Poor data quality is adversely impacting businesses with an average of $15 million per year in losses, according to a Gartner’s Data Quality Market Survey.
But, first, what is data quality? Data quality refers to the status of your data measured against several criteria.
And what are the criteria used? They vary depending on your business requirements. But to ensure that your data is of the highest quality, we recommend measuring against these criteria:
- Validity
- Accuracy
- Uniqueness
- Completeness
- Consistency
- Timeliness
- Integrity
- Conformity
Data quality can’t be achieved overnight, nor is it a one-time event. It starts with the assessment of your data health. This assessment will give you a clear picture of the follow-up actions to take to improve your data. After all, you can’t improve what you can’t measure.
The assessment outcome can become an input for your data governance framework that prescribes the rules, roles, and policies on the use of data. With this framework in place, you can start embedding data quality within your business. This leads to better-informed decision-making, improved compliance, and enhanced operational efficiency—essentially, all your organisational goals.
What are the key milestones for data health?
To ensure that your data is ready for the subsequent remediation and governance processes, our team of experts have recommended the following key milestones in assessing your data health and coming up with the diagnoses. These were devised based on years of experience engaging customers from different industries.
Explore
We begin by exploring and understanding the sources of your data according to business towers. We’ve defined business towers based on the most logical grouping of critical master data areas. They are:
- Finance and Project data
- Asset and Maintenance data
- Customer data
- Supply Chain and Procurement data
- HR and People data
Of course, this depends on how you run your business. If your business deals primarily with capital projects and you think it makes more sense to have a separate business tower for project data, then the grouping can be changed.
At this stage, we’ll also identify potential Data SME and Business Lead for each business tower. The Data SME is responsible for the data belonging to the business tower and becomes the single point of contact. While the Business Tower Lead takes up the business part of things and helps the Data SME wherever possible in gaining access to relevant information.
Connect
Once the data sources have been identified, we’ll establish connection with these data sources and initiate the groundwork for business and technical profiling.
Technical profiling sets up the technical attributes of metadata, e.g., the source, format, length, and data type. While business profiling refers to the business attributes, e.g., definition, usage, and context.
Profile
This is where the technical and business profiling of data takes place using MDO. It provides an initial understanding of the data areas for each business tower, surfacing key highlights on the current data health.
Define
Once we have clarity on the levels of your data health as reported via MDO dashboard tools, we’ll proceed by defining new business rules for each business tower. Defining the rules for each of the 8 data quality categories will be the most ideal so you can address current and future data anomalies holistically.
This is where we start engaging your Data SMEs and Business Tower Leads by sharing the results and business rules with them.
The findings may vary between industries and even business towers. This is normal because as a business, you’d have several evolved and mature data areas as compared to others.
For example, if you already have customer data consolidation process in place to upsell and cross-sell your products, you’d have fewer customer data issues to address. But when it comes to your asset and maintenance data, it could be a different story altogether.
Analyse & recommend
Now that your business stakeholders have seen the results, our enterprise consultants will deep dive into the findings and walk through the recommendations with them. The outcome of this exercise is Data Quality Registers detailing the main data issues encompassing all business towers.
And depending on the data issues, we’ll suggest suitable measures such as implementing active or passive governance, executing spare parts’ data cleansing before categorising them, etc.
Remediate
We’ll proceed by discussing optional remediation actions with your business people, as well as populating the Data Quality Registers with these details for their easy reference. The execution of the remediation plan depends largely on the maturity of the respective industry or business tower. Upon obtaining business approval, the plan will be set in motion.
Conclusion
These data health key milestones were designed to help you advance to the data governance stage. It will pave the way for data quality awareness and ownership amongst your people. While the decision is entirely up to you in carrying out the remediation steps, you’ll gain a better picture of the state of your data at the end of this assessment.
With MDO and our team of experts and consultants, you’re well-positioned to embark on your data quality journey and future-proof your strategic aspirations.
Written by: Shigim Yusof
