HomeCareerBlogsFrom Data to Steering: My journey as a BI Specialist in secondment
Blog
From Data to Steering: My journey as a BI Specialist in secondment
Working in secondment at CIMSOLUTIONS means stepping into a new world every single time: new teams, new processes, new data landscapes, and above all, new needs that demand clear answers. That is precisely what has made this work so inspiring to me for years. It is like getting to look around the inner workings of a fascinating organization every single time.
Every client has its own dynamics, but the common thread is always the same: how do you translate complex business processes and fragmented data into reliable steering and accountability information that allows an organization to move forward?
The Levels of Steering and Accountability
The concepts of steering and accountability information suggest a hierarchy within an organization where steering and accountability take place. Within that hierarchy, there will therefore be one or more levels that are steered and one or more levels to which accountability is exercised. Only the highest level steers alone, and the lowest level accounts only.
In a large organization, steering and accountability are usually performed at three levels.
Operational
The operational level executes business processes. Employees record data during their work: customer data, orders, and status changes. Managers provide monthly reports to tactical management, such as order numbers, items sold, and lead times. These reports serve as both steering and accountability information.
Tactical
Tactical management forms the link between strategy and operations. They use reports, including financial reports, to provide direction and to account to strategic management.
Strategic
The strategic level steers at a high level, makes decisions based on financial and substantive steering information, and uses this to give direction and execution to the organization’s established vision and strategy.
Requirements Engineering for Management and Accountability Information Needs
It is my task and responsibility to translate the information needs into usable management and accountability information at the three described levels. In practice, this means:
Conducting interviews with stakeholders and holding brainstorming sessions and workshops with, for example, management, team leaders, functional administrators, business analysts, and domain experts to identify requirements for the management and accountability information. This is called ‘eliciting’. In an Agile Scrum environment, the requirements for concepts, metrics, key performance indicators (KPIs), and associated standards are initially formulated broadly and at a high level, as are the dashboards and reports to be built. Later, these requirements are further refined in one or more iterations for each Agile Scrum sprint;
Documenting and validating the requirements with the stakeholders;
Managing the requirements, capturing and documenting further refined requirements for the sprint, as well as authorizing and granting access to the requirements to all stakeholders. The latter is particularly important for developing, testing, and accepting a dashboard and reports; the requirements also form a repository or data catalog in which all concepts, metrics, KPIs, and standards are described in detail for all stakeholders;
Analyzing data in the source systems for completeness, integrity, and usability for the requested management and accountability information;
Defining concepts, (aggregated) metrics, KPIs, and associated high or low standards;
Designing usable data vault and data mart models (Kimball);
Organizing and leading refinement sessions (Agile Scrum) with back-end (ETL) and front-end (dashboards, reports) developers. During the refinement sessions, it becomes clear to the developers exactly what needs to be built in bi-weekly sprints. At the end of each sprint, a product demo is presented to the product owner. Based on the demo, the product owner presents new requirements and adjustments for the next sprint.
These tasks result in a final product that enables an organization to substantiate its work processes with reliable, uniform, and source-based management and accountability information. Concepts and aggregated metrics of the work processes are no longer raw datasets, but provide clear insights that enable all levels of the organization to manage and account for their actions.
Process mining: what do organizations actually do?
Classic business intelligence provides us with an aggregated metric on the output side of a critical business process or sub-process. For example, that a certain figure of a metric must be achieved per month to assess whether a critical business process is performing adequately. We call this a KPI, a Key Performance Indicator.
A standard is also agreed upon for a KPI. For example, for the concept of ‘Sickness Absenteeism’, a KPI ‘ZPM’, ‘Sickness Absenteeism Percentage per Month’, is agreed upon. The standard is set that this may not exceed 2%. This is called a low standard: in principle, we should preferably not exceed this. A high standard means that a certain minimum value must be achieved per month. We should preferably not fall below this. “When constructing a highway, at least one kilometer of asphalt must be laid per day on average.” When the low or high standard is not met, it is important to find out why this is the case.
Classical business intelligence can subsequently be used to perform analyses on the aggregated metrics against various perspectives or dimensions. In the example of exceeding the 2% sickness absence per month, one can examine which department was affected. Regarding the construction of the highway, where at least one kilometer of asphalt was not laid on average per day, one can identify which team and during which period this occurred. In this way, one can ‘put one finger on the sore spot’ and, most importantly, take action, or in other words, make timely adjustments.
However, there are conceivable business processes that are too complex to analyze from a few perspectives and draw conclusions from. When multiple complex subprocesses, feedback loops, and/or decision points are present in a business process, a classical analysis of the (aggregated) metrics from a few perspectives may no longer suffice. In such cases, we really want to be able to look into the large amount of data within the process itself, even somewhere in the middle of the process, and not just on the process output side!
A team leader in charge of a small, manageable process can steer it without detailed process information, but not when dealing with a complex business process. A KPI is not achieved: “Where exactly do I need to intervene in my process?” In effect, we are ‘blind’ to seeing exactly what has taken place within the process.
In such cases, process mining can help discover exactly where the bottlenecks in a business process lie. As a result, the blind black ink spot disappears.
In addition, process mining can assist with process simulations for process adjustments, performing compliance and conformance checks, and predicting the outcome and remaining lead time of ongoing processes.
There is a wealth of information available on the internet regarding process mining and a wide range of process mining software. A good starting point could be an introductory course in process mining. A very good course I took myself is the one from Coursera: https://www.coursera.org/learn/process-mining.
Why secondment at CIMSOLUTIONS works for me
What do client assignments and projects at CIMSOLUTIONS have in common? Impact. But also variety, growth, and the trust to take responsibility. At CIMSOLUTIONS, you are not a pawn; you are seen as a professional and as a person. You are given the space to learn, explore, experiment, and help colleagues grow.
For me, secondment means:
Working on assignments where you truly make a difference
Constantly getting to know new organizations and data landscapes
Broadening, deepening, and immediately applying your knowledge
Collaborating with committed colleagues who share the same drive
Continuing to develop through training, certifications, and knowledge sharing
In 2021, I was elected Employee of the Year for the Rotterdam office: this felt like wonderful recognition for my hard work, but actually, every successful and interesting assignment is a reward in itself.
Finally: data is only valuable if you make it understandable
Whether it concerns social services, customer contact, urban development, or enforcement, data always tell a story. My job is to make that story clear, reliable, and usable. The combination of business analysis, data analysis, data modeling, process insight, and technical knowledge helps with this every day.
And that is precisely where the strength of secondment lies for me: every assignment is a new chapter, but the common thread remains the same. You help organizations move forward by bringing structure and insight into what they run on, both daily and in the medium and long term: their processes and their data.
Frank Wouts Senior Data & Business Intelligence Specialist
Also interesting
Blog
From Data to Steering: My journey as a BI Specialist in secondment
How do you translate complex business processes and fragmented data into reliable steering and accountability information that allows an organization to move forward?
From need to realization: why the path to value is rarely linear
By combining design and validation and giving the Business Analyst a directing role before realization, organizations can deliver value faster that truly aligns with operations.
From raw data to business value – Medallion Architecture
The Medallion Architecture, also known as the Multi-Hop Architecture, has received a lot of attention in CIMSOLUTIONS' knowledge-sharing sessions over the past year. This architecture forms the foundation of modern data platforms for a growing number of organizations.
Current AI news is mainly dominated by generative AI and Large Language Models (LLMs). Although these models deliver impressive performance, they also have their limitations.
Quantum computers can perform complex calculations and analyses many times faster and more efficiently than the traditional computers we work with every day.