6 minute read 26 Aug 2021

Data Quality Management has become a must. But what does that mean and how do you create actionable insight into the quality of that data?

5 tips to properly integrate Data Quality Management into your organization

Authors
Patrice Latinne

EY Belgium Data & Analytics Leader, Financial Services

Passionate leader in broad data science and artificial intelligence (AI) systems. Energized by team empowerment, success and focus on client satisfaction. Married and father of three.

Ali El Maghraoui

EY Belgium Financial Services Data & Analytics, Information Strategy Executive Director

Passionate about Data & Analytics, Technology & Innovation. Focused on enabling organizations to unleash the full potential of their data. Obsessed by client satisfaction. Proud husband and father.

Riccardo Magnani

EY Belgium Consulting Associate Partner

Great passion for new challenges and curiosity for new professional adventures.

6 minute read 26 Aug 2021

Data Quality Management has become a must. But what does that mean and how do you create actionable insight into the quality of that data?

In brief

  • Data Quality is the data’s ability to serve its purpose and can be measured through a series of dimensions.
  • Following some tips and tricks can help you succeed in implementing Data Quality Management within your company.

Most organizations continue to collect, store and manage more and more data. Since they make decisions and take action based on  information and insight coming from this data, poor data quality can have disastrous consequences….

This begs the question: what is “poor (or high) quality data”, and how do you measure it?

This article sheds a light on the topic and gives some tips & tricks for successful Data Quality Management in your organization.

1. What is Data Quality and how do you measure it?

Data Quality is the assessment of the data against its purpose, and its ability to serve that purpose.

In other words, data quality depends on context and on the needs of the data consumer.

To optimize it, proper Data Quality Management is necessary. This can be defined as “a set of roles and responsibilities, organization setup, processes, procedures and policies used to define, measure, control, maintain and improve data quality”. It helps determine whether your data is good enough to use.

Defining…

Organizations collect high amounts of data, and not all of it is of equal importance. The best practice is thus to focus your effort on defining your critical data and its quality criteria. This means selecting the data that has the most impact on your (business) needs and is the most important to your organization and data consumers.

Measuring and controlling …

Data quality can be measured and controlled along the whole lifecycle of your data, using Data Quality dimensions. These are measurable features that facilitate better communication among data actors about data quality objectives, challenges and remediation approaches.

As data quality depends on the context and the needs of its consumers, it is up to every organization to select the most relevant dimensions

The most common Data Quality Dimensions are the following:

  • Completeness indicates whether all the data needed has an assigned value and is available to be used. It is not necessarily about having 100% of your data fields completed, but the ones that are critical for the purpose for which it is used.
  • Uniqueness indicates whether duplicate records exist. There is a higher risk of duplication when combining data sets. This is why it is important to check this dimension, as unique records build trust in the data.
  • Timeliness indicates the degree to which data represents reality from the required point in time. This is particularly crucial when information is highly time-sensitive. For example, during the pandemic, data with high-quality timeliness led to more responsive health care provisions and saved lives.
  • Consistency indicates the degree to which data is consistent between two or more systems/data sets. This increases the possibility to link data from multiple sources which improves the usefulness of your data.
  • Accuracy indicates the degree to which data correctly describes the “real-world” object. This dimension is quite challenging to monitor, but high data accuracy facilitates trusted analytics and confident decision-making.
  • Validity indicates whether the data conforms to the defined syntax (format, type, range). However, we have to keep in mind that “valid” values don’t necessarily mean “accurate” values.

Maintaining and Improving…

Of course, Data Quality Management does not end here. Once you have measured your data quality, you can identify the “poor quality” data and the issues that are keeping your organization from achieving its business objectives. Data that does not meet its purpose represents a risk and cost for your organization, hence the importance of identifying the nature, root cause and impact of the issue. Once defined and prioritized, you can tackle the root causes by implementing remediation plans and actions to, ultimately, improve data quality.

2. Tips & tricks for an effective Data Quality Management

Once you know what Data Quality Management means and requires, it all comes down to one question: what makes a Data Quality Management program succeed or fail?

TIP # 1 Tailor it to your needs

Every organization is different, which means needs, challenges and priorities vary from one business to another. Similarly, each organization needs to design its own Data Quality Management, taking into account its constraints and realities. This journey starts by correctly identifying what is critical to realize business value and what is not, and by identifying the most valuable data as well as defining their minimum quality requirements.

TIP # 2 Define and rollout your Data Quality framework

This may seem obvious, but you should take time to design simple, repeatable and scalable practices, processes and ways of working that are fit for your purpose. Define clear roles and responsibilities for data quality, and take time to make sure your people can develop the necessary skills. Also, you can set up appropriate ceremonies to regularly review your critical data quality, discuss the common issues, recognize and share good data quality practices. This will foster a good visibility on the situation and fast update of practices when necessary.

TIP # 3 Involve all your stakeholders

Data Quality Management is not a stand-alone activity, and will be most effective when integrated to your general way of working, processes and procedures. Everyone in your organization should be – and should feel – involved. Issues affecting the quality of your  data can occur during its whole lifecycle from data entry to data collection, use and reporting. This means that everyone should understand their role and responsibility in reaching high-quality data. Everyone who uses data to perform their job within your organization need to drive this.

TIP # 4 See it as a culture, not a project

As for your data management as a whole, your Data Quality Management is not a project, it is a work culture. It is made of iterative and repeatable activities that cannot exist in isolation, but must be fully integrated within your business. As mentioned under TIP # 3, it stems from a true commitment, not only from the leadership team, but from people of all levels. Data Quality Management should become a cross-functional engagement of your whole organization, as high or poor quality data impacts various areas of your company. This will also foster a proactive approach, rather than a reaction to the identification of bad quality data.

TIP # 5 Be transparent

Transparency leads to trust. We encourage you to be transparent about your data, create a culture where poor data quality can be talked about, where people can share their experience with quality problems and the solutions they implemented. Also, communicating on your data quality practices and how they impact your business helps everyone understand why all of this matters, and feel involved in the program. By being transparent, you can reach a quality mindset and, in time, a culture where high data quality is achieved from the start.

3. Conclusion

Much like any endeavor one wishes to carry out, Data Quality Management rests on one crucial basis: purpose. There is no such thing as reaching your destination if you do not know first where you are going. Why do you need the data? What goal does it serve? Only by answering these questions can you strive to reach high-quality data. And once you have done this, the real journey begins: fostering a culture that onboards all internal stakeholders on the way to high-quality data.

Newsletters EY Belgium

Subscribe to one of our newsletters and stay up to date of our latest news, insights, events or more. 

Subscribe

Summary

Data Quality is key for organizations to make good decisions. This calls for proper Data Quality Management, which can only be reached through an accurate view on relevant dimensions and a widespread culture within the business.

About this article

Authors
Patrice Latinne

EY Belgium Data & Analytics Leader, Financial Services

Passionate leader in broad data science and artificial intelligence (AI) systems. Energized by team empowerment, success and focus on client satisfaction. Married and father of three.

Ali El Maghraoui

EY Belgium Financial Services Data & Analytics, Information Strategy Executive Director

Passionate about Data & Analytics, Technology & Innovation. Focused on enabling organizations to unleash the full potential of their data. Obsessed by client satisfaction. Proud husband and father.

Riccardo Magnani

EY Belgium Consulting Associate Partner

Great passion for new challenges and curiosity for new professional adventures.