Business-Driven Data Organization
Business-driven data organization holds the benefits of clearer value and alignment of business and data initiatives. However, it does not develop without deliberate actions related to business data capabilities. We can help you with that.
Written by — Sami Helin, Data Architect
Becoming data-driven is an objective pursued by many organizations. While the term’s objectives of developing data utilization and generating value from data are relevant and good, it can also direct to too technical data approaches: data is thought before business – data is the driver. And as data resides in different business systems, data becomes a systems development topic. Even worse, data is often considered a separate topic disconnected from business development.
Data-driven vs. business-driven data approach is somewhat semantics, but certainly, data needs to better align with business. And of course, data and technology unlock new business opportunities and affect the way we conduct business. But talking data through systems enforces siloed data thinking and distracts communication. It also switches business focus from making data-related business problems and objectives tangible. The focus is on understanding systems and technologies and what is possible with the data at hand.
Reasons for separation of business and data
How did this happen? Why did data understanding get distracted from business understanding since data is by principle just a persistent representation of business operations? One big reason has been the advancement in technology and systems. Many systems have become so complex that system experts are needed to translate the data in systems into valuable information for decision-making. And when businesses typically run quite a number of systems, understanding how things relate to each other requires even more experts primarily with systems and technical understanding.
Furthermore, initiatives where data collection is planned, typically have quite an operational focus. For people directly working with the system collecting the data, data understanding and terminology in data also starts to align with systems. But decision-making typically needs an end-to-end process understanding rather than a specific system’s part. It becomes difficult for decision-makers to pinpoint the data they need as it is scattered in different systems and possibly not aligned or even collected consistently.
Implications of systems-driven data development
How has this affected data solutions development? A number of data roles have evolved with different types of technological and system expertise. And a whole range of technological expertise is combined into an IT-driven data approach.
Data development starts with technologies and architectures and many data roles get nominations from people who understand more aspects of the systems than the business objectives and development initiatives. This leads to data operations starting to form their own problem reality drifting further apart from business priorities and decision-making.
One tangible example is data governance which states noble objectives of better data to support business objectives and prevent misuse of data. In practice, it is typically seen as a restrictive element rather than something really useful to business. And certainly, trying to enforce a data governance program in an IT and system-driven landscape will be a pain. But the next chapter details some development suggestions that make data governance a normal business operation - without really emphasizing we do data governance.
Developing business capabilities and understanding of data
How to get business to the driver's seat in data initiatives? Many organizations have started decentralizing data operations and that is certainly one pattern. However, this really doesn’t change a lot of the fundamental issues. It just brings the people-speaking system data closer to the business needs. Data proficiency and understanding among the people doing business decisions and driving business initiatives typically do not evolve that much – the translators are just closer to business.
A more profound impact requires raising the business views of data above systems. In practice, this means focusing on making business language, business decisions, and objectives explicit. Discussing what we want to achieve, what we are planning to do, and how we measure our operations without any reference to any systems is a good starting point. And this focus is needed in the initial business development initiative when it is decided what data is collected and how – not in the end when all operations are done and someone tries to separately make sense of where the data is and what the content is.
Changing mindset from systems to business regarding data is a many-sided operation. We are talking about quite a profound change. However, the means and steps should not be overwhelming. It is more about making a business perspective explicit in different data decisions and operations. Below are some items to redefine the role of business and data experts. With the role of data experts, I mean roles whose focus is on systems and data methodologies, whereas business involvement comes from people responsible for business decisions and operations.
- Own the data: Do the data decisions and describe what the business needs data about.
- Make the purpose of the data explicit: Why do we need the data, what is good data? New purposes will evolve over time and this understanding develops as business evolves.
- Explain business language: What do we mean by the terms we use?
- Describe operations and decisions: Metrics that we follow and operate upon.
- Clarify data organization: Who decides metrics, what business information is sensitive to show, and to whom?
- In initiatives: What are expected outcomes and who should make them happen?
- Ensure leadership support for transformation.
- Participate in cross-functional teams around business end outcomes.
Data expert involvement
- Translate the business data model into system mapping (not the other way around) and make it explicit.
- Support with data capabilities and take lead in developing and managing certain areas.
- Give guidance on system and technology-initiated data opportunities.
- Participate in cross-functional teams around business end outcomes.
And this does not mean that we stop doing business and start thinking about those topics mentioned above. It means doing those things explicitly in new business initiatives and keeping the business model intact. Data needs to evolve as your business evolves, iteratively and continuously. And the support needed from data experts is of course big - the system and technological landscape is not getting easier. It is about changing the initial mindset and developing the collaboration that is crucial.
If you don’t have a data catalog yet, I strongly suggest some approach in that area to accumulate this understanding. However, this is not a tool exercise, but primarily a matter of responsibilities, communication, and skill development.
Benefits of business-driven data organization
How is this type of approach better than the one originating from system understanding? One clear implication is that starting with system understanding is slow and difficult - it typically needs a lot of investigation and clarification.
By starting with business, data decisions start to relate to business decisions: how do we restrict visibility to certain data, how can I find information about sales, customers etc. We are referring to terminology and topics business can and should make decisions upon. We don’t need to refer to the complex capabilities and structures the systems need to implement those things.
Furthermore, rather than starting to think about data topics in a reporting initiative, bigger business involvement in understanding the data we collect or need to collect, will have a big impact on the speed of utilizing the data. Basically, business involvement is shifted from an actor on system outcomes to clarifying the business where systems and data should fit in.
Business-driven data organization holds the benefits of clearer business value and alignment of business and data initiatives. However, it does not develop without deliberate actions related to business data capabilities. Business needs to take responsibility for data and data decisions. Business operations’ relation to data also needs better visibility as clarifying the purposes for data needs to develop. Those purposes and actions related to data enable deriving its value and understanding quality requirements for data and thus enabling focus on the most important elements.
Of course, business is involved in system projects also at the moment clarifying the processes, data, and reporting needs. However, it is rarely the case that there would be a model on decisions and data needs for the system to fulfill. Often the expectations are more around what the system provides. And as there is no business data model to map the system against, often the system black box of data is put in place.
This shift will require new skills and certainly a lot of support from today's data experts. But for organizations taking the business-driven approach, the holy grail of fast, high-business-value, data initiatives with good quality data is within reach. And with the digitalization development, data quality and availability requirements really don’t leave a lot of options - profound changes can’t be made with data as an afterthought.