White Paper: Agile Data Management in 3 steps
Data is the oil of the digital economy. And the quantity and diversity of data just continues to grow. More and more sources of data from within and outside an organization can be accessed and integrated, analyzed and made available to the organization. As an increasing number of directors and managers want to be able to make decisions on the basis of data, the quality of that data is of great importance. This paper aims to show how robust data management can be configured in such a way that data, in whatever quantity, can become and remain a reliable resource. Clearly defined ownership and management of data pave the way to a truly data-driven organization. A way that leads to proper coordination of people, processes and technology so that data can be used efficiently and effectively. And good data management makes organizations agile.
In this paper you will read how to make your organization ready for a data-driven future in three steps using the independent DMBoK data management model. Quint Wellington Redwood has used this approach successfully many times. Two cases illustrate that.
Sustainability integrated within the Value Chainleft
Introduction: What is a data-driven organization?
The term ‘data-driven’ means that a business or institution makes decisions on the basis of data, rather than on intuition or personal experience. This is also called fact-based or evidence-based decision-making. So, it is decision-making based on facts, at all levels of the organization: business strategy, marketing, sales, finance – in all business units and departments. Businesses that have started up since the turn of the century are almost all data driven: it is, as it were, in their DNA. And they have great success with new digital products and business models that allow them to derail existing markets and threaten the established order. This is not scaremongering, it is a fact. Half of the businesses on the Fortune 500 list in 2000 are no longer trading; the number of digital, data-driven enterprises is thus growing rapidly. That underlines the importance of digital services and data-driven decision-making. It implies that existing businesses need to transform into digital enterprises if they want to remain relevant in the digital age.
Why become a data-driven enterprise?
The most significant reason for becoming a data-driven enterprise is to create a competitive advantage. General market information is the same for all providers on the market; what makes the difference is the data – for instance on customers – that an enterprise possesses. The collation of specific information, smart combinations of data and analyzing these elements intelligently with methods developed in-house sharpens the competitive edge. Analytics also makes it possible to further improve the customer experience, which in this online age really equates to a competitive advantage. Management consultants McKinsey did a study into the motivations of organizations see this link
Data is an asset
A growing number of organizations are discovering the potential of data. They play with data (and so-called ‘big data’) and realize that data-driven decision-making is of great benefit, and want to make even more use of data. They are aware that this places high demands on the quality and continuity of the data, and that it is absolutely imperative to set up their data management properly. In this endeavor, they have powerful support from the top. Or at least that’s the idea. Alas, in practice it seems that not all high-level managers realize that data is crucial to the enterprise as a whole, regardless of what people throughout the organization say or how passionately they advocate it. The plans for big data analytics also often appear to be no more than ambitions in an environment in which day-to-day business intelligence is not yet properly set up. One in which the springing of data leaks shows that data governance is a mess. And in which a data-driven approach is not on the board’s agenda.
Factors behind successful data management
Senior leadership, an organization in which the processes are in line with the use of data, and good data governance are essential conditions for implementing a data-driven approach. Data is an asset, although many organizations fail to manage it in the same way as other assets, such as people, resources, capital, IT infrastructure and applications. Only then can an asset be valuable, and remain so. Good data management is essential to organizations. Organizations with well-structured governance, in which the ownership of data sets 1, 2 and 3 is crystal clear. Organizations in which it is clear what information needs special attention due to its mission-critical, distinctive nature: so-called ‘golden data’. Organizations that employ data professionals, not ‘data enthusiasts’.
Professionally organized data governance:
- provides coherent governance company-wide
- keeps data under control
- maintains high data-quality levels
- makes data reliable and ensures compliance
- does not take a ‘scattergun’ approach to data – the real value lies in just a small portion of it
- makes it possible to get information to the business rapidly
- creates agility – when something has to be changed, that is possible. And it can happen relatively quickly.
How mature is the organization?
Once management has developed a strategy and vision on the use of data in the organization, the data revolution can commence. The primary question is whether or not the enterprise or institution is ready. Is everyone, at all levels of the organization, prepared to adopt a different working method? Is top-level management giving its unconditional backing? Are the processes by which this is to be achieved in place, and is IT ready? In other words, how mature is the organization and what steps need to be taken for data to be made into an asset? A tried and trusted method for determining this is to assume that the desired situation is up and running, and then backtrack to the current situation. This is a good way of accurately visualizing all the steps that need to be taken. Of course, that only works if the targeted to-be situation can be established accurately and independently.
Technically speaking, almost anything is possible
Storing large amounts of data is not a problem in technical terms. The problem of accessing and collating different data – from a range of sources, in different formats, and with varying levels of quality – is also one that can be resolved in technical terms. Strategic sourcing of IT services can help in that regard, for instance by outsourcing certain services and combining that with the infrastructure on site. The problem lies elsewhere. Not for nothing do we call it a data-driven organization. Becoming data driven demands different processes and working methods, different IT, and different skills and competencies. In short, a different way of thinking. At all levels of the organization. Without that revolution data will never become an asset: it will be a ball and chain. It is up to management to guide and supervise this process of culture change.
How to gain control of data, and keep on top of it
The question is how an organization can gain control of the mushrooming quantity of data and the increasing number of data streams. Who is responsible for what, who does what and who is allowed to do what? What tasks can and are being differentiated and, more to the point, is there coherence and how can coherence be safeguarded? The Data Management Body of Knowledge (DMBoK) was developed by the DAMA International community. DAMA thus offers data professionals the world over a supplier-independent best-practice resource that covers all disciplines within the domain of data management. DMBoK is well on track to become the standard in the digital age.
What is DMBoK?
The data management body of knowledge is considerable and is continually growing. To face up to this challenge, DAMA International has published the DAMA Guide to the Data Management Body of Knowledge (DAMA DMBoK) as the ‘definitive introduction’ to data management. DAMA DMBoK defines a standard view of data management functions, terminology and best practices, without detailing specific methods and techniques. Because DAMA DMBoK is not an authority on any single topic, it refers readers to acknowledged and widely accepted publications, articles and websites. The first edition of DAMA DMBoK is available in print and PDF versions from Technics Publications or the website of DAMA International www.dama.org
Data management with DMBoK in practice
DMBoK offers an independent and complete foundation for setting up all aspects of data management. On that basis, it is possible, in practice, to accurately map out three important factors:
- The current maturity level of the organization
- The desired/required level
- The roadmap tracing the route from the current situation to the new one
An assessment measures each individual aspect of data governance. An average score can be derived from this, which is a snapshot of a company’s maturity level. The assessment also shows which elements need extra attention; these can be focused on to reach the next maturity level
To-be situation: Using the strategy, vision and mission of the organization regarding becoming data driven, it is possible to determine the maturity level required to achieve the objectives. That too is done using the various individual levels of the DMBoK model.
Roadmap: The output of the assessment and the desired situation can be converted into a practical roadmap that shows how processes, people and systems need to work to form a combined, data-driven organization. The assignment of responsibilities in this respect is crucial, as is the ability to think in terms of information chains, and viewing and treating data as a company asset.
How DMBoK helps organizations become data driven
Decisions are made on the basis of facts, rather than ‘gut feeling’. These facts must be based on reliable information and data. And that requires good data management. DMBoK is the ultimate data management model for this purpose, a model that can reveal the maturity level of the organization and trace the roadmap to becoming data driven. DMBoK covers the most important aspects and processes to structure data management properly and implement it throughout the company, because data is not something that only concerns the IT department. To achieve a structured set-up, Quint has devised a maturity model which allows organizations to create a roadmap highlighting the foundations for a data-driven organization. Robustly structured data management does not make an organization rigid, but agile: if something has to change, it can be changed relatively quickly. Towards robust, agile data management in three steps:
- Assessment: determination of the current maturity level of the organization.
- To-be situation: the maturity level required to be able to achieve the business objectives
- Roadmap: the actions and measures to be taken in order to trace out the route from the current situation to the new one.
Some practical examples
Reference Case 1 – Vitens sets up data governance using DMBoK
As a supplier of drinking water Vitens has a great number of business processes, each of which demands a large amount of data that is growing by the day. Far-reaching automation and the use of sensor technology have made data an essential component of Vitens’s operations, with quality and availability being of singular importance, for instance, for predictive maintenance. What Vitens really wanted to do was to manage data as an asset, and to set up its data management professionally.
As well as being a production business that pumps water to the surface, filters it and tests it in its own laboratory, Vitens is also a distributor, transporting water via its own network to consumers and charging them by means of water meters. Each of the required business processes uses large quantities of data which grow by the day. This, combined with far-reaching automation and the use of sensor technology, has made data an essential part of business operations. Management’s vision is that data is an asset, and therefore has to be managed.
A large number of different types of IT converge within Vitens. As well as traditional IT, with back-office systems such as SAP and GIS, water production is also largely automated, and the network is now managed based on sensor data. The more processes that can be steered by data, the higher the quality of that data needs to be. An urgent need has therefore emerged at Vitens to gain full control of the data. “Data is one of our primary assets,” says Popke Graansma, Head of IT. “As such, it has been incorporated by senior management in the reformulated Vitens strategy.” To organize sound data management and ensure proper governance, he called in the assistance of Edwin Eichelsheim, consultant at Quint Wellington Redwood.
Together with Eichelsheim, Graansma embarked on a process to give shape to data governance and data management policies at Vitens. In a workshop, Eichelsheim introduced the DMBoK model (see box) developed by the DAMA International community. The maturity of the organization was assessed using the DMBoK model, and this position was used to shape the approach to be used to set up management and governance. The initiative for the data project therefore lies with the IT department. But, Graansma continues, “Because IT lies at the core and given my experience with data, it was the business that asked me to write a plan. But it is certainly a ‘co-production’ between business and IT, and this is also reflected on the work floor.” The current plan sets out the primary lines of his vision of data management and was well received by the board and first-line management. “They fully appreciate that data is an asset and as such it must be managed and organized. Although no hard business case has been prepared, overall there is a good enough case that justifies investments in people, time, resources and attention.”
Vitens’s Asset Management department is responsible for managing all assets, such as the pipeline network, the pumping stations and a host of filtering technologies, water softeners, etc. Furthermore, Vitens has a number of water production sites which are subject to specific legislation and regulations. This too is an important aspect for Asset Management. The board has followed up on the belief that data is an asset by assigning responsibility for data management to the Asset Management department. For the team, this is an entirely new discipline, and a number of data stewards have therefore been added. Despite the fact that data management is outside the IT organization, the IT department has provisioned the required infrastructure and application management processes expeditiously. The data analysis method is also based on DMBoK. The SAP BO BI tool was rolled out by IT to make it easier to access and combine certain data from a data warehouse. In relation to Geographic Information Systems, access has been improved to the extent that users can now present information at map level.
Working with Quint
Most of the new people, including a program manager, will make a start on data governance early in 2016. Graansma’s assessment is that in the initial period, the focus will mainly be on writing the work plan. “Things such as the governance structure, working arrangements and protocols all need to be set down in detail on paper. But there also needs to be a gradual transition from thought to action, with a concrete data improvement process. All this still needs to be decided. The first results must be visible by the end of 2016. From then on, it will be an ongoing activity.” He looks back positively on the cooperation with Quint: “Their expertise and practical experience with data and DMBoK provides a solid foundation for an efficient and effective approach to data management and data governance at Vitens.”
Reference Case 2 – DMBoK helps Stedin move closer to data-driven services
With modern metering technologies and smart applications, Stedin Meetbedrijf provides companies with insight, an overview and a variety of ways to make energy savings. Stedin Meetbedrijf asked Quint to carry out an assessment to ascertain the maturity level of its data management. Among other things, this study would help the further development of business services based on data. It thus marks the starting point of the road to a new, data-driven enterprise.
Edwin Eichelsheim of Quint Wellington Redwood was asked to perform the assessment in the second quarter by Ashwand Prahladsingh and the then director of Stedin Meetbedrijf. He investigated the maturity of data management in the organization. This is important because a high level of data management maturity is needed to execute the envisaged strategy, and to make Stedin Meetbedrijf a data-driven organization. Based on the Data Management Body of Knowledge (DMBoK), Eichelsheim observed that for most DMBoK focus areas, the company had a maturity level of 1. One of the things that this tells us is that the data vision has not been properly formulated as a business case, something that would help the company as a whole to adopt a data-driven approach, and that data management is therefore in its infancy. In terms of compliance, security and privacy, in contrast, the company was performing particularly well. Based on DMBoK, Quint also provided a roadmap for data management for newly introduced services and products. The assessment gives Stedin Meetbedrijf a solid foundation for making further progress, and there is still much work to be done.
In accordance with its mission statement, Stedin Meetbedrijf has the ambition to become a data-driven organization. The initial applications of data analyses were primarily aimed at improving internal processes. Workflow analyses, marketing, data validation, portfolio analyses and organizing the billing process have already led to many improvements and savings. The next steps encompass developing products and services based on data. Although – unlike in the consumer market – regulation is not an issue, the confidentiality, security and quality of the data is obviously the top priority.
A specific example of a newly developed business service: with the customer’s permission, the consumption history – sometimes going back as many as eight years – is examined to assess whether the capacity contracted with the network operator differs from average consumption. Adjusting the contracted capacity could lead to big cost savings for the customer. A simple example, but the idea is to apply advanced analytics to all available data to identify divergent consumption patterns for customers. Have circumstances changed, or have things remained the same? Have there been any cases of fraud or abuse? Business customers benefit from this.
Data warehousing as a driver of adoption
In order to introduce the board and Operations to the opportunities afforded by data analysis, back in 2014 , innovation manager Tjerk Poot started up his own data warehouse within the metering company’s Corporate Clients division. In actual fact, the initiative came from employee David van der Velden, who had specific ideas about the opportunities it had to offer, and a deep understanding of the business side of the company. With the help of a few third parties, it was possible to create such a data warehouse which brought together four sources. Together, these sources account for several terabytes and around 80% of the available data. The source systems on their own are not suitable for analysis, commercial use or combinations of data. The data warehouse initiative lies emphatically with the business rather than IT. “That might seem a strange way of doing things,” says Tjerk Poot, who is actually responsible for developing new services and products. “But at the time I could see that ideas would be increasingly based on data. So the driving force behind the data warehouse was primarily new business development, not IT.”
No matter how conventional, the data warehouse plays an important role in the adoption of a data-driven approach, says Poot. “You have to keep it small and simple to begin with: by means of conventional reports and dashboards you can furnish management with familiar information, the sort they like to see, which they can use to make decisions. In that way, data shows that it has added value. Moreover, the available dashboard provides much more information than was previously available, attracting the attention of many managers as well as people on the work floor. And the transparency offers many new insights into work processes, exposing not just difficulties but also giving pointers to solutions. The well-thought-out approach of providing the organization with analyses in small, incremental steps thus ensures that the use of data in day-to-day operations is embraced by the organization.”
Assessment as a foundation
The process which will transform Stedin Meetbedrijf B2B into a data-driven enterprise is in full swing. There is plenty of data, but a data-driven company needs other processes, makes new demands on IT, and requires other skills and knowledge from the people who work there. In brief, it requires a change in culture, a new mindset across all parts of the enterprise. The way data management – including data governance – is organized still depends to a large extent on the situation. This means that an issue is dealt with as it occurs, and then a management decision is needed. “In terms of data management, Stedin Meetbedrijf B2B is at an early stage of development, and is still very reactive. They recognize this: they are real go-getters and they like to get stuck in,” says Prahladsingh. “Nevertheless, big steps have already been taken on the path towards becoming a more data-driven organization, and the structure is improving. Quint’s assessment provides a good foundation and serves as guide.”