Data and analytics governance model is currently in the department not formally defined.
Many of the mega problems as described in (9) such as 1. Existing analytical solutions are not used properly
2. Lack of understanding of existing reports, where to find them and what information are shown
3. Duplication of reports by different users and many different versions in Excels shared over emails
4. Different data sources with different rules are used for same metrics
5. Not clearly defined business terms, metrics, exceptions from calculations 6. Outdated or missing documentation of reports, data sources and IT architecture
7. Documentation exists mostly in the form of SOP of how to generate report or dashboard, however, does not describe who is using it, why, what are the metrics and how they are calculated
8. Overall documentation of data sources and IT infrastructure is outdated as it is delivered usually during project completion, however any changes are afterwards not integrated
9. Lack of trust in data and data quality, that leads to creating further dashboards and reports by individual users
10. Speed and flexibility of delivering analytical solution is not according to business expectation
11. Big boom of using Power BI across different departments, with lack of strategy, governance, support, and trainings.
Data governance has been in the company mainly domain of IT, therefore we can see that governance proves on IT side is set up and managed. On business side, there is currently no data & analytics governance process set up in international logistics. Management of the
36 international logistics is aware of it, therefore new role of Data Product Owner has been created which will take responsibility to set up the governance process together with implementation of the data lake project in 2021 – 2022, which aims not only to create data storage capabilities, but overall data & analytics strategy for international logistics department.
Data governance frameworks and approaches are very widely documented by many different organisations, following key principles for governance implementation defined by (8), which will be applied on our use case to ensure successful governance of the solution.
1. Analytics environment must be owned
2. Dedicated person for implementing the governance 3. Governance system must be easy to understand and apply 4. Create rules and governance based on settled practise 5. Set up based on successful implementation
6. Use existing resources and respect maturity level 7. Implement gradually
Data governance can be described in different areas and layers (8) that can cover the entire development and life cycle of the analytic environment.
Figure 20 Conceptual Data governance model
5.1 GAP Analysis
Following the key principles data governance for our supply chain losses visibility and to understand where to start, I have been first creating gap analysis of existing governance, problems comparing with standards.
37 Following areas have been identified as most problematic and they will be tackled as priority for setting up data governance for international logistics supply chain losses. After successful implementation and testing, they can be copied to other analytical used cases in department:
1. No business owner is defined for data & analytic
2. Lack of documentation of data sources, metrics, business terms 3. Low data literacy in department
4. No dedicated analytical resource to support business users on daily basis 5. No data and analytics strategy set up for the department
6. No data stewardship
5.2 Roles and Responsibilities
First area that must be defined are different roles and responsibilities related to data and analytics environment. For easy implementation which is one of the recommended principles (8), we will use resources available in international logistics department, reflecting their current knowledge and capabilities and adding some of new responsibilities to their roles.
Table 12 Roles & responsibilities
Role Responsibilities
Data product Owner
• Data stewardship
• Power BI development backlog responsibility
• Responsible for continuous improvement in data management area, best practices, benchmarking
• Responsible for data governance set up
• Data catalogue NEW
• Data literacy increase in the department NEW
International logistics Capability Manager
• Responsible for building new capabilities for international department in terms of processes, automation, and digital experience
• Aligning project priorities within department
• Responsible for budget delivery
• Define Data & analytics strategy for International logistics NEW
Power BI developers • Responsible for development of new dashboards in Power BI
38
• Monitor performance of current Power BI Dashboards and Data sets
• Responsible for delivering documentation as part of project finalisation
• Responsible for performance optimisation of data model and architecture
Reporting specialists
• Responsible for delivering data related activities such as extractions on daily basis for existing reports and
dashboards
• Monitor usage of reports and dashboards
• Identify opportunities for improvements
• Responsible for documentation of data
& analytics area NEW
• Support end user with minor Power BI adjustments NEW
Business Expert
• Business expertise in certain area
• Business owner for data in respective area of expertise NEW
• Responsible for defining data quality rules and identify actions to improve it NEW
Master data expert
• Responsible for master data management
• Responsible for defining data quality rules and identify actions to improve it NEW
5.3 Documentation
Existing documentation related to data and analytics exist mainly in forms of SOP on how to perform certain activities, data extractions, reports and is used primarily by onboarding of specific person into role, rather than broader knowledge management document.
Currently department documentation related to other business areas is being migrated into new tool, which enables better knowledge management, schedule periodical review of the document content and great search functionalities. Documentation there is open to anyone in the organisation.
39 To close the gap of lacking documentation in the data & analytics management in the department, we will use this tool also for creating documentation in data related area, by defining common standards, that should be followed for every analytical solution.
1. Define responsible people for every document as per agreed roles:
2. Standard set of documentation requirements that must be documented for every report or dashboard has been defined on top of existing SOPs focusing on step-by-step creation of the report or dashboard
• Metrics and calculation methods
• Description of data elements
• Data Sources
• Exceptions from calculation
• Business process related to the report or dashboard
• Objective of the dashboard/report
• Frequency of update
• Data history
• Distribution method
3. Set up review cycle periodicity as minimum 1 year Figure 21 Documentation RACI model
40
5.4 Data quality management
Data quality is currently one of the most problematic area identified by business users and it is being mentioned in many internal surveys, workshops, or projects. Lack of trust in the data is exceptionally low. Also, it is not clearly formulated on what does it actually means that the data are correct, what is the expected level of data quality and which is the leading data source. All those problems are indicating that data quality is not managed and will require broader strategy in the department.
For this analytical solution, we have defined following data quality rules, that should be implemented and regularly checked to ensure correctness of the dashboards:
1. Relevant transactional data must be saved in dedicated folder by each zone 2. Column names in the zone transactional data must be kept
3. Cost list per SKU for EUR export must be available for all active SKUs 4. Budget rates must be updated on yearly basis
5. Any changes in current transactional data must be communicated to Power BI developer for adjustments
5.5 Governance roll out
Implementing all defined areas of data and analytics governance process for international logistics department will be done gradually during 2021 as different areas requires shorter or longer period for preparation, implementation itself and most importantly change management. Important part of aligning roles and responsibilities and defining data governance area as one of the priorities with individual targets set for respective roles is critical milestone that can ensure it successful roll out.
Detailed governance roll out plan will be completed in Q1 following timelines of dependent project of implementing data lake as certain areas will be rolled out together.
New responsibilities within the existing roles have been agreed with management of respective teams. As a next step, alignment with HR department to define training requirements and identify suitable trainings for building new required skills and preparing people reprofiling training plan related to Power BI, Power query knowledge, general data, analytics, and new technologies literacy.
Documentation as per newly agreed standards has been agreed to start immediately and target to finish at the end of H1 2021 for existing reports and dashboards
Most complex and problematic area is related to creating data business dictionary, understanding of data sources in respective areas, where there is long term lack of knowledge. This area is expected to be delivered together with implementation of data lake project.
41