for the PESTE factors.

Table 3.5 Normalized pairwise comparison matrix for the PESTE factors.

Table 3.6 Priority vector for PESTE factors.

Table 3.7 Priority vector for PESTE factors.

Table 3.8 Pairwise comparison matrices for alternative international markets.

Table 3.9 Priority vectors for alternative international markets.

Table 3.10 Criteria selected.

Table 3.11 Subjective pairwise evaluation of social acceptance for types of power plants.

Chapter 4

Table 4.1 Dimensions to evaluate suppliers.

Table 4.2 Dimensions to evaluate buyers.

Table 4.3 Dimensions to substitutes.

Table 4.4 Dimensions to evaluate the threat of entrants.

Table 4.5 Dimensions to evaluate competitive intensity.

Table 4.6 Five forces analysis of the airline industry.

Table 4.7 Process to manage RMS.

Table 4.8 Five Forces

analysis for the easyJet case.

Chapter 5

Table 5.1 Factors driving the evolution of industries through the life cycle model

.

Table 5.2 Measures to track the evolution of the industry.

Table 5.3 Key uncertainties affecting managerial decision making during industry evolution.

Table 5.4 Evaluating behavioral

types, strategic actions and industry implications.

Table 5.5 Prior industry affiliations of the companies that entered the consumer digital camera market.

Chapter 6

Table 6.1 Cost drivers.

Table 6.2 Differentiation drivers.

Table 6.3 Cost drivers and the use of Management Science methods to manage costs.

Table 6.4 Differentiation drivers and Management Science methods to control differentiation drivers.

Table 6.5 Differences in decision‐making styles using Porter's generic strategies.

Table 6.6 Main decision‐making processes existing in each model of the company.

Chapter 7

Table 7.1 Differentiation drivers and their resources

and capabilities.

Table 7.2 Cost drivers and related key resources

and capabilities.

Table 7.3 Resource mapping

process.

Table 7.4 Integrating external and internal analyses: scenarios and resource mapping.

Table 7.5 Decision tree process.

Table 7.6 EMV calculation.

Table 7.7 Risk profile.

Chapter 8

Table 8.1 Generic set of organizational operational and management processes.

Table 8.2 Business process modelling techniques and methodologies.

Table 8.3 Applications of Little's formula

in the analysis of business processes.

Table 8.4 Most common units in a hospital.

Chapter 9

Table 9.1 Financial performance measures

.

Table 9.2 Financial ratios for four large European low‐cost airlines in 2016.

Table 9.3 Systems thinking

strategic maps.

Table 9.4 Systems thinking

strategic maps process for a fishing company.

Table 9.5 Systems thinking

strategic maps of a fishing company.

Table 9.6 Analytical tools to describe data: descriptive analytics.

Table 9.7 Analytical tools to identify the root causes of performance

.

Table 9.8 Companies involved in the study.

Table 9.9 Adoption of performance measurement systems in the companies of the study.

Table 9.10 Perceived usefulness of performance management systems on the companies.

Chapter 10

Table 10.1 Elements of a business plan.

Table 10.2 Key questions for a business model

.

Table 10.3 Key success factors.

Table 10.4 Cash flow components

: critical assumptions and scenarios factors.

Table 10.5 Example of a profit and loss

statement.

Table 10.6 Example of a profit and loss

statement.

Table 10.7 Sources of funding.

Table 10.8 Performing Monte Carlo

simulation using Microsoft Excel®.

Table 10.9 Steps for performing a multi‐criteria decision analysis for a winery.

Chapter 11

Table 11.1 Key changes occurring in the industry during the industry life cycle.

Table 11.2 Strategic options for concentrated growth, and market andproduct development

.

Table 11.3 Examples of international strategies

.

Table 11.4 Metrics for six projects related to building operational capabilities.

List of Illustrations

Chapter 1

Figure 1.1 The analytics field..

Figure 1.2 Management science styles..

Figure 1.3 Basic categorization of modeling methods in management science.

.

Figure 1.4 Strategic management process.

Chapter 3

Figure 3.1 External factors affecting the organization.

Figure 3.2 A PESTE analysis

to evaluate international market attractiveness.

Figure 3.3 Management science styles.

Chapter 4

Figure 4.1 Financial performance of the airline industry. (a) Airline share prices by region. (b) Airline operating profits by region.

Source:

IATA Economics, 2015b. Reproduced with permission from IATA Economics.

Figure 4.2 Price and capacity performance of the airline industry. (a) Average return fare worldwide and USA airline yields. (b) Airline fleet development.

Source:

IATA Economics, 2015b. Reproduced with permission from IATA Economics.

Figure 4.3 Load factors for passenger and freight: the two main revenue streams in the airline industry. (a) Load factors for passengers and freight markets. (b) Total passenger market with seasonality.

Source:

IATA Economics, 2015b. Reproduced with permission from IATA Economics.

Figure 4.4 Simple description of RMS.

Figure 4.5 Trend of tweets for the companies during October 2011.

Source:

He al. (2013: figure 2, page 467. Reproduced with permission from International Journal of Information Management.

Figure 4.6 Management science styles.

Source:

Adapted from Walker (2009). Reproduced with permission from Elsevier.

Figure 4.7 The dynamics of potential passengers.

Figure 4.8 Balancing loop and other causal links describing rivals' retaliation through fare reduction.

Figure 4.9 The dynamics of route saturation.

Chapter 5

Figure 5.1 Industry life cycle

in four stages.

Figure 5.2 Industry life cycle

and the number of companies in an industry.

Figure 5.3 Industry as a feedback system.

Figure 5.4 Dynamic behavioral

model of industry evolution.

Figure 5.5 Worldwide shipments from Japanese camera makers.

Source:

CIPA, 2015. Reproduced by permission of John Wiley & Sons.

Figure 5.6 Market development model.

Figure 5.7 Market development simulations. (a) Growth in the number of adopters and the decline in potential adopters over time; and (b) rates at which the transition occurs.

Figure 5.8 Company growth model.

Figure 5.9 Company growth simulation. (a) Production capacity adjustment. (b) Learning curve and unit costs. (c) Financial performance. (d) Operating profits.

Figure 5.9

Continued

Figure 5.10 Replacement sales.

Figure 5.11 Simulated industry performance over time and differences in profitability.

Source:

Caldart and Oliveira (2010: figure 6, page 101). Reproduced with permission of Elsevier.

Figure 5.12 Management science styles.

Source:

Walker (2009). Reproduced with permission of Elsevier.

Figure 5.13 Sales of cameras from 1977 to 2014.

Source:

CIPA (2015).

Figure 5.14 The photography industry landscape.

Source:

Taylordavidson.com.

Chapter 6

Figure 6.1 Mapping market segmentation: (a) mapping age/behavior vs. price and market size (size of boxes indicates market size); and (b) matrix: age vs. type of drinks in pubs.

Figure 6.2 Competitive advantage

paths and generic strategies.

Figure 6.3 Value chain.

Figure 6.4 Activity system map

for a full service airline in Asia.

Figure 6.5 easyJet Business Model Canvas.

Figure 6.6 Mobile telephone handset strategy canvas.

Figure 6.7 Sales in million units for bar and liquid soap.

Figure 6.8 Management science styles.

Figure 6.9 Mapping the sectors of the model with the concept of value

chain.

Figure 6.10 Market dynamics.

Figure 6.11 Operational resources.

Figure 6.12 Operational resources.

Chapter 7

Figure 7.1 Managerial decision‐making process responsible for resource management.

Figure 7.2 Resource development concept. Source: Adapted from Morecroft (2002).

Figure 7.3 Resource map of a bookstore chain.

Figure 7.4 Decision tree for the European option.

Figure 7.5 Management science styles.

Figure 7.6 Resource map of a wine

brand sold through specialist wine stores.

Chapter 8

Figure 8.1 Typical organization structures: functional, multidivisional and matrix represented using organizational charts.

Figure 8.2 Organizational network in the finance area includes direct and non‐direct linkages as well as links with diverse intensity.

Figure 8.3 Six organizational structures in the software industry by “Manu Cornet”.

Source:

www.bonkersworld.net (creative common license).

Figure 8.4 Management science styles.

Source:

Walker (2009).

Figure 8.5 Broad comparison between discrete event, agent‐based and system dynamics

modeling..

Chapter 9

Figure 9.1 What gets measured gets done.

Source:

Tapinos and Dyson (2007: figure 11.1). Reproduced with permission of John Wiley & Sons.

Figure 9.2 A strategic performance measurement system.

Figure 9.3 World capture fisheries and aquaculture production. Source: FAO (2016: Figure 1).

Figure 9.4 Management science styles.

Chapter 10

Figure 10.1 Example of a breakeven analysis using three scenarios for price and variable costs.

Figure 10.2 Management science styles.

Source:

Walker (2009). Reproduced with permission from Elsevier.

Chapter 11

Figure 11.1 Strategy selection matrix.

Source:

Adapted from Pearce and Robinson (2000).

Figure 11.2 The BCG Growth‐Share Matrix.

Source:

Adapted from Proctor and Hassard (1990).

Figure 11.3 The McKinsey/GE Industry Attractiveness‐Business Strength Matrix. Source: Adapted from Pearce and Robinson (2000).

Figure 11.4 Management science styles. Source: Walker (2009). Reproduced with permission of Elsevier.

Figure 11.5 Spreadsheet with linear programming

formulation.

Figure 11.6 Linear programming formulation using Solver in Excel.

Figure 11.7 Solver results from the problem.

Figure 11.8 Spreadsheet with the goal programming

formulation.

Figure 11.9 Configuration of the Solver.

Figure 11.10 First iteration by minimizing the deviation for return on investment.

Chapter 12

Figure 12.1 A model of the turnaround process

. Source: Adapted from Robbins and Pearce (1992: figures 1 and 2, page 291).

Figure 12.2 Example of a DEA model implemented in Excel. DMU 1 and 2 are inefficient.

Figure 12.3 Management science styles.

Source:

Walker (2009). Reproduced with permission of Elsevier.

Figure 12.4 Benefits realized through projects fill the gap between current and desired situation (Serra and Kunc, 2015).

Figure 12.5 Project implementation processes

at strategic and tactical levels.

Figure 12.6 Structure of the simulation model to evaluate remedial actions to correct deviations in projects.

Guide

Cover

Table of Contents

Begin Reading

Pages

C1

iv

xi

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

109

112

110

111

113

114

115

116

117

118

119

120

121

122

123

124

125

126

133

127

128

129

130

131

132

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

307

309

308

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

371

372

373

374

375

E1

Strategic Analytics

Integrating Management Science and Strategy

Edited by

Martin Kunc, Ph.D.

Copyright

This edition first published 2019

© 2019 John Wiley & Sons Ltd

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.

The right of Martin Kunc to be identified as the author of this work has been asserted in accordance with law.

Registered Offices

John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA

John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

Editorial Office

9600 Garsington Road, Oxford, OX4 2DQ, UK

For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.

Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print versions of this book may not be available in other formats.

Limit of Liability/Disclaimer of Warranty

While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

Library of Congress Cataloging‐in‐Publication Data

Names: Kunc, Martin, author.

Title: Strategic analytics : integrating management science and strategy / Martin Kunc, University of Warwick.

Description: Hoboken, NJ : John Wiley & Sons, [2019] | Includes bibliographical references and index. |

Identifiers: LCCN 2018025510 (print) | LCCN 2018028470 (ebook) | ISBN 9781118943694 (Adobe PDF) | ISBN 9781118943687 (ePub) | ISBN 9781118907184 (hardcover)

Subjects: LCSH: Management science. | Decision making. | Strategic planning.

Classification: LCC HD30.25 (ebook) | LCC HD30.25 .K8165 2018 (print) | DDC 658.4/012–dc23

LC record available at https://lccn.loc.gov/2018025510

Cover design by Wiley

Cover image: © Michel Leynaud/Getty Images

About the Companion website

This book is accompanied by a companion website:

www.wiley.com/go/kunc/strategic-analytics

The website includes:

Power points

Excels

Models

1Introduction to Strategic Analytics

Objectives

To explain Strategic Analytics

To introduce the main pillars of the book: the fields of analytics, management science

, information technology

, statistics and strategic management

To explain the fields of analytics, management science

, information technology

, big data

analytics and strategic management

Learning outcomes and managerial capabilities developed

Managers can learn to tackle complex problems in strategy through the integration of analytics within strategic management

processes

Principles of Analytics

: tools, support systems and methods

Identification of strategic problems

In today's environment, managers face turbulence and crisis. More information than ever is generated continuously: social media, financial performance management systems, customer relationship management systems, and internal and external reports. There are powerful trends: measuring and quantifying everything and skills in quantitative subjects are widely available. The problem for managers is not only to make sense of abundant quantitative information but also to engage with staff possessing analytical skills. Simultaneously managing multiple factors under pressure requires new managerial capabilities. Managers need to develop their strategies using clear strategy processes supported by the increasing availability of data. This situation calls for a different approach to strategy, such as an integration with analytics, as the science of extracting value from data and structuring complex problems.

Managers' decision processes can fall into a continuum which has on one hand pure analysis, which relies on established processes, and pure synthesis that involves identification of patterns and new ideas (Pidd, 2009). Strategic planning involves a mix between both extremes: analysis to identify problems and synthesis to observe trends and emerging situations before competitors. However, there is not a clear process when combining the two. Sometimes, emerging situations are discovered and then analysis is employed to confirm their impact while transforming them into new emergent strategies. In other circumstances, there is a careful planning process involving extensive data gathering and the construction of complex financial, operational and other type models to validate the new strategies.

The idea behind “Strategic Analytics” is to answer a simple question: how can quantitative and qualitative information be used to make strategic decisions? Strategic Analytics does not imply turning managers into quantitative analysts or quantitative analysts into expert strategists. Organizations need interdisciplinary teams comprised by members who can talk to each other sharing a common language. Thus, Strategic Analytics works on the basis of providing a reasonable understanding of how a variety of quantitative methods, in conjunction with structured and unstructured data, can be used to help strategic decision making in any organization. There is also the intent to show the real and practical benefit of Strategic Analytics. Strategic Analytics does not pretend to offer easy solutions to strategic problems but different ways of analyzing and solving strategic problems beyond the traditional qualitative approach to strategic management. Strategic Analytics is also an understanding of the context and processes in which analytics skills can be applied to support strategic management.

Future managers are taught a wide variety of concepts in strategy subjects but they are not taught how to apply them or even to connect them to related problems. Future managers need to develop capabilities to tackle problems that are not structured in a neat way like case studies are. In that sense, each chapter focuses on a case study with limited information that has to be solved applying a combination of theoretical concepts and analytical methods (quantitative and qualitative). The aim of this founding principle is to integrate strategic concepts with analytical tools in a unique set of capabilities (skills) to help future managers to tackle strategic problem using multiple sources of information. The main benefit is that quantitative methods, which are usually seen as a difficult experience for managers, are connected with a hands‐on subject like strategy. Therefore, managers can learn capabilities to tackle complex problems in strategy through the integration of analytics (quantitative/qualitative methods to extract value from data) within strategic management processes. Therefore, the book will provide a bridge to integrate quantitative methods with their application in strategy adding rigorous methods to solve real issues in strategy.

The rest of the chapter introduces the main pillars of the book: the fields of analytics, management science, information technology, statistics and strategic management.

1.1 What is Analytics?

Organizations are competing using analytics because there is an increasing amount of data, people with capabilities to use data and, in a highly competitive environment, it is more difficult to compete effectively. While organizations can use basic descriptive statistics from any of their existing data, organizations using analytics apply modeling to understand their environments, predict the behavior of key actors, e.g. customers and suppliers, and optimize operations. Organizations can obtain competitive advantage using multiples analytics applications but it requires a new type of organization and management (Davenport, 2006):

a companywide embrace of analytics impels changes in culture, processes, behavior and skills for many employees. And so, like any major transition, it requires leadership from executives at the very top who have a passion for the quantitative approach… CEOs leading the analytics charge require both an appreciation of and a familiarity with the subject. A background in statistics isn't necessary, but those leaders must understand the theory behind various quantitative methods so that they recognize those methods' limitations‐which factors are weighed and which ones aren't… Of course, not all decision should be grounded in analytics… For analytics‐minded leaders, then, the challenge boils down to knowing when to run with the numbers and when to run with their guts (Davenport, 2006: pages 102–103)

An analytical perspective is important, when data has become a key strategic asset of organizations in recent years, and analytics creates value by delivering systematic decision support in a well‐timed way (Laursen and Thorlund, 2010; Holsapple et al., 2014). Business analytics, one of the multiple branches in Analytics, comprises three key elements: Information Systems, Human Competencies, and Business Processes (Laursen and Thorlund, 2010). Business analytics reflects the convergence of three disciplines: statistics, information systems, and management science (Laursen and Thorlund, 2010). While the supporting disciplines are traditional, the innovation lies in their intersections. For example, data mining aims to understand characteristics and patterns among variables in large databases using a variety of statistical analysis, e.g. correlation and regression analysis. Information technology provides data and supports decision support systems, which are sustained by management science tools together with statistical analysis, for the development of analytics. Business analytics is rooted in advances of information technology systems, which involve the acquisition, generation, assimilation, selection and presentation of data, together with tools, statistics and management science, to develop the data into knowledge to support decision making. In a similar definition, Mortenson et al. (2015) suggest analytics is the intersection of basic disciplines: technologies (electrical engineering and computer science), decision making (psychology and behavioral science) and quantitative methods (mathematics, statistics and economics); and their applications: information systems, artificial intelligence and operational research.Figure 1.1 shows Mortenson et al.'s (2015) representation of the concept of analytics.

Figure 1.1 The analytics field..

Source: Mortenson et al. (2015: figure 1, page 586). Reproduced with permission of Elsevier

In terms of types of analytics, Davenport proposes three types: descriptive, predictive, and prescriptive (Davenport, 2013), which are described below.

D

escriptive Analytics

.

It employs traditional statistical skills to present data collected from internal organizational activities and external data. It is utilized to understand what happened during their past business activities in order to disclose whether the current business objectives have been obtained. Then the next step is to investigate the reasons behind the results by drilling down into more detailed data and explore scientifically, e.g. test and validate or reject hypotheses. It is characterized traditionally as data mining, business intelligence and dashboards.

Predictive Analytics

.

It involves statistical and mathematic techniques to predict future unknown events or behaviors based on historical data to support operations. Some of the popular techniques in this area cover decision tree, text analytics

, neural networks, regression

modeling, and time‐series forecasting

. Predictive analytics

also relies on data mining and other machine learning algorithms to identify trends, patterns, or relationships as volume, variety, and velocity of data increases.

Prescriptive Analytics

.

It uses optimization

and/or simulation

to identify the best alternatives to improve performance. Prescriptive analytics is used in many areas of business, including operations, marketing, and finance. For example, finding the best pricing or portfolio of new product investments strategy to maximize revenue. Prescriptive analytics can address questions such as: How much should we produce to maximize profit? Should we change our plans if a natural disaster closes a supplier's factory and if so, by how much?

In terms of analytics, there is a wide variety of tools so an exhaustive list of tools will be beyond the scope of this book. Table 1.1 shows a list of the most common tools extracted from Gandomi and Haider (2015) and Chen et al. (2012).

Table 1.1 A selection of Analytics tools.

Method

Brief description

Suggested further reading

Text analytics

Text analytics is employed to extract information from textual data such as social network, blogs, online forums, survey responses, documents, news and any logs from interactions with customers. The basic tools to perform the analysis are statistical analysis, computational linguistics and machine learning. Some of the outcomes from their use are:

Information extraction generates structured data from unstructured text

Text summarization provides summaries from multiple documents to present the key information existing in the original text considering location and frequency of text units

Sentiment analysis analyze text, which contains opinions, to infer positive or negative sentiment. The analysis can be performed at document, sentence and aspect levels

Aggarwal and Zhai (

2012

)

Data analytics

Data analytics

comprises technologies based on data mining and statistical analysis. More specifically, data mining algorithms, such as statistical machine learning (Bayesian networks, Hidden Markov models), sequential and temporal mining, spatial mining, process mining and network mining, are the key tools. Data mining

algorithms perform classifications, clustering

, regression

, association analysis and network analysis

.

Berry and Linoff (

1997

) Hand et al. (

2001

)

Network

analytics

It evolved from bibliometric analysis, citation networks and co‐authorship networks tools. The current focus is on link mining, which is the prediction of links – social relationships, collaboration – between end users, customers, etc., and community detection, which involves representing networks as graphs and applying graph theory to identify communities Another focus is on the dynamics of social networks

using agent‐based models, social influence and information diffusion models

Aggarwal (

2011

)

Visual analytics

The objective is to facilitate analytical reasoning through interactive visual interfaces

that synthesize information from big, ambiguous and dynamic data. It integrates information visualization with data management and data analysis together with human perception and cognition fields. Some applications are in:

Spatial data

such as geographic measurements, GPS data, and remote sensing

Temporal data

such as patterns, trends and correlations of data over time

Network

data related to different real networks such as transportation, electric power grids, communities, etc.

Andrienko and Andrienko (

2005

) Dykes et al. (

2005

) Simoff et al. (

2008

)

From a process perspective, Liberatore and Luo (2010) suggest analytics follows a four‐step process. First, data is collected from diverse systems. Then data extraction and manipulation are two tasks performed together with the objective to obtain and organize the data for analysis. The second step is data analysis comprising three activities: analysis; predictive modeling; and optimization. Analysis can involve presenting and analyzing data using interactive tables, charts and dashboards. Predictive modeling is employed to estimate trends, classify data and validate relationships. Optimization attempts to obtain the optimal solution. The methods employed in this stage are discussed in the next section. The third step is the interpretation of the analysis into insights. Data visualization offers insights into what happened in the past. The purpose of predictive modeling is to help to foresee future issues if the current trends continue. Optimization gives potential solutions under certain situations. The fourth step involves translating the insights into actions related to operational aspects, redefining processes or defining/adjusting strategies.

It is important to remember that big data and data science terms are different: big data refers to the data in large volume; and data science to the methods to analyze this data. One of the recent fields emerging from the rise of big data and analytics is data science. Data science is as an interdisciplinary field that combines statistics, data mining and other tools to generate analytical insights and prediction models from structured and unstructured big data (George et al., 2016). Data science focuses on the systematic study of the organization, properties, and analysis of data and the process of inference together with confidence in the inference (Dhar, 2013). Big data, as the raw material for data science, has a set of characteristics: volume, variety and velocity. Building on the notion of volume, “data scope” refers to the level of completeness of data describing events (George et al., 2016). Scope implies a wide range of variables, whole populations rather than samples, and numerous observations on each participant. A higher number of observations shifts the analysis from samples to populations (McAfee and Brynjolfsson, 2012). Variety relates to the extreme diversity in the type of data provided from internal and external sources from numeric to textual data. In terms of velocity, data is streaming in at unprecedented speed and must be dealt with in a timely manner. Another characteristic in addition to increasing velocity and variety of data is variability since data flows can be highly inconsistent with periodic peaks occurring in different units of time, e.g. minutes, days, months.

For the analysis of big data, there are a number of challenges because traditional statistical concepts apply to situations where a sample of the population is analyzed; however big data can capture the entire population (George et al., 2016). First, there is a (very) large number of potential explanatory variables available. Secondly, the data are too large to be processed by conventional personal computers. Thirdly, a model that shows a strong relationship between independent variables and the dependent variable together with strong predictive validity, may not demonstrate a causal relationship between variables. Causality needs a field experiment in which an independent variable is manipulated for a random sample of the target population. Fourthly, statistical significance becomes less meaningful when working with big data. Variables that have a small effect on the dependent variable will be significant if the sample size is large enough and spurious correlations will appear when using a large number of variables (George et al., 2016).

Laursen and Thorlund (2010) offer a useful guide to connect questions, type of analysis and competencies required to perform analytics, see Table 1.2.

Table 1.2 Approaches to analyze big data.

Type of question

Type of analysis

Competencies required

Is there any relation between an action and the results obtained?

The most basic analysis is to retrieve a set of data that provides evidence of the actions, e.g. price changes, and results, e.g. sales changes. Then, the data is presented in a certain format, e.g. a contingency table, bar charts, etc., leaving the interpretation to the users. This is usually the realm of visual analytics (Reference) or descriptive analytics

Data manager or report developer to retrieve data and organize in a visual format. No knowledge of the business is required

What is the behavior of a certain customer/ performance?

Sometimes, there is no knowledge about the performance of a certain variable so the result of this analysis is simple descriptive statistics such as sums, average, range and standard deviations. This is usually the realm of descriptive analytics (Keim et al.,

2008

)

Data manager or report developer to retrieve data and organize in a visual format. No knowledge of the business is required

What is the correlation between a certain event and the results? Is the action causing any effect on the results?

Hypothesis‐driven analytics employs statistical analysis and the intention is to validate the correlations between events and results. Interestingly, the use of traditional statistical tests to validate correlations, e.g. 95% confidence interval, may not be relevant in the context of big data

since the data contains the whole population (so a comparison of the average between variables may be enough) (George et al.,

2016

). In terms of analysis, they can be cross‐sectional, e.g. regression

analysis, or longitudinal, e.g. time series analysis

Analytics

to execute appropriate statistical tests and business knowledge to ensure the quality of the selection of the variables, e.g. variables clearly describe the business

What can be learnt from the events that occurred/actions taken?

Data‐driven analytics, e.g. data mining and explorative analytics, attempts to create models for specific decision support. This is not theory driven but data driven. In other words, the algorithms will find the optimal model without restrictions. The quality of the findings depends on the performance in the data set not the theoretical significance. However, there are validation procedures such as testing the model in a different subset of data than the one employed to develop it (Lismont et al. 2017) Data‐driven is best suited for complex tasks due to changes in data, large amounts of data and variables, and limited initial knowledge

Computer science to develop algorithms, e.g. neural network, decision trees, binary regression

, and machine learning that search across the data and creates models explaining the relationships in the data Business is also required to evaluate the reasonability of the model

What is the best model to describe the factors driving the results with and without a clear variable in mind?

Data‐driven analytics can also be performed with a specific variable in mind but no knowledge about the drivers of the variable. Then, the model is employed in predictive mode to find out future events (predictive analytics

). In this case, data mining has a target variable and a lot of input variables without knowing their impact on the target variable Another analysis is just simply looking for patterns in data. This is employed when there is a large set of variables that do not generate sufficient information so there is a need to reduce, or group, them into a smaller, and more meaningful, set of variables. This is usually performed through cluster analysis (Kaufman and Rousseeuw,

2009

), cross‐sell or basket analysis models (Berry and Linoff,

1997

) and up‐sell models (Cohen,

2004

)

Computer science to develop algorithms to search for variables to create a model

1.2 What is Management Science?

While analytics provide the context for the use of data for decision making, management science, together with statistics, is one of the engines behind analytics. According to Mayer et al. (2004), the six purposes of management science in the area of supporting strategy making are:

Re

search and analyze:

Management science can generate knowledge in a specific domain, and specific issues, to develop deeper understanding of the impact of issues on the performance of organizations or society.

Design and recommend:

This activity focuses on translating available knowledge into new strategies by making recommendations or designing the strategies.

Provide strategic advice:

Its role is to provide advice to a client on a strategy for achieving certain goals given a certain context (e.g. environment, responses of other actors, etc.). It involves understanding the requirements of the decision makers and the organization.

Clarify arguments and values:

In this role, management science

practitioners analyze the values and argumentation systems that underpin the organizational debate. Given this purpose, analysts seek to improve the quality of the discussion detecting biases or limitations in the arguments

Democratize:

Analysis activities can have normative and ethical objectives related to the stakeholders

. Powerful interests can be involved affecting the discussion so management science

analysts can support views and opinions overlooked due to the lack of power. For example, simulation

models may be useful to support claims over environmental impact.

Mediate:

In this role, management science

designs the rules and procedures for the strategic making process and manages the interactions and progress of the process, especially facilitating meetings with different stakeholders

and strategic decision makers.

The top part of Figure 1.2 shows management science mostly oriented towards objects and focused on systems, strategies, and models, while the bottom part indicates its orientation to subjects, which comprises working with people (decision makers, stakeholders, experts) and their interactions during the strategic management process. The arrows point to “styles” of analysis. When management science encompasses systems and strategies, the styles are rational, argumentative and client‐advisory as it provides reasons and arguments to support the decision‐making process. When management science approaches the subjective part of strategic management processes, the styles will be participative, interactive and process‐oriented since their role is to generate forums for discussions where modes are “transitional objects” (Mayer et al., 2004).

Figure 1.2 Management science styles..

Source: Walker (2009). Reproduced with permission of Elsevier

In practical terms, the translation of a strategic problem to the analytical world is accomplished through the process of finding meaning and structure to the strategic problem through the development of a model (Mayer et al., 2004). The model helps to visualize the data required together with precise definitions reducing ambiguity and confusion. The structure of the model can be validated to identify incoherent statements and spurious assumptions. The results of the model reflect the current and future behavior of the strategic problem given the assumptions about the problem. Finally, experiments on solutions and selection of strategic choices can illuminate a satisficing solution. The importance of each of the previous steps, translation, visualization, validation, simulation and experimentation, depends on the emphasis of the use of the model as indicated in Figure 1.2. A key aspect to highlight is the model cannot be only one solution but part of multiple models or methodologies to triangulate the strategic insights obtained. Figure 1.3 shows a categorization of modeling methods in management science reflecting the richness of the field.

Figure 1.3 Basic categorization of modeling methods in management science..

Source: Adapted from Williams (2008). Reproduced with permission of John Wiley & Sons

From a management science perspective, qualitative methods are predominantly employed to structure problems that are ill‐structured (not clearly visualized in the minds of the strategic decision makers) or parameters difficult to quantify (Williams, 2008). The methods cover multiple set of issues. For example, one strategic problem may consist of understanding interconnectedness inside the organization and across industry (see Soft Systems Methodology, Cognitive mapping and Causal loop diagrams in Table 1.3) or the need to manage interactions with other actors/stakeholders (see Drama theory in Table 1.3). Other strategic problems addressed are uncertain futures (see Scenario analysis in Table 1.3) or uncertainty in the strategic choices existing (see Robustness analysis in Table 1.3). Qualitative methods can also work in situations where parameters are difficult to quantify (Williams, 2008) so the aim of these methods is to support decision makers to quantify parameters leading to an initial quantitative evaluation of the strategic problem. For example, decision trees can evaluate strategic actions considering subjective probabilities (see Decision trees in Table 1.3), and analytic hierarchy process and multi‐criteria decision analysis (see Analytic hierarchy process and Multi‐criteria decision analysis in Table 1.3) focus on the quantification of preferences in order to evaluate strategic options and choices.

Table 1.3 Qualitative methods in Management Science.

Method

Brief description

Suggested further reading

Soft Systems Methodology

Approach used to understand and model human activity systems. The process involves appreciating the problem situation, uncovering different views from the key stakeholders

, defining human activity systems and identifying desirable and feasible changes to improve the problem situation

Checkland and Scholes (

1990

)

Cognitive mapping

Group‐based process to make sense of complex problems developing shared understanding using individual decision makers' construction of reality

Eden and Ackermann (

2013

)

Causal loop diagrams

It is a method to represent the feedback structures responsible for the behavior of systems. It involves identifying causal relationships among concepts as well as the type of linkage (positive or negative). It can be a group‐ or individual‐based method

Sterman (

2000

)

Drama theory

A modeling method to understand activities between two or more competing or co‐operating entities considering that humans, as decision makers, may be irrational so employing drama can be a useful framework of analysis. It involves the use of role play

Bryant (

2003

)

Scenario

analysis

It is a group process to help organizations to learn about future events by considering alternative possible outcomes and how the future will affect them through stories

van der Heijden (

2005

)

Robustness analysis

It supports decision making when there is radical uncertainty about the future, so it attempts to support rational decisions today given unknowable future conditions. It resolves the paradox by assessing decisions in terms of the attractive future options that they may keep open. It involves the use of decision trees to evaluate the diverse configurations to be faced at the end of the decision

Rosenhead and Mingers (

2001

)

Decision trees

It basically consists of a simple diagram of decisions and consequences considering the uncertainty of their outcome through their probabilities. The decisions and outcomes are represented as branches and nodes, so the tree represents a sequential path of the set of decisions and their outcomes. Then decision makers choose the optimal decision, which reflects the maximum expected values

No specific reading is suggested since it is a fairly standard technique available in quantitative methods books

Analytic hierarchy process

It implies selecting between different options considering the multiple goals a decision maker has. The idea is to structure the criteria set hierarchically, with the overall goal of the decision at the top of the model, and then comparing between pairs of criteria while maintaining consistency using decision‐weights

Saaty (

1980

)

Multi‐criteria decision analysis

It is similar to the analytic hierarchy process

but the main focus is on the use of diverse methods for evaluating the weights on the criteria and then applying the weights to the criteria scores for each individual option

Belton and Stewart (

2001

)

Methods to calculate one parameter or a certain attribute of a system involve two types: deterministic; and stochastic. Deterministic methods can include a simple profits statement analysis, where the attribute is profit, to the evaluation of the relative efficiency across multiple decision‐making units, considering one dimension. Stochastic models assume that there is not one unique result of a certain attribute but a range of values. Depending on the purpose of the analysis, they can be evaluated as Statistical, Monte Carlo or Probabilistic (Table 1.4).

Table 1.4 Quantitative methods to calculate one attribute in Management Science.

Method

Brief description

Suggested further reading

Data Envelopment Analysis

This method evaluates the relative efficiency of similar decision‐making units in terms of inputs and outputs. The graphical nature of the tool allows the decision maker to see the optimum units at the frontier and highlight the inefficient units

Cooper et al. (

2006

)

Statistics

There is a large range of methods for analyzing data but the methods attempt to basically evaluate relationships between variables (correlation), independence between groups (factor analysis), differences between groups and treatments (analysis of variance), grouping items (clustering

) and finding explanatory factors (regressions)

Cortinhas and Black (

2012

)

Monte Carlo

It is a broad class of computational algorithms that rely on repeated random sampling to obtain a range of possible outcomes and the probabilities they will occur for any choice of action. It allows considering risk in quantitative analysis and decision making

No specific reading is suggested since it is a fairly standard technique available in quantitative methods books

Probabilistic

It involves the use of probabilities to evaluate the future result of a certain event based on judgmental, historical or experimental information. There are many different methods among them Bayesian methods, Fuzzy methods, Stochastic/Markov processes and Risk analysis

Cortinhas and Black (

2012

)

Methods to replicate are utilized to understand how a system behaved in the past, behaves currently or how it will behave in the future based on models reflecting the interactions between the different components of the systems. The aim of the methods is to achieve a representation of the system that is accepted by the users (validation) while simple enough to understand how it works (conceptualization). The methods can be classified as deterministic and stochastic. The main deterministic method is System Dynamics, a methodology developed at MIT in the late 1950s (Table 1.5). Stochastic methods consider the existence of uncertainty in the system and its impact on the performance of the system. The uncertainty may be generated by randomness (discrete event simulation) or aggregation of multiple individual agents (agent‐based simulation) (Table 1.5). Modeling can be very useful when the organization is facing up to complex issues as an interpretative approach to solve them. However, it is impossible to provide a comprehensive validation that a model is completely correct because of the interpretative approach to the complex problem. The key is the process of modeling, especially problem and model formulation (Robinson, 2004).

Table 1.5 Quantitative methods to replicate in Management Science.

Method

Brief description

Suggested further reading

System Dynamics

It represents organizations as a set of stocks and flows, which reflect accumulation processes, and feedback loops

representing information flows and causal relations between the components of the organization. The representation of the behavior is mainly developed through time series of variables. Models are developed in either group‐ or individual‐based processes

Morecroft (

2015

)

Discrete Event Simulation

This method evaluates the impact of randomness in the behavior of people or materials as they move through business processes considering resources

available. People or materials form queues until they are served in the case of scarce resources. The behavior is represented using rules and logical relationships between the components of the system. The main outcome of the analysis is a probability distribution of selected parameters

Robinson (

2004

)

Agent‐based simulation

This method focuses on the decision‐making processes of individual entities and their emergent behavior over time. It usually involves evaluating the emerging behaviors resulting from the interaction of multiple individual entities. The focus is on understanding how individuals behaving in a certain manner and interacting with other individuals generate macro‐level behavior

Macal and North (2005)

Simulation models can be used in different modes. The evaluation of their use is discussed in Pidd (2009) and it is based on Lane's folding star (Lane, 1995). For example, ardent modeling consists of the development of a formal mathematical model, which is able to run simulations, in order to develop a set of recommendations for changes in the real world. In qualitative modeling, the development of the model provides common language for managers or stakeholders to discuss their views of the system/problem under consideration. The model is mainly used to facilitate interpretation of the problem so there is no need to generate equations and gather data to run simulations. The main objective is to appreciate the situation. When the model use is discursive modeling, a formal mathematical model is developed with the objective to support training through experimentation. Its aim is to foster learning and appreciation of the complexity involved in certain situations.The model acts as a practice or training ground for managers (Lane, 1995). The model is generated to convey important theoretical elements, e.g. feedback processes, and includes guidelines to structure user’s experience (Lane, 1995). However, the interaction between the model and the users cannot be completely structured so interpretation of the situation will also arise. Finally, theoretical modeling implies the development of a formal mathematical model, which does not represent a specific problem or situation, to offer policy insights and communicate the conceptualization of a problem in general terms without specific users and with the intention to raise awareness.

Optimization methods develop models to find the “best” or “optimal” solution from a set of possible solutions. These models have generally two elements: the objective function, which represents the set of decision variables whose optimal value needs to be found; and a set of constraints, which limit the values that the decision variables can take. There are two major types of models: deterministic; and stochastic (Table 1.6).

Table 1.6 Quantitative methods to optimize in Management Science.

Method

Brief description

Suggested further reading

Deterministic optimization

There are different techniques under this method: linear programming

(single objective function, linear combination of decision variables and the decision variables can have any value

except non‐negative); integer and mixed programming (the only difference with linear programming is that all or some of the decision variables can only have integer values); non‐linear programming

(the objective function and the constraints are related non‐linearly); goal programming

(there are multiple competing objectives but if the objectives can be weighted, then the problem will be presented as single‐objective linear programming); and dynamic programming

(the problem is divided into stages with a decision at each stage subject to the state to move in the next stage and the objective function has to satisfy a recursive requirement such as cost or distance)

Winston and Goldberg (

2004

)

Stochastic optimization

It is the process of maximizing or minimizing a mathematical or statistical function when one or more of the input parameters are subject to randomness. Stochastic processes always involve probability The objective is to maximize the function's output value

in the face of numerous random input variables

Heyman and Sobel (

2003

)

Given strategic management processes are complex and uncertain, it is important to explore the consequences of strategic decisions and plans before implementing them. Often the process of exploring the consequences is performed intuitively by highly experienced managers based on their long experience. While it is definitively important to have experience, as well as deep knowledge, in crafting strategy (Pidd, 2009), managers also have limited rationality and are subject to biases and heuristics. Biases and heuristics are useful in detecting patterns that resemble previous experience but strategies tend to be forward looking and uncertain, so previous experience may hinder rather than help managers.

Consequently, the role of external and explicit models, as the main outcome of management science processes (an integral discipline in Analytics) is to provide the platform to capture the critical aspect of the issues the decisions and plans are addressing. Management science models will not replace strategic decision makers but they can be valuable tools to direct their thinking and articulate existing knowledge. Models can account for conditions where decision makers are unsure about the result of a choice, so a formal model can provide optimized options for the set of choices under consideration (prescriptive analytics). In other circumstances, the model may help to understand the current situation to identify opportunities or threats using either existing big data or eliciting assumptions (descriptive analytics). Models are also able to present evidence of future issues when certain strategic decisions are implemented (predictive analytics). In other words, models facilitate “procedural rationality” (Pidd, 2009). Thus, models are concerned with the nature of strategy processes (search, design and evaluation) rather than the outcome of the strategy process (e.g. differentiation or low costs, diversification or integration). In that way, models and modeling can be considered systematic procedures to support decision making of bounded rational decision makers (Pidd, 2009).

Strategy Analytics requires formal, illustrative and metaphorical models as they are the most suitable for the issues faced during the strategy process. Realistic management science models involving more operational detail, big data and process dimensions can become a competitive advantage. This is usually the realm of strategic operational research (Bell, 1998