21 March 2025
Master data, often overlooked but not to be ignored
Master data plays a key role in processes and supports decision making. Why do many companies not have a robust process, governance and resources tied into good master data management? Not paying enough attention to master data management leads to people chasing data, wasting time and effort on doing manual rework and to drawing wrong conclusions. Follow the five step approach to improve the quality of master data.
Master data is non-transactional, think of product, customer, supplier, employee or location data. Especially in multinationals, globally active in multiple countries and having a long history in mergers & acquisitions, a complicated IT landscape exists. Having accurate master data that can be relied on and serves as a single source of truth in the entire organization, is a challenge.
BCI Global recognizes the challenges clients have while working with their data during projects. Take for example a supply chain network study where a big pile of data, including master data is collected to create an overview of the current supply chain set up. Organizations typically have difficulties on providing high quality data. Common datasets have blank, incorrect and duplicate values and even missing records. Or several datasets show contradictory master data. What part of the data records collected can be trusted? The data collection phase in a project often is an iterative process that takes too much time and budget.
- Appoint data owners & stewards
- Determine business critical data attributes
- Create data rules
- Use & maintain data rules
- Measure & report on data quality
Appoint data owners & stewards
The first step is to make the scope manageable. Don’t try to set up the end state master data governance organization, but start small. Start within a part of the existing organization structure and involve several employees with different levels of seniority, having good knowledge of the business, key processes and systems. They could become the future data owners or stewards. Data owners are responsible for data within a domain, for example product related data. Data owners are supported by data stewards who are responsible for filling in the daily activities related to data quality. These roles will contribute to govern master data in a structured way and will create ownership within an organization where this was missing before.
Determine business critical data attributes
Determine business critical data objects and attributes, incl. their source system. It is not the objective to create a comprehensive list containing all possible data fields within the organization. Focus on the ones that are causing pain in your daily operation and prevent well-founded decision making. Think of attributes that cause customer complaints or parameters used by a planning tool to calculate orders. Generate buy-in and involvement from business representatives that are heavily impacted in their daily work by these master data fields. Do they agree this being the list with critical data fields?
Create data rules
Data owners, stewards and business representatives have to determine and document a set of rules describing the allowed values for the key data fields identified. Some fields will always have a fixed value or will have a limited variety of values. Try to end up with as little as possible attributes where the allowed values are unique and cannot be captured in a list of rules. This helps setting up and maintaining master data and makes the auditing process less challenging.
Use & maintain data rules
As reality changes, also the set of data rules is not written in stone and needs maintenance to remain useful. Imagine the effect on the business when data stewards would not maintain the rules for a year. New master data records are set up incorrect, quality checks are done against outdated rules, processes are executed with errors, rework is required and (strategic) decisions are taken based on incorrect master data.
Measure & report on quality
Without measuring and reporting master data quality, an organization is still kept in the dark and does not know whether its efforts are paying off. Data stewards should periodically compare the values against the set of data rules and check for deviations. During the first rounds of this exercise, a full comparison will be wise. Set up KPI’s and report on the scores. A monthly drumbeat could look like:
- Day A: create a report containing the records and key fields in need to be verified
- Day B: compare the values from against the data rules
- Day C: know the matching records and the ones in need of correction
- Day D: make the corrections in the source system
- Day E: report on quality by sharing the KPI scores within the organization
- Day F: initiate root cause analysis on key data fields not meeting their targets
- Day G: implement a structural improvement to prevent repetition.
Improvements are visible after a couple of cycles. The need for a complete check can be replaced by randomly sampling a part of the master data records for each of the critical attributes.
Building on this foundation
The organization now has a better structure in place, is more in control and is aware of its master data quality. The higher standard of master data contributes to first time right execution of processes by employees, but also significantly improves the quality of outputs generated by systems that use master data inputs. Master data that is used in repetitive or one-off projects or reports do not require time consuming clean-ups first, can be trusted as-is and is instantly ready for decision making. The company has to decide whether this foundation is sufficient or needs to be built upon further by including more data objects and attributes, systems or parts of the organization.
For more information related to master data, contact our Senior Consultant in related contacts.
