Big Data Maturity Models

Big Data Maturity Models (BDMMs) are the next step in creating advanced analytics, which is aimed at better understanding its importance for the development of an organization. The existing BDMMs are largely a continuation and improvement of maturity models developed for the needs of BI. Many of them relate to the assessment and development of processes, analytical capabilities, and their role in the development of new business models, achieving organizational success, and creating a competitive advantage.

According to Braun (2015), BDMMs are the artifacts used to measure BD maturity. These models aid organizations in creating structure around their BD capabilities and in identifying where to start (Halper & Krishnan, 2014). They provide tools that assist organizations in defining goals around their BD program and in communicating their BD vision to an entire organization. BDMMs also provide a methodology to measure and monitor the state of firm’s BD capability, the effort required to complete the current stage or phase of maturity and to progress to the next stage. Additionally, BDMMs measure and manage the speed of both the progress and adoption of big data programs in an organization. The goals of BDMMs are (1) to provide a capability assessment tool that generates specific focus on BD in key organizational areas; (2) to help guide development milestones; and (3) to avoid pitfalls in establishing and building BD capabilities.

De Bruin et al. (2005) classify BDMM’s as (1) descriptive—assess the “as-is” situation; (2) prescriptive—focus on relationships to business performance and how maturity improvement can positively affect business value; and lastly (3) comparative-enable benchmarking across industries or regions. Mettler and Rosner (2009) identify three factors or aspects in a maturity model, namely, (1) process maturity - to which extent a specific process is explicitly defined, managed, measured, controlled, and effective; (2) object maturity - to which extent a particular object reaches a predefined level of sophistication; and (3) people capability - to which extent the workforce is able to enable knowledge creation and enhance proficiency.

The research literature provides many examples and descriptions of BD maturity models. The most frequently cited and used ones are presented in the following.

The TDWI Big Data Maturity Model

One of the most popular models to measure the maturity level of an organization in the context of BD analytics has been proposed by The Data Warehouse Institute (TDWI). The TDWI BDMM describes the steps that every organization has to follow while undertaking BD initiatives. The model shows how to transform an organization to fully utilize BD. The model consists of five stages (Halper & Krishnan, 2013—2014): (1) nascent, (2) preadoption, (3) early adoption, (4) corporate adoption, and (5) mature/visionary. Each stage is assessed from the perspective of five dimensions: organization, infrastructure, data management, analytics, and governance.

The nascent stage means that an organization has a low awareness of BD or their value. There is little to no executive support for the effort, and only some people in an organization are interested in the potential value of BD. During the preadoption stage, an organization starts taking the first initiatives concerning BD analytics. An organization may have invested in some new technologies, such as Hadoop, in support of BD.

The next stage, called early adoption, is characterized by one or two proofs of concept (POCs) which become more established and production ready. Organizations tend to spend a long time in this stage, often because it is hard to bridge the gap that leads to corporate-wide adoption of BD and BD analytics. Then, corporate adoption stage is the major crossover phase in any organization’s BD journey. During corporate adoption, end-users might have started operationalizing BD analytics or changing their decision-making processes. Most organizations would have repeatedly addressed certain gaps in their infrastructure, data management, governance, and analytics. Halper and Krishnan (2013—2014) stress that only a few organizations can be considered as visionary in terms of BD and BD analytics. During this stage organizations are able to execute BD programs and to use highly mature infrastructure. They also have BD governance strategies that are supported by executives. Employees share a level of excitement and energy around BD and BD analytics.

Big Data and Analytics Maturity Model (IBM model)

IBM Institute of Business Value (IBV) proposed Big Data and Analytics Maturity Model (Not, 2015). The model comprises six categories: business strategy, information, analytics, culture and operational execution, architecture, and governance. Each category is described by five levels of maturity: ad hoc, foundational, competitive, differentiating, and breakaway.

According to Not (2015), the first step with any advanced technology capability is to recognize that its use needs to be business-driven. While underpinning technology is needed to acquire data and execute analytics, business expertise is necessary to derive meaningful insight and use it to differentiate outcomes. Differentiation can be achieved by enriching customer engagement and driving operational improvements. It demands the organizational capacity to explore data for new opportunities and an ability to construct quantified business cases. Mature organizations are able to harness available data and apply analytics to it to innovate and create new business models.

The second category in the mentioned model is information. The utilization of data to manage the business is the base capability. However, highly mature organizations recognize that data are a first-class, strategic business asset. Not only is this view based on existing transactional systems - the systems of record - but also on systems that support individual— the systems of engagement - and external data sources. Furthermore, mature organizations provide governed access to data wherever they reside in an organization and are able to provide them with the meaning and context.

Not (2015) stresses that analytics is necessary to understand why something has happened or to predict what is likely to happen. Mature use of analytics optimizes the business. Organizations will already be reporting to show their financial performance and to demonstrate regulatory compliance, but analytics is necessary to understand why something has happened or to predict what is likely to happen. The resulting insight aids the improvement of customer engagement and operational efficiency. Analytics is used to make data-driven decision-making pervasive in an organization, and it requires timely insight into context.

Culture and operational execution reflect the need to engage staff, managers, and top executives in analytics, learning, and sharing knowledge in an organization. Access to data and use of analytics to derive insight builds no business value in and of itself. An organization realizes benefits when its people and systems have a desire to seek out and make use of insight as it operates. Trust in insight is essential, as is an ability to visualize, share, and provide feedback to learn and improve. A mature organization can offer rich data and analytics services that are aligned to and evolve with business priorities.

Advanced analytics based on BD requires coherent architecture and technology. It enables ease of access by end-users, agility in the capabilities required to address current business needs, and a managed approach to accessing required data.

The last category — governance — stresses that information governance is a CSF for BD projects. Policies need to be established and enforced to a degree of confidence in information, and so that resulting insights are understood and reflected in decision-making efforts. Policies also need to include data provenance and accuracy, data quality, master data and metadata, lifecycle management, security, privacy, and ethical use.

Dell Data Maturity Model

The Dell Data Maturity Model defines four levels of maturity (Dell, 2016): data aware, data proficient, data savvy, and data driven. The first level of maturity is referred to as data aware. Firms at this level are aware of the importance of the data they have and of the fact that they can gain immediate benefits. Flowever, most of the work associated with the creation of reports and the acquisition of data from various systems and information sources is done manually.

At the second level, data proficient are firms that despite the incomplete integration of all data, sources have already achieved a fairly high level of standardization in reporting. Such firms are more aware of how data are collected and stored. They seek to standardize reports and automate the measurement of key indicators for a given area.

Data savvy is the third level of maturity in the approach to data. Firms at this level experience real business effects resulting from the utilization of BD. This has a visible impact on their operational activities as well as results in noticeable savings and acceleration of the decision-making process. It is argued that it is only at this stage that information held by an organization begins to be used to achieve a real competitive advantage.

The full utilization of collected data occurs in organizations at the fourth level, known as data driven. These are organizations in which relevant indicators have been defined for all key business processes and are analyzed on an ongoing basis. The information obtained is used to analyze trends and optimize, predict, and create new values. These organizations are well aware that BD enable them not only to measure but above all to analyze and draw conclusions, because such analytics makes it easier to run the business and increase the competitive advantage.

Big Data Business Model Maturity Index

Schmarzo (2013) proposed the Big Data Business Model Maturity Index as a tool for assessing the maturity of a business model in the context of BD utilization. As Schmarzo claims, organizations may use this index to (1) get information on the stage of utilizing advanced BD analytics, on their value creation process, and on business models - that is, get information on current state and (2) identify desired target state.

Schmarzo’s model is composed of five levels (stages): business monitoring, business insights, business optimization, data monetization, and business metamorphosis. Tlie first three levels focus on the organization itself and on optimizing its internal business processes. The last two levels focus on organization’s environment - its customers and markets.

The Hortonworks Big Data Maturity Model

The Hortonworks Big Data Maturity Model assesses organization’s BD capabilities across five domains: (1) data sponsorship, (2) data and analytics practices, (3) technology and infrastructure, (4) organization and skills, and (5) process management. Each domain is described by four different components. Data sponsorship domain refers to vision and strategy, funding, advocacy, and business case. The second domain concerns data collection, storage, processing, and analysis. Domain of technology and infrastructure is assessed by hosting, functionality, tools, and integrations. The next domain is characterized by analytic and development skills, talent strategy, leadership, and collaboration models. The last domain, called process management, is assessed by planning and budget, operations, security and governance, program measurement, and investment focus.