Management of Data Value Chains, a Value Monitoring Capability

Maturity Model

Rob Brennan

, Judie Attard

and Markus Helfert

ADAPT Centre, School of Computer Science and Statistics, O’Reilly Institute, Trinity College Dublin, Dublin 2, Ireland

Lero Centre, ADAPT Centre, School of Computer Applications, Dublin City University, Dublin 9, Ireland

Keywords: Data Value, Business-IT Alignment, Business Value of IT, Governance of IS Organisation.

Abstract: This paper identifies management capabilities for data value chains as a gap in current data value research.

It specifies a data value management capability framework and a first data value monitoring capability

maturity model (CMM). This framework and CMM will enable organisations to identify and measure the

current state of their data value monitoring processes, and show how to take steps to enhance value

monitoring in order to exploit the full data value potential in their organisation. This new approach to data

value management is needed since, despite the success of Big Data and the appeal of the data-driven

enterprise, there is little evidence-based guidance for maximising data value creation. To date, most data

value optimisation has focused on technological gains such as data platforms or analytics, without bridging

the gap to organisational knowledge or human factors research. The evidence of best practice gathered here

from the state of the art shows that there is a hierarchy of data value dimensions for data value monitoring,

starting with cost and peaking with utility (understanding value creation). The models are validated by a

case study of three organisations that are managing data value and using it to support strategic decision-

making.

1 INTRODUCTION

With the advent of Big Data and the trend of

maintaining ever larger datasets in the hope of

supporting value creation, the concept of data value

is gaining increasing traction (The Economist,

2017). Despite the uptake in data use across all

sectors of society, there is a severe lack of guidance

for data value management for effective data

exploitation. The data value chains depicted in the

literature hark back to manufacturing value chains

(Crié & Micheaux, 2006; Latif, Us Saeed, Hoefler,

Stocker, & Wagner, 2009; Lee & Yang, 2000; Miller

& Mork, 2013; Peppard & Rylander, 2006), and

focus on the acquisition, manipulation, exploitation,

and distribution of data. These manufacturing-

inspired data value chains cater for the use of

intangible data as a product, but they do not cater for

the vital aspect of managing the value creating

processes within the chain (Rayport & Sviokla,

1995). It has been observed (Otto, 2015) that

measures for managing data as a strategic resource

have focused on technological endeavours such as

data architecture or analytics, but topics such as

providing evidence for the effectiveness of

organisational measures and practices have remained

relatively unexplored.

Effective data value chain management, and

hence optimised value creation, depends on an

understanding of the context of use, the value

creation process, data value measures, and hence the

nature of data value. However there is a lack of

understanding of how data value dimensions

combine into data valuations (Viscusi & Batini,

2014) or contribute to undoubtedly complex data

value creation processes (Moody & Walsh, 1999).

Nonetheless data value assessment metrics are

available in the literature (Higson & Waltho, 2010)

and have been used for specific applications. End-to-

end value chain monitoring of data value is an

important step towards data value management

processes.

This paper explores the research question “What

capabilities are required for enterprises to best

support data value management processes?” To

address this question, we bring together the diverse

experiences of leading practitioners in data value

management to identify the key capabilities and

Brennan, R., Attard, J. and Helfert, M.

Management of Data Value Chains, a Value Monitoring Capability Maturity Model.

DOI: 10.5220/0006684805730584

In Proceedings of the 20th International Conference on Enterprise Information Systems (ICEIS 2018), pages 573-584

ISBN: 978-989-758-298-1

573

processes for data value maximisation, and arrange

them in a new capability model. The data value

monitoring and assessment capability is then

examined in detail, as it acts as a basis for many

subsequent data value management capabilities.

Then, a monitoring capability maturity model with

associated data value metrics is specified. The data

value monitoring capability model is then evaluated

in a case study of three data-centric organisations (a

public knowledge base publisher, a data-driven

retailer, and an open government data initiative) that

use data valuations to drive operational performance

and strategic decision-making.

The contributions of this paper are; the

description of a novel data value chain management

capability framework, the specification of a

capability maturity model for data value monitoring

(including identification of practices, outcomes,

value dimensions and metrics for each maturity

level), and evidence supporting this model provided

by a synthesis of the state of the art and the case

study analysis.

The rest of this paper is structured as follows:

section 2 presents background work on data value

chains and capability maturity models, section 3

provides related work and defines management

capabilities for data value chains and the monitoring

capability maturity model, section 4 provides a case

study of data value monitoring in practice and

section 5 provides a concluding discussion and

future work.

2 BACKGROUND

In this section we first discuss value chains in the

context of data value and then the application of

capability maturity models to managing value.

Value chains have been used for decades,

particularly in the manufacturing industry, to

identify what manufacturer activities gave additional

value to the product being created, and therefore

resulting in a competitive advantage. The term

“Value Chain” was introduced by Porter, where he

defined a value chain to be the strategically relevant

interdependent activities undertaken by a firm in

order to achieve its goal (Porter, 1985). Porter lists

five main activities in his value chain, namely (i)

inbound logistics, (ii) operations, (iii) outbound

logistics, (iv) marketing and sales, and (v) service.

Therefore, the value chain was an indispensable tool

that enabled the analysis of the interactions between

the different activities in order to identify where

value was being created. However, despite the

success of the value chain concept for this aim, in

recent years products and services have become

increasingly digital, and therefore intangible. For

this reason, a number of authors have provided

newer definitions of the value chain concept whilst

catering for this digital dimension. This revised data

value chain provided stakeholders with a new

perspective on creating value on a digital, intangible

product. For example, Lee and Yang define a value

chain for knowledge (Lee & Yang, 2000), Crié and

Micheaux provide a more generic value chain that

considers raw data (Crié & Micheaux, 2006), Miller

and Mork define a data value chain with a big data

perspective (Miller & Mork, 2013), Latif et al. target

value creation on Linked Data (Latif, Us Saeed,

Hoefler, Stocker, & Wagner, 2009), The Big Data

Value Association (BDVA) provides a data value

chain specifically for Big Data (see Figure 1), and

Attard et al. define a data value network having a

structure that caters for the fluid nature of data

(Attard, Orlandi, & Auer, 2017).

Figure 1: BDVA Big Data Value Chain adapted from

(Zillner, S., Curry, E., Metzger and A., Auer, S., 2017).

As useful as all the aforementioned data value

chains are, none of them consider the management

of value creating processes or the capabilities

required to achieve this. Whilst some publications

(Crié & Micheaux, 2006; Lee & Yang, 2000)

consider the management of the data product itself,

the literature on data value chains described here

does not specify data value management capabilities.

In order to exploit the highest value of data, and

obtain the highest impact, we deem the data value

management capabilities of any entity to be vital in

any data exploitation endeavour.

Unlike current (data) value chain models,

Capability Maturity Models (CMMs) are specifically

created to identify how an organisation manages its

processes with the aim of becoming more successful

in achieving its goals. This is usually done by

assessing the current situation of an organisation,

deriving and prioritising improvement measures, and

controlling their implementation progress (Iversen,

Nielsen, Nerbjerg, & Norbjerg, 1999). Thus CMMs

guide organisations in the definition of progressive,

qualitative self-assessments that identify areas of

gradual improvement.

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

574

Humphrey defined one of the first CMMs that

covered practices for planning, engineering, and

managing software development and maintenance

(Humphrey, 1989). This CMM consists of five

levels of software process maturity, namely:

1. Initial Level - the organisation typically does

not provide a stable environment for

developing and maintaining software;

2. Repeatable Level - policies for managing a

software project and procedures to implement

those policies are established;

3. Defined Level - the standard process for

developing and maintaining software across

the organisation is documented;

4. Managed Level - the organisation sets

quantitative quality goals for both software

products and processes; and

5. Optimising Level - the entire organisation is

focused on continuous process improvement.

Since Humphrey’s CMM, there have been many

developments on the topic. Maturity models have

become more popular in Information Systems and

also cater for the maturity analysis of various

domains. For example, de Bruin and Rosemann

discuss a maturity model that can assist

organisations to improve in business process

management (De Bruin & Rosemann, 2005),

Gottshalk explores the digital government domain

(Gottschalk, 2008), Paulk focuses on software

engineering (Paulk, 2002), Ali et al cover inter-

organisational systems (Ali, Kurnia, & Johnston,

2008), de Bruin et al. discuss knowledge

management (de Bruin et al., 2005), Lacerda discuss

usability engineering (Lacerda, Gresse, &

Wangenheim, 2017), etc.

The mentioned CMM literature provides varying

approaches for analysing the current maturity of an

organisation, product, service or initiative, and

guidance on how to improve. Despite this, these

maturity models do not focus on the capabilities

required to manage data value explicitly. The

emerging recognition of data as an organisational

asset that spans many products or services must

change this. The lack of data value-oriented CMMs

may be because: (i) data value is a multi-

dimensional quantity that is highly context

dependent (Viscusi & Batini, 2014) and (ii) data

value creation processes are still not well

understood. Another dependency of CMM

approaches is the need for the documentation of best

practice in terms of practices, outcomes and metrics

in order to formulate an evidence-based maturity

model and, before the work presented here, no-one

has collected this evidence.

Thus, there is a need for identification of the

required data value chain management capabilities

to maximise value creation from data, and a

capability maturity model to assess organisational

readiness and guide organisational development.

3 DATA VALUE CHAIN

MANAGEMENT

CAPABILITIES

Defining data value chain management depends on

examining how value grows from organisational

capabilities, identifying a set of data value

management capabilities and specifying capability

maturity models for them. Due to space limits we

only specify a full CMM for the fundamental data

value monitoring capability as an example of our

approach.

3.1 Organisational Capabilities

Organisational capabilities focus on internal

processes, functions, and systems to meet customer

needs. Thus, organisational capabilities foster unique

service-specific routines (Felin, Foss, Heimeriks, &

Madsen, 2012) which enable competencies to

harness competitive advantage and are typically

directed towards achieving defined goals and

strategies (Winter, 2000). There are numerous

definitions throughout management literature which

examines the multidimensional nature of

organisational capabilities. Collis (Collis, 1994)

classifies many of the definitions into three broad

categories:

 Performance (capabilities): capabilities that

reflect the ability to perform the basic functional

activities of the organisation. This suggests that

capabilities are developed in functional areas

which have defined business processes;

 Repeated processes (capabilities): capabilities

are repeated processes that are responsive to

market trends and short lifecycles. This suggests

that an organisation must be agile to adapt to

customer needs by refocusing organisational

capability deployment;

 Strategise value creation (capabilities):

capabilities that realise the value of resource

allocation to execute and enable novel strategies

by deploying organisational resources.

Management of Data Value Chains, a Value Monitoring Capability Maturity Model

575

Thus, from a technological viewpoint, we can view

capabilities as being socially embedded routines

which may be captured through technological

means, i.e. automated processes. They support the

transformation of inputs into outputs. These outputs

are the product of each individual process that are

networked together to contribute to the entire service

system.

3.1.1 Routines

An organisational capability may be described as “a

high level routine (or collection of routines) that,

together with its implementing input flows, confers

upon an organisation’s management a set of decision

options for producing significant outputs of a

particular type” (Winter, 2000). It is interesting that

Winter identifies ‘routine’ as the contributing factor

of capabilities since it conjures a notion of learned

behaviour which follows a specific execution pattern

that is repetitive in nature. Routines are relatively

fixed, unchanging objects which reflect an

agreement of how things are done, i.e. imposing a

control mechanism through repetitive patterns

(Feldman & Pentland, 2003). Thus, routines encode

organisational capabilities and knowledge in a

learning cycle (see Figure 2).

3.1.2 Skills and Dynamic Capabilities

Chandler defines organisational capabilities as an

organisation’s collective physical facilities and skills

of employees, and the expertise of top management

layers (Chandler, 1994). Thus, capabilities provide a

building block towards organisational core

competencies (Coulter, 2002), for example, research

and development. In addition, considering the

concept of ‘dynamic capabilities’, Collis suggests

that dynamic capabilities govern the rate of change

of ordinary capabilities (Collis, 1994). Dynamic

Figure 2: Data Value Management Capabilities (Adapted

from (Gupta & George, 2016)).

capabilities are traditionally believed to involve

patterned activities which originate from objectives.

According to (Winter, 2003) (p.2) dynamic

capabilities “are those that operate to extend, modify

or create ordinary capabilities…[and] involves a

patterning of activity”.

We define capabilities as a partial representation

of the collective ability to carry out specific business

processes across a network in a cyclical, efficient,

and relatively predictable manner to contribute

towards organisational performance. Figure 2

describes the three key building blocks that form the

basis for performance, repeated processes and

strategic value creation for data value management.

There have been some efforts to understand the

role of IT in value creation. For example, transaction

cost theory (Clemons & Row, 1991) is one approach

to examine how technology reduces transaction

costs. Thus, resources are central to the

sustainability of competitive advantage. This is true

for a resource-based view of IT business value.

(Melville, Carroll, Kraemer, & Gurbaxani, 2004)

propose that “IT and non-IT resources and the

business processes of electronically connected

trading partners shape the focal firm's ability to

generate and capture organisational performance”

(Melville Carroll, Kraemer, & Gurbaxani., 2004) (p.

307). IT business value research examines how

organisational performance results from IT

investment and to what extent the application of IT

leads to improved performance. By performance,

scholars have referred to productivity enhancement,

profitability improvement, cost reduction,

competitive advantage, inventory reduction, and

other measures of performance (for example, (Hitt &

Brynjolfsson, 1996). Nowadays, the emphasis often

lays on optimising internal routines and capabilities

though individual processes. In most cases,

managers opt for CMM and IT Service Management

(ITSM) as traditional approaches to evaluate the

utilisation and alignment of IT resources in service

optimisation.

It is well established that value can flow from

data and that technology solutions can enhance the

value creation process. In addition if an organisation

has effective IT services to leverage data and

technology then it has a multiplier effect on value

creation. Most of the work to date has focused on

optimising the performance of these data production

capabilities (Otto, 2015).

3.1.3 Data Value Management Capabilities

We must identify additional data value management

capabilities that will span the data value chain and

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

576

optimise value creation. These capabilities must be

capable of: (1) driving data value performance-

related capabilities such as data integration by

dynamically re-configuring tools and workflows to

optimise value creation (an enterprise objective)

rather than to optimise correctness or speed

(technical objectives); (2) define and promulgate

effective repeated data value processes, work

practices and routines throughout the organisation,

especially in ways that support automation; and (3)

will support a strategic control loop for data value

and thus enable new product innovation and

exploitation based on data value management.

In order to support these requirements for data

value chain management we propose that the

following data value management capabilities are

required across the length of the data value chain,

split into performance, processes, and strategy

groupings (Figure 3). In addition, we identify a

minimal set of capabilities that will be built upon by

the other management capabilities. These are termed

fundamental data value management capabilities:

data value monitoring, profiling and prediction. It

must be observed that it is expected that more

capabilities will be identified in future work. A brief

definition of each capability is given next.

The fundamental capabilities are as follows:

 Data value monitoring capability – supports

data value assessment and reporting at all

stages in the data value chain;

 Data value profiling capability – the ability

to specify sources of value for a specific

organisation;

 Data value prediction capability – capacity

to analyse and predict data value trends or

sources from current indicators, patterns and

context.

The performance capabilities for data value

management are:

Figure 3: Data Value Management Capability Framework.

 Dataset optimisation capability – ability to

acquire, curate, maintain and enhance datasets

based on value maximisation;

 Infrastructure optimisation capability –

ability to design, deploy and reconfigure the

enterprise data infrastructure to maximise data

value;

 IT services optimisation capability – the

ability to evolve, adapt and orchestrate the

data processes, support services and

workflows towards value-based goals and

outcomes.

The processes and human factors oriented capability

for data value management are:

 Data value communication capability –

level of integration of data value into IT

leadership and governance infrastructures;

 Data value training capability – the ability

to enhance organisational management and

technical skills about data value issues;

 Data value process definition capability –

value-centric data process design and

integration of data value considerations into

wider organisational processes.

The strategy capabilities are:

 Data innovation capability – the ability to

innovate in services or products based on

novel use of (or insights from) data;

 Decision-making capability – organisational

capacity to create a data value-based control

loop that feeds into organisational level

decisions;

 Organisational learning capability – the

extent to which the organisation can evolve

based on data value considerations;

 Strategic data planning capability –

identification of longer-term data acquisition,

data architecture, system integration, IS

priorities, business process and people

required;

 Strategic change capability – identification

of the changes in organisational goals based

on data value.

In the next section we provide a capability maturity

model for the data value monitoring capability as an

example of a fundamental capability that must be

specified in order to realise our vision.

3.2 Data Value Monitoring Capability

The data value monitoring capability focuses on

assessing and reporting of data value throughout the

value chain by gathering metrics on datasets, the

Management of Data Value Chains, a Value Monitoring Capability Maturity Model

577

data infrastructure, data users, costs and operational

processes. Monitoring forms a fundamental function

of any control capability whether embodied as the

fault management or performance management

functions in a telecommunications network (3GPP,

2009) or a MAPE (monitor, assess, plan, execute)

control loop in IBM’s autonomic computing

initiative (Jennings et al., 2007). Monitoring data

value is essential to providing feedback to any data

value control system, whether automated, semi-

automated or manual.

There have been a number of documented uses

of data value monitoring for enhanced control of

elements of the data value chain, especially in the

application areas of file-storage management

(Turczyk, Heckmann, & Steinmetz, 2007),

information lifecycle management (Chen, 2005),

information pricing (Rao & Ng, 2016), data

governance (Stander, 2015), and data quality

management (Even, Shankaranarayanan, & Berger,

2010). We use those and other examples from

practice as the basis for our approach.

This capability aims to be holistic in assessing

the key dimensions of data value for a particular

organisation. Unfortunately, there are a wide range

of known dimensions of data value (Viscusi &

Batini, 2014) and there is not yet a consensus on

their definitions, how they are related, or how

practical data value metrics in information systems

relate to monetary value, as measured in accounting-

based measures of value. Viscusi and Batini break

data value down into information capacity and

information utility (Viscusi & Batini, 2014).

Capacity is then subdivided into quality, structure,

diffusion and infrastructure. In their scheme, utility

is based on financial value, pertinence and

transaction costs. In contrast, the models of (Moody

& Walsh, 1999) and (Tallon, 2013) strongly

emphasise usage as a key dimension of value. It is in

the area of usage-based data value that the most

progress has been made for practical data value

monitoring systems. Hence, we must give it most

prominence in the following analysis. Given a lack

of a comprehensive metric framework for measuring

data value, it has been found that having at least

some data value measurement capability is better

than none.

Ease of measurement is another important

concept to consider when deciding how to monitor

data value in an organisation. Some data value

dimensions have well known metrics and may even

have recommended data or metadata formats, for

example the W3C’s data quality vocabulary

(Albertoni et al., 2016). Other data value

dimensions, such as business utility or impact are

very difficult to measure since they depend on

having models and information about the business

processes, outcomes and data dependencies in order

to identify measurable metrics for the contribution of

datasets to profit or operating efficiencies.

One important distinction to be made when creating

a monitoring framework is the identification of data

value metrics that can be evaluated independently of

any knowledge of the context of use of the dataset,

for example frequency of access, dataset volume or

dataset purchase cost, and context-dependent metrics

such as appropriate data quality metric thresholds or

the relevance of a particular dataset for a specific

task. We label this latter class of metrics as

“contextual”.

Table 1 defines a capability maturity model for

data value monitoring based on our analysis of the

existing literature for data value monitoring or

metrics discussed above. As can be seen from the

table, lower levels of maturity are associated with

more limited collection of data value dimensions and

there is a hierarchy of monitoring dimensions that

emerges from the literature on data value metrics.

The second major trend seen in the table is the

increasing connection between technology-centric

metrics such as data volume, and enterprise-centric

metrics such as average financial contribution per

record. Ultimately, we posit that attaining a level

five capability maturity level is dependent on being

able to understand or model how data value chain

processes or steps convert data into value (utility).

This corresponds to understanding the value creation

process in a business, at least to the extent of a

reliable black box predictor function.

One challenge in tackling this level of business

monitoring is the need to converge the views of the

technology-centric data administration functions of

the business, with the financial and operations

functions (and ultimately the strategic planning

functions). This approach has a natural foundation in

the increasing importance of data governance within

overall governance structures (Tallon, 2013).

4 CASE STUDIES

4.1 Aims and Methodology

The case study comprises three organisations. The

three cases represent four different industries, with

headquarters located in three different countries and

a scope of operations ranging from a single country,

to a handful of through worldwide. For an

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

578

Table 1: A Capability Maturity Model for Data Value Monitoring including practices, outcomes, and metrics pertaining to

data value monitoring.

organisation to be included in the case study, three

criteria had to be met: (1) the organisation should

have experience in managing data value chains and

data value monitoring, (2) the organisation’s data

value management concept should cover multiple

data value monitoring dimensions, and (3) the

organisations should allow the authors of the article

to get in touch with subject matter experts.

For each organisation, the data value monitoring

CMM (Table 1) was applied to assess their current

processes. This involved investigating their

methodology and assigning their metrics to specific

data value dimensions. For data collection, various

methods and data sources were used, ranging from

structured questioning of subject matter experts to

the analysis of publicly available material (annual

reports, for example), research papers, presentations,

and internal documents from joint research projects

that focused on data quality, data management,

governance and data process improvement for

increased agility and productivity. It must be

observed that the diversity of the cases selected

make the use of “natural controls” as often required

for case study research (Lee, 1989) difficult to

achieve, and this is left for future work.

4.2 Case Studies

4.2.1 DBpedia - Public Data Infrastructure

for a Large, Multilingual, Semantic

Knowledge Graph

DBpedia (Lehmann et al., 2015) is the centre of the

current web of data. It extracts, integrates, manages

and publishes authoritative open RDF-based datasets

that are used as a common point of reference for

interlinking and enriching most of the structured

data on the web today. It relies on an automated data

Maturity

Level

Practices Outcomes Value

Monitoring

Dimensions

Example Metrics

1 No formal practices are

expected at this level.

Potential for collection of

pertinent metrics for non-

value reasons. No

coherent analysis

erformed.

Inconsistent approach to

data value monitoring. Data

value monitoring is not

formalised. Ad hoc local

assessments based on

perceived business unit

needs may exist.

No value-

specific

monitoring

No metrics

2 Only perform reactive

data value assessment

based on crisis events and

execute reviews of crisis

handlin

as needed.

Can establish links between

service or product failures

and data. Know the costs

associated with specific

datasets.

Outage events,

Cost

Time since last file access

(Chen, 2005), Cost of

procurement (Moody &

Walsh, 1999)

3 Active monitoring of data

value using a mix of

context-independent

measures.

Know who is using what

data and for what purposes.

Have defined a set of

metrics for value

assessment. Can assess a

value for each data

resource.

Usage,

Quality

metrics, IT

Operations,

Intrinsic

Data volume (Short, 2014),

storage cost (Tallon, 2013),

completeness (Laney, 2011),

Trustworthiness (Viscusi &

Batini, 2014)

4 Characterise the links

between data, value,

business operations and

revenue. Specify

thresholds for business

context-dependent

dimensions of data value.

Can attribute specific

process or product

outcomes to a specific data

resource. Can distinguish

between data resources that

are fit for business purposes

and those that are not.

Contextual,

Value

contribution

Business user satisfaction

(Higson & Waltho, 2010),

Shelf life of data (Moody &

Walsh, 1999), Data quality

thresholds, intrinsic record

value (Even &

Shankaranarayanan, 2005)

5 Monitoring supports the

optimised exploitation of

data value. Monitoring

feedback is provided to

other data value

mana

ement ca

abilities.

Understand value creation

processes at each stage in

value chain. Data value

monitoring feeds back to

whole organisation and

acts strate

Utility Estimated benefit to business

from using data (Higson &

Waltho, 2010), Change in

KPIs for business objectives

(Laney, 2011), Potential lost

revenue

(

Tallon, 2013

)

Management of Data Value Chains, a Value Monitoring Capability Maturity Model

579

extraction framework to generate RDF data from

Wikipedia documents, and is now mining new

sources like WikiData, EU Open Data portals,

DBpedia association member data and the open web.

The data is published in 125 languages in the form

of file dumps, Linked Data and SPARQL (SPARQL

Protocol And RDF Query Language) endpoints.

With the help of the H2020 ALIGNED project

(Gavin et al., 2016) it has semi-automated the data

publishing process through the use of the dataset

metadata specification DataID (Freudenberg et al.,

2016) and a supporting toolchain for generation,

validation, knowledgebase staging and

dockerisation. DBpedia supports higher quality

knowledge extraction through tools for community

generated mappings, a repository of links to other

open data sets and a human-curated reference

ontology. Quality management of interlinks to other

datasets also remains a crucial problem in web-scale

data (Meehan, Kontokostas, Freudenberg, Brennan,

& O’Sullivan, 2016). Managing data value is critical

in the DBpedia Association

(http://wiki.dbpedia.org/dbpedia-association) which

is a non-profit that is responsible for overall

governance of DBpedia activities. As part of

ongoing automation, innovation support and quality

improvement initiatives in DBpedia, for example

through the ALIGNED project a number of data

value monitoring activities have been initiated.

4.2.2 MyVolts – A Data-Driven Online

Retailer

myVolts (http://www.myvolts.com/) is a data-driven

SME that relies on high quality, agile data

integration and business intelligence analytics to

operate extremely efficiently. MyVolts develops and

operates a highly automated internet retail and

business intelligence system (Frodo) that reduces

their costs and staffing. They have served over 1

million customers and are a leading source of

consumer device power supplies in the markets

where they operate: the USA, Ireland, UK, France

and Germany. MyVolts collect, manage and analyse

data on their customers, the evolving market of

power supply device specifications, and the power

supply needs of all consumer electronics devices.

This involves monitoring social media, web sales

data such as Amazon top seller lists, customer

queries and complaints, and device manufacturer

homepages. New consumer electronic devices must

be discovered, categorised, profiled for potential

sales value and have their power supply technical

specifications (voltage, polarity, tip type and

dimensions) mined from open web data. There are

an estimated 5.5 million consumer electronics

devices on sale today and the number of powered

devices is growing rapidly. MyVolts use data value

estimates to drive data acquisition, cleaning and

enrichment, and product logistics or marketing

decisions.

4.2.3 Data.gov.uk – An Open Government

Data Initiative

Data.gov.uk is the United Kingdom government’s

effort towards opening up its data to encourage its

re-use. This effort is led by the Data team in the

Cabinet Office, working across government

departments to ensure that data is released in a

timely and accessible way. The main aims behind

this effort are to enable the government to be more

transparent, to support businesses and academics, to

foster innovation, and to enable citizens to become

more informed. Data.gov.uk focuses on publishing

non-personal, non-sensitive data in a single

searchable website. Available datasets are sourced

from all central government departments, and a

number of other public-sector bodies and local

authorities. This data includes environment data,

government spending data, transport data, crime

data, health data, etc. There are currently close to 43

thousand datasets currently published on the portal.

Data.gov.uk uses data value management to sustain

the open government data portal. Decisions on

maintenance of datasets or the publishing of new

datasets are based mostly on usage and data quality

statistics.

4.3 Case Analysis

All organisations in the case study are assessing

several dimensions of the value of their data

resource (see Table 2) and seeking to apply data

value in order to perform strategic data planning.

In DBpedia data value monitoring started in

2015 as part of the H2020 ALIGNED project quality

improvement program for DBpedia processes, data

and tools. It had been found that due to the success

of DBpedia as a research platform that provided an

extensive formal knowledge base as a free

service that industrial entities such as Wolters

Kluwer Germany (Kontokostas et al., 2016) had

adopted DBpedia as a datasource. This in turn

highlighted data quality issues and the difficulty of

maintaining innovation and a stable service offering

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

580

Table 2: Data Value Monitoring Case Study Evaluations.

Monitoring

CMM

Level

Case A: DBpedia Case B: MyVolts

Case C: Open Government

Data

Usage

Log analysis Identification

of data consumers

Number of views, Number of

downloads, Publisher activity

log

Cost

Creation cost, maintenance cost Dedicated staff

costs

Publishing cost, maintenance

cost

Quality

Test driven assessment, Interlink

validation, user surveys

Number of trouble

tickets

Openness, Re-usability of data,

Broken links, Missing

resources

Intrinsic

Volume, number of classes,

number of properties, number of

Wiki

edia

es ma

Number of

records, record

rowth rate

Number of available datasets,

Operations

User surveys, trouble ticket

analysis, staff surveys

Not measured Not measured

Contextual 4 Not measure

Not measure

Utilit

5 Not measure

Not measure

Economic Value

at DBpedia

and led to the formation of the DBpedia

Association to oversee the development of DBpedia

from a research infrastructure to a public

infrastructure. The Association needed to perform

strategic planning and the data quality, data value,

and productivities studies were used as an input to

that as well as automation and process improvement

research. The baseline studies took place in May-

June 2015 and have continued on a periodic basis

since then. The data collection methods included

analysis of the DBpedia data assets in the form of

the DBpedia ontology, the mappings wiki and

quality test reports on the DBpedia data. In addition

to these direct measurements of the dataset a number

of other monitoring initiatives were setup to get a

wider view, these included: analysing GitHub

repositories for DBpedia extraction code, monitoring

Wikipedia logs (the source of DBpedia data), a staff

survey, a user survey and analysis of traffic on the

public DBpedia mailing list.

In contrast MyVolts Ltd. have only recently

started to monitor data value sources. As a small

enterprise, information naturally flows easier within

See for example DBpedia director Sebastian Hellmann’s

announcement on the public DBpedia mailing list on the 25th

Sept 2017 https://sourceforge.net/p/dbpedia/mailman/dbpedia-

discussion/?viewmonth=201709&viewday=25

their tight-knit teams and the organisation has seen

many years of increasing success despite processes

that have been less formalised. Nonetheless the high

level of automation and the use of a single integrated

IT system in their business has facilitated rapid

analysis of the sources of value. As an SME they

have in the past relied on informal but effective

qualitative staff transfer of knowledge about the

context of use and utility of data to direct their

strategic data management and ultimately

management of their data value chain. They are now

taking the first steps to automation of data value

monitoring and collecting statistics on the number of

electronic device records, the growth rate of this

data, the number of data quality-related trouble

tickets and the costs directly associated with dataset

creation. This will be combined with existing

informal knowledge on data utility and context of

use to enable data value to be estimated.

Data value monitoring was part and parcel of the

Data.gov.uk open government initiative from the

very start of the project. The initiative required the

selection of key datasets for initial publishing. These

key datasets comprised National Health Service,

Education and Skills, Criminal Justice, Transport,

and Government financial information datasets.

These datasets where considered to be key datasets

since they directly targeted the specified goals of the

Management of Data Value Chains, a Value Monitoring Capability Maturity Model

581

data.gov.uk initiative. Therefore, data value was

here used to enable more informed decision making.

Thereafter, and throughout the lifetime of the

initiative, data value monitoring was implemented to

ensure sustainability. Dataset usage (through number

of views and downloads) and dataset publishing

frequency are constantly being tracked. This enables

all stakeholders, whether publishers or consumers, to

identify popular or highly-requested datasets, and

their update rate. In turn, dataset maintenance can be

planned accordingly. Data quality is also being

monitored, both from the initiative itself, and also

from the dataset consumers. Focus is usually

stronger on openness and usability of the datasets.

The financial aspects of the initiative, whilst

monitored, are mostly estimates. In addition, the

utility of the published datasets is also estimated

through the resulting contribution to the economy.

For example, the National Audit Office estimated

that the datasets published by April 2012 could

potentially add economic value in the region of 1.6

to 6 billion Pounds a year.

5 DISCUSSIONS AND

CONCLUSIONS

With this paper we discuss management capabilities

for data value chains and argue that the monitoring

capability is fundamental. Underpinned by three

cases the paper presents a first data value monitoring

capability maturity model (CMM) and data value

metrics based on current best practice. The cases

illustrate the wide variety of ways to operationalise

the data value monitoring component, and

demonstrate capabilities along five maturity levels.

The analysis of our data value monitoring capability

model supports the general hypothesis, i.e. that there

is a hierarchy of data value dimensions when it

comes to monitoring - usage and cost are the easiest

to implement. It also shows the fundamental

character of Data Value Monitoring, Profiling and

Prediction capabilities in order to increase overall

data value management capabilities.

The presented cases in this paper demonstrate

that monitoring data value is a necessary prerequisite

to strategic data management and ultimately data

value chain management within the enterprise. Our

research, as evident for example from the

Data.gov.uk open government initiative, stresses the

importance of the Data Value Monitoring capability

even on a lower maturity level. Furthermore, the

DBpedia case highlighted data quality issues and the

difficulty of maintaining innovation, which again

emphasises the importance of Data Value

Monitoring. An example of the use of data value

estimates to drive data acquisition, cleaning and

enrichment is illustrated in the MyVolts case. We

have shown that it is possible to assess the maturity

of data value monitoring processes and use our

capability maturity model to guide enterprises on the

dimensions of data value to assess and what metrics

to use. We also validated our maturity model by

demonstrating that the hierarchy of data value

monitoring dimensions we envisage is seen in

nature.

The proposed Data Value Monitoring

component is embedded into our overarching

Capability model that provides valuable guidelines

to industry. However further research in needed on:

metrics to assess data value, how metrics can be

combined to give data value estimates, data value

monitoring infrastructure; formal models describing

metrics, dimensions and how they relate (ontologies

or data models), techniques for comparison of

assessment methods, understanding the relationship

between other CMM models to the data value CMM

proposed here. However, as our research indicates, a

specific data value oriented maturity framework

provides value, and this new value exploitation

perspective can complement the other, often output

or service oriented maturity frameworks.

ACKNOWLEDGEMENTS

This research was partially supported by Science

Foundation Ireland and co-funded by the European

Regional Development Fund through the ADAPT

Centre for Digital Content Technology [grant

number 13/RC/2106] and grant 13/RC/2094 co-

funded under the European Regional Development

Fund through the Southern & Eastern Regional

Operational Programme to Lero - the Irish Software

Research Centre (www.lero.ie)

The authors wish to thank Dr Pieter De Leenheer

of Collibra Inc. for feedback on this work.

REFERENCES

3GPP. (2009). (V8.1.0, 3GPP TS 32.111-2). Retrieved

from http://www.qtc.jp/3GPP/Specs/32111-2-810.pdf.

Albertoni, R., Isaac, A., Guéret, C., Debattista, J., Lee, D.,

Mihindukulasooriya, N., & Zaveri, A. (2016). Data

Quality Vocabulary ({DQV}). Retrieved from

http://www.w3.org/TR/vocab-dqv/

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

582

Ali, M., Kurnia, S., & Johnston, R. B. (2008). A Dyadic

Model of Interorganizational Systems (IOS) Adoption

Maturity. In Proceedings of the 41st Annual Hawaii

International Conference on System Sciences (HICSS

2008) (pp. 9–9). IEEE.

https://doi.org/10.1109/HICSS.2008.18.

Attard, J., Orlandi, F., & Auer, S. (2017). Exploiting the

value of data through data value networks. In

Proceedings of the 10th International Conference on

Theory and Practice of Electronic Governance

ICEGOV’17 (pp. 475–484).

https://doi.org/10.1145/3047273.3047299.

de Bruin, T., Rosemann, M., Freeze, R., Kulkarni, U., &

Carey, W. (2005). Understanding the Main Phases of

Developing a Maturity Assessment Model. In

Australasian Conference on Information Systems

(ACIS),.

Higson, C. & Waltho, D.. (2010). Valuing Information as

an Asset. Sas the Power To Know, (November), 1–17.

Retrieved from

http://www.eurim.org.uk/activities/ig/InformationAsse

t.pdf.

Chandler, A. D. (1994). Scale and scope : the dynamics of

industrial capitalism. Belknap Press of Harvard

University Press. Retrieved from

http://www.hup.harvard.edu/catalog.php?isbn=978067

4789951&content=reviews.

Chen, Y. (2005). Information Valuation for Information

Lifecycle Management. In Second International

Conference on Autonomic Computing (ICAC’05) (pp.

135–146). https://doi.org/10.1109/ICAC.2005.35.

Clemons, E. K., & Row, M. C. (1991). Sustaining IT

Advantage: The Role of Structural Differences. MIS

Quarterly, 15(3), 275. https://doi.org/10.2307/249639.

Collis, D. J. (1994). Research Note: How Valuable are

Organizational Capabilities? Strategic Management

Journal, 15(S1), 143–152.

https://doi.org/10.1002/smj.4250150910.

Coulter, M. K. (2002). Strategic management in action.

Prentice Hall.

Crié, D., & Micheaux, A. (2006). From customer data to

value: What is lacking in the information chain?

Journal of Database Marketing & Customer Strategy

Management, 13(4), 282–299.

https://doi.org/10.1057/palgrave.dbm.3240306.

De Bruin, T., & Rosemann, M. (2005). Towards a

business process management maturity model, 26–28.

Retrieved from

https://eprints.qut.edu.au/25194/1/25194_rosemann_2

006001488.pdf.

Economist (2017), “The world’s most valuable resource is

no longer oil, but data,” Economist, May 6, 2017.

Retrieved from

https://www.economist.com/news/leaders/21721656-

data-economy-demands-new-approach-antitrust-rules-

worlds-most-valuable-resource.

Even, A., & Shankaranarayanan, G. (2005). Value-driven

Data Quality Assessment. … Conference on

Information Quality. Retrieved from

http://mitiq.mit.edu/ICIQ/Documents/IQ Conference

2005/Papers/ValueDrivenDQAssessment.pdf.

Even, A., Shankaranarayanan, G., & Berger, P. D. (2010).

Evaluating a model for cost-effective data quality

management in a real-world CRM setting.

https://doi.org/10.1016/j.dss.2010.07.011.

Feldman, M. S., & Pentland, B. T. (2003).

Reconceptualizing Organizational Routines as a

Source of Flexibility and Change. Administrative

Science Quarterly, 48(1), 94.

https://doi.org/10.2307/3556620.

Felin, T., Foss, N. J., Heimeriks, K. H., & Madsen, T. L.

(2012). Microfoundations of Routines and

Capabilities: Individuals, Processes, and Structure.

Journal of Management Studies, 49(8), 1351–1374.

https://doi.org/10.1111/j.1467-6486.2012.01052.x.

Freudenberg, M., Brümmer, M., Rücknagel, J., Ulrich, R.,

Eckart, T., Kontokostas, D., & Hellmann, S. (2016).

The Metadata Ecosystem of DataID. In Special Track

on Metadata & Semantics for Open Repositories at

10th International Conference on Metadata and

Semantics Research, (2016). Retrieved from

http://aksw.org/Groups/KILT.

Gavin, O., Kontokostas, D., Dirschl, C., Koller, A.,

Davies, J., Francois, P., … Brennan, R. (2016). The

ALIGNED Project - Aligned, Quality-centric Software

and Data Engineering Driven by Semantics. In EU

Project Networking Session at 13th ESWC. Retrieved

from http://www.tara.tcd.ie/handle/2262/76242.

Gottschalk, P. (2008). Maturity levels for interoperability

in digital government. Government Information

Quarterly, 26, 75–81.

https://doi.org/10.1016/j.giq.2008.03.003.

Gupta, M., & George, J. F. (2016). Toward the

development of a big data analytics capability.

Information and Management, 53(8), 1049–1064.

https://doi.org/10.1016/j.im.2016.07.004.

Hitt, L. M., & Brynjolfsson, E. (1996). Productivity,

Business Profitability, and Consumer Surplus: Three

Different Measures of Information Technology Value.

MIS Quarterly, 20(2), 121.

https://doi.org/10.2307/249475.

Humphrey, W. S. (1989). Managing the software process.

Addison-Wesley. Retrieved from

https://dl.acm.org/citation.cfm?id=64795.

Iversen, J., Nielsen, P. A., Nerbjerg, J., & Norbjerg, J.

(1999). Situated assessment of problems in software

development. ACM SIGMIS Database, 30(2), 66–81.

https://doi.org/10.1145/383371.383376.

Jennings, B., van der Meer, S., Balasubramaniam, S.,

Botvich, D., O Foghlu, M., Donnelly, W., & Strassner,

J. (2007). Towards autonomic management of

communications networks. IEEE Communications

Magazine, 45(10), 112–121.

https://doi.org/10.1109/MCOM.2007.4342833.

Kontokostas, D., Mader, C., Dirschl, C., Eck, K.,

Leuthold, M., Lehmann, J., & Hellmann, S. (2016).

Semantically Enhanced Quality Assurance in the

JURION Business Use Case. In 13th International

Conference, ESWC 2016, Heraklion, Crete, Greece,

Management of Data Value Chains, a Value Monitoring Capability Maturity Model

583

May 2016 (pp. 661–676). https://doi.org/10.1007/978-

3-319-34129-3_40.

Lacerda, T. C., Gresse, C., & Wangenheim, V. (2017).

ARTICLE IN PRESS Systematic literature review of

usability capability/maturity models. Computer

Standards & Interfaces, 13(0), 17–1.

Laney, D. (2011). Infonomics : The Economics of

Information and Principles of Information Asset

Management. The Fifth MIT Information Quality

Industry Symposium, July 13-15, 2011, 590–603.

Latif, A., Us Saeed, A., Hoefler, P., Stocker, A., &

Wagner, C. (2009). The Linked Data Value Chain: A

Lightweight Model for Business Engineers. In

Proceedings of International Conference on Semantic

Systems (pp. 568–576).

Lee, A. S. (1989). A Scientific Methodology for MIS Case

Studies. MIS Quarterly, 13(1), 33.

https://doi.org/10.2307/248698.

Lee, C. C., & Yang, J. (2000). Knowledge value chain.

Journal of Management Development, 19(9), 783–

794.

Lehmann, J., Isele, R., Jakob, M., Jentzsch, A.,

Kontokostas, D., Mendes, P. N., … Bizer, C. (2015).

DBpedia – A large-scale, multilingual knowledge base

extracted from Wikipedia. Semantic Web, 6(2), 167–

195. https://doi.org/10.3233/SW-140134.

Meehan, A., Kontokostas, D., Freudenberg, M., Brennan,

R., & O’Sullivan, D. (2016). Validating Interlinks

Between Linked Data Datasets with the SUMMR

Methodology (pp. 654–672). Springer, Cham.

https://doi.org/10.1007/978-3-319-48472-3_39.

Melville, N., Carroll, W. E., Kraemer, K., & Gurbaxani,

V. (2004). Review: Information Technology And

Organizational Performance: An Integrative Model Of

It Business Value 1. MIS Quarterly, 28(2), 283–322.

Miller, H. G., & Mork, P. (2013). From Data to Decisions:

A Value Chain for Big Data. IT Professional, 15(1),

57–59. https://doi.org/10.1109/MITP.2013.11.

Moody, D., & Walsh, P. (1999). Measuring The Value Of

Information: An Asset Valuation Approach. Seventh

European Conference on Information Systems

(ECIS’99), 1–17. https://doi.org/citeulike:9316228.

Otto, B. (2015). Quality and Value of the Data Resource

in Large Enterprises. Information Systems

Management, 32(3), 234–251.

https://doi.org/10.1080/10580530.2015.1044344.

Paulk, M. (2002). Capability Maturity Model for

Software. In Encyclopedia of Software Engineering.

Hoboken, NJ, USA: John Wiley & Sons, Inc.

https://doi.org/10.1002/0471028959.sof589.

Peppard, J., & Rylander, A. (2006). From Value Chain to

Value Network: European Management Journal,

24(2–3), 128–141.

https://doi.org/10.1016/j.emj.2006.03.003.

Porter, M. E. (1985). Competitive Advantage:Creating and

sustaining superior performance. New York (Vol. 15).

https://doi.org/10.1182/blood-2005-11-4354.

Rao, D., & Ng, W. K. (2016). Information Pricing: A

Utility Based Pricing Mechanism. Proceedings - 2016

IEEE 14th International Conference on Dependable,

Autonomic and Secure Computing, DASC 2016, 2016

IEEE 14th International Conference on Pervasive

Intelligence and Computing, PICom 2016, 754–760.

Rayport, J. F., & Sviokla, J. J. (1995). Exploiting the

Virtual Value Chain. Harvard Business Review, 73,

75.

Short, J. (2014). Valuing Enterprise Data : CLDS

Research Project Brief. Retrieved from

http://www.sdsc.edu/Events/ipp_webinars/ipp_webina

r_data_value_project_feb_13_2014.html.

Stander, J. B. (2015). The Modern Asset : Big Data and

by, (December).

Tallon, P. P. (2013). Corporate governance of big data:

Perspectives on value, risk, and cost. Computer, 46(6),

32–38. https://doi.org/10.1109/MC.2013.155.

Turczyk, L. A., Heckmann, O., & Steinmetz, R. (2007).

File Valuation in Information Lifecycle Management.

Managing Worldwide Operations & Communications

with Information Technology, 347–351.

Viscusi, G., & Batini, C. (2014). Digital Information Asset

Evaluation: Characteristics and Dimensions (pp. 77–

86). Springer, Cham. https://doi.org/10.1007/978-3-

319-07040-7_9.

Winter, S. G. (2000). The Satisficing Principle in

Capability Learning. Strategic Management Journal,

21(10–11), 981–996.

Winter, S. G. (2003). Understanding dynamic capabilities.

Strategic Management Journal, 24(10), 991–995.

Zillner, S., Curry, E., Metzger, A., Auer, S. (2017),

European Big Data Value Partnership Strategic

Research and Innovation Agenda. (2017). Retrieved

from

http://www.bdva.eu/sites/default/files/EuropeanBigDa

taValuePartnership_SRIA__v3_0.pdf.

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

584