Every organisation that works with meaningful volumes of data should conduct periodic data quality assessments and or regular data quality audits, to ensure that data quality is adequate and that they are not exposed to unnecessary risks. A data Quality assessment is an essential component of any data quality solution.
A Data Quality Assessment accurately measures Data Quality
There should be no debate that poor-quality data costs money. Indeed, according to Gartner Research, “the average financial impact of poor data quality on organizations is $9.7 million per year.” But at which end of the scale do you fall?
Is the quality of your data better, or worse, than average?
Data quality assessments accurately quantify the state of data quality in the assessed datasets, and uncover hidden issues that may be impacting your business operations.
Have you failed a financial or regulatory audit?
Are you implementing a new system?
Is your data appropriate to deliver the required reports?
How will you address known (and unknown) data quality issues?
How bad is your data quality really?
A data quality audit or assessment will give you an accurate answer and should be performed before engaging in any significant data-related project.
The Importance of a Data Quality Assessment
A Data quality assessment plays a crucial role in ensuring the accuracy, completeness, and reliability of organizational data assets.
Conducting regular data quality audits identifies data errors, inconsistencies, and anomalies that may impact decision-making processes or business operations. With thorough data quality audits, organizations can proactively address data issues and implement corrective measures to improve overall data quality and mange data risks
Does your data comply with data policies and standards?
How will the Protection of Personal Information (PoPIA) impact your business?
How does your data support your requirements to comply with regulations and legislation such as RDARR, PoPIA, the Solvency and Assessment Management regime, and many more?
Where is your data held? What needs to change for compliance?
Can you identify where data does not comply with required business standards, and track improvements over time?
Data Quality Assessment for Big Data
In the era of big data, data quality metrics and expertise are more important than ever.
As organizations collect and analyze vast amounts of data, ensuring data accuracy, reliability, and relevance becomes increasingly challenging. By leveraging data quality metrics and expertise, organizations can evaluate the quality of their big data sets and make informed decisions based on trustworthy insights. Discover why data quality assessments are crucial to unlocking the full potential of big data.
Why Data Quality Audits are Critical for Reference Data Sets
Reference data sets serve as foundational components for various business processes and analytics initiatives.
Therefore, ensuring the accuracy and reliability of reference data is paramount for organizational success. Regular data quality audits on reference data sets identify and rectify inaccuracies, inconsistencies, and discrepancies that may impact business operations.
Are you concerned about data quality?
Understand how we use market-leading data quality tools supported by experienced consultants, to rapidly assess the fitness of your environment to meet your specific business needs.
Data Quality Assessments for Continuous Improvement
Data quality Assessment and monitoring is essential for ensuring data remains accurate, reliable, and consistent.
By regularly monitoring data quality metrics and identifying potential issues, organizations can proactively address data quality issues and drive continuous improvement initiatives. With a robust data quality monitoring process, organizations can detect anomalies, trends, and patterns that may impact decision-making processes or business outcomes.
The Case for a Data Quality Dashboard
A data quality dashboard provides organizations with a centralized platform for monitoring and managing data quality metrics.
By visualizing key data quality indicators and trends, a data quality dashboard enables stakeholders to quickly assess the health of their data assets and identify areas for improvement. With customizable dashboards and interactive features, organizations can tailor their data quality dashboards to meet their specific needs and objectives.
How to Craft Meaningful Data Quality Metrics
Crafting meaningful data quality metrics is essential for assessing and improving the quality of organizational data assets.
A Data Quality Assessment can uncover "de facto" business rules and metrics that can be used to define and formalise relevant metrics for ongoing Data Quality Audits. By defining clear and relevant metrics, organizations can measure the accuracy, completeness, and consistency of their data and identify areas for improvement. With well-crafted data quality metrics, organizations can track progress over time and prioritize initiatives to enhance data quality.
When to perform a data quality audit
Every organisation that works with meaningful volumes of data should conduct periodic data quality assessments, to ensure that data quality is adequate and that they are not exposed to unnecessary risks.
However, there are some specific circumstances that should trigger a data quality audit, including:
Before making an important, data-dependent decision.
Understand where existing data may need to be enhanced to be updated to be suitable for its new use.
For example, one client invested tens of millions in a new client segmentation model, only to find that the data points proposed by the external consultant were poorly captured and useless for analytics.
A data quality audit would have allowed them to plan the project better, focus on the data they had, and deliver the required outcomes.
Before implementing a new system:
Any data migration carries risk. Whether you are rolling out a new ERP, implementing master data management, or moving your EDW to the cloud, a data quality assessment identifies data risks before they become issues.
For example, one client rolled out an integrated planning application not realising that the data in their ERP was not sufficiently accurate, complete, or consistent to be used in the new planning system.
The result - the project failed to deliver the expected returns.
A data quality audit would have allowed them to plan the project better, adjust for the data issues, and deliver the required outcomes.
To comply with regulations:
Data issues can, in turn. lead to non-compliance with data-dependent regulations such as the PFMA, PoPIA or the Basel Accords.
For example, at one client we helped them to identify and resolve data issues that had caused them to fail a financial audit
Before beginning a data clean-up:
It may sound obvious, but many organisations engage in data cleansing with no idea of how bad the problem is. Often this results in wasted time and money.
A data quality assessment gives you a baseline against which you can prioritise, identify opportunities for process or automated improvement, and measure the progress of manual cleansing.
The Data Quality Audit Process
Each data quality assessment is tailored to your specific business drivers and can be delivered much more quickly, and more accurately than manual / SQL approaches.
A data quality assessment typically comprises 5 steps:
- Information gathering: A high-level understanding of business goals, processes and known issues
- Identification of data: Confirm which data is in scope, where it is stored, and how to access it.
- Data profiling: Measure data quality including completeness, uniqueness, consistency and referential integrity
- Confirmation of findings: Discussions with stakeholders to confirm findings and impact
- Report: Finalise reports and other deliverables with business context
Actual measures of data quality issues (known and previously unknown) will be exposed to your business and technical stakeholders to allow you to accurately identify the scope and risk of any remediation effort.
Deliverables of our data quality assessment include:
- Review of key business objectives and relevance of approach in this regard.
- Various assessment reports and recommendations (sample exec summary below)
- Roadmap/plan for delivery
How to Create a Data Quality Scorecard
A data quality scorecard provides organizations with a systematic approach to assessing and managing data quality across various dimensions and is critical for ongoing data quality audits
By defining key performance indicators (KPIs) and measurement criteria, organizations can create a comprehensive scorecard that evaluates data quality based on specific criteria and objectives. With a data quality scorecard in place, organizations can track progress, identify areas for improvement, and drive continuous data quality initiatives.
Reuse your investment
The last thing you need is to invest millions in new systems only to find that the data quality issues that plagued you previously have resurfaced within months of your go-live date.
Our approach allows new systems to reuse data quality rules and automate data cleansing and management processes both to improve data quality and to maintain desired levels.
FAQ
What is the difference between a data quality audit and an assessment?
A data quality audit is the process of evaluating the quality of your business data to ensure its accuracy and reliability, typically in response to a specific event. A data quality assessment is repeated periodically, to ensure that data quality standards are being met and maintained or improved.
Why are data quality assessments important?
Data quality assessments are important because they ensure that your business decisions and operations are based on accurate and reliable data. Poor data quality can lead to incorrect insights and decisions, which can have serious consequences for your business, and is proven to cost businesses millions each year due to operational errors and issues.
How often should I perform a data quality assessment?"
The frequency of data quality assessments depends on the nature of your business and the volume of data you handle. However, it's recommended to perform a data quality assessment at least once a year, and more frequently for critical data sets.
What are the common data quality issues?
Common data quality issues include missing and incomplete data, incorrect data, duplicate data, inconsistent data, and outdated data.
|
Group |
Quality |
Issues Considered |
Example of DQ Problem |
|
Relation to other data |
Referential Integrity |
Do records exist where expected? |
Counter-party records carry a link for a Legal Entity that does not appear in the table |
|
Cardinality |
Is the structure of relationships among entities and attributes maintained consistently? |
A customer has more than one VAT number. |
|
|
Structure of fields |
Format |
Do values follow consistent formatting standards? |
Company registration number appears as xx/xxxxx/xxxx, xxxx/xxxxx/xxxx, xx-xxxxx-xxxx etc |
|
Standard |
Are data elements consistently defined and understood? |
Gender code = M, F, U in one system and |
|
|
Consistent |
Do values represent the same meaning across systems and files? |
Profit margin is calculated differently across units, using two different formulas |
|
|
Content within data structures |
Complete |
Is all necessary data present? |
72% of Company Registration Number is blank |
|
Accurate |
Does the data accurately represent reality or a verifiable source? |
A Supplier is listed as ‘Active’ but went out of business six years ago |
|
|
Valid |
Do data values fall within acceptable ranges defined by the business? |
Transaction Data = 02/07/1982, but business started in 2001 |
|
|
Fit for purpose |
Is the information valuable to the business? |
A person has a SIC code |
Dimensions of Big Data Quality
Big data creates new challenges and adds new dimensions that go beyond accuracy, completeness, consistency, timeliness, validity, and relevance. Our post on the new dimensions of big data quality offers a more complete picture of complex data sets.
Why do I need to use a data profiling tool?
Data quality issues are frequently poorly understood and hidden. Data profiling tools provide results that are up to 90% more accurate in a fraction of the time taken to deliver similar insights using SQL or BI tools
Data profiling tools, like Trillium, automatically discover data quality issues inherent in the data. They make unknown data issues visible to business data stewards and other stakeholders, quickly and at scale. This keeps business owners in the loop and helps to build integrity
By contrast. SQL approaches depend on (typically) junior technical resources asking the right questions. Questions that they do not have the subject matter expertise to ask.
Learn more about the importance of data governance to support data quality initiatives, and about the broader role of a data quality assurance framework.
Got a Question? 
