How can you improve data literacy for data users and consumers?

Defining data quality standards might be just one step closure to make data readable and reliable. However, in my opinion, data literacy can only be achieved by playing more and more with data you have, e.g., profiling it, finding out patterns and correlations in various data points available. Further, to accelerate the data learning process, data engineering plays an important role in an organisation. The data should be cleansed, standardised, organised and labelled for easy understanding.

While creating data quality standards to align with respective risk dimensions of accuracy, completeness, consistency etc. its equally important to understand where and how should data citizens apply them. Origination data quality rules apply at the system of origin (SOO) and Transmission data quality rules could apply as the data moves throughout the supply chain of SOO to SOR (System of Record) or SOR to ADS (Authorized Data Source) or other downstream data consumers. When designing some of these rules it’s important to assess the nature of these rules - A preventive rule will stop bad data from entering into the system whereas a detective rule is used to identify bad data that is in the system which may require remediation.

Actually I don't agree that defining data quality standards is the first step to improving data literacy. I would take a step back and look at the current data maturity state. Outside of data and technology, how are the employees using, interpreting and analyzing data? Do they have a clear understanding on origin of data? Do they ask qualifying questions on the data set before reporting on it? Secondly, I would look at the data and business strategy and identify the direction and use case valuation that will drive the acceleration of data literacy within the organization. As a data leader, I would start by building a culture that focuses on data-driven analysis and results. After these steps, I would tactically look at data quality rules.

I disagree that this is the first step in improving an organization's data literacy. First, I would recommend identifying the various roles and responsibilities of all those who use or create data - there will be different roles and will have different and similar responsibilities. Know these roles and their current data activities, it is important to evaluate their current skills in creating and / or consuming data. What are they doing well, what needs to be changed - and why - and how is this affecting the organization now? Defining data quality standards is a much later step in this process.

The first step in improving data literacy among users and consumers is to define what constitutes high-quality data within your organization. Establish specific, comprehensive standards for data accuracy, completeness, consistency, and timeliness. Communicate these standards to all data stakeholders so they understand the requirements and the role these standards play in making data-driven decisions. This foundation will enable users not only to recognize high-quality data but also to understand the processes involved in achieving and maintaining these standards.

Most organizations do not engage in training for data quality - not documenting and implementing DQ expectations and rules, not stressing the importance of high data quality for the organization's success (and compliance where appropriate), and not offering the right training for business people (especially data stewards). DQ training should be part of every onboarding training for all levels, including leaders, and should be revised and presented regularly.

It is crucial to provide comprehensive training programs tailored to the needs of different user groups within the organization. Training should focus on the basics of data quality, such as how to identify and rectify data issues, as well as more advanced topics tailored to specific roles or departments. Incorporate hands-on sessions where users can practice using data quality techniques and techniques. Regularly updated training ensures that all personnel are informed about the best practices and new developments in data management.

An alternative approach to improve data literacy could involve leveraging peer-to-peer learning and mentorship. In this setup, individuals who are more experienced in data handling could guide less experienced team members. This allows for hands-on learning and real-time feedback, which can often lead to a more practical understanding of data quality issues. Additionally, incorporating data quality discussions and problem-solving into regular team meetings can also reinforce its importance and keep it top of mind.

Once the current state of data literacy is established, the next step is to identify the desired level of proficiency and competence that the organisation aims to achieve. This involves aligning data literacy goals with business objectives and identifying the specific skills and capabilities required to meet those goals. Based on the assessment findings and the desired outcomes, a tailored data literacy programme can then be designed to address the identified gaps and empower employees with the necessary skills and knowledge. This programme may include training sessions, workshops, hands-on exercises, and access to resources and tools aimed at improving data literacy skills.

Enhancing data literacy through comprehensive data quality training has been a fundamental aspect of empowering our teams to handle data more effectively. By educating our data users and consumers on critical topics like data validation, cleaning, and analysis, we equip them with the necessary skills to identify and address data quality issues proactively. This training not only builds confidence in using data but also instills a sense of responsibility for maintaining its integrity. The impact is clear: teams that understand data quality contribute to better decision-making processes and more reliable outcomes, reinforcing the value of data as a strategic asset.

Provide your users with the best tools to maintain and enhance their data quality. Utilize intuitive data quality software that helps users in their daily activities, such as data validation, cleansing, and analysis. Ensure that these tools are seamlessly integrated into the systems and processes they already employ to minimize disruption and encourage widespread adoption. Provide training on how to use these tools effectively and ensure that support is available to assist users in troubleshooting problems.

Choosing a tool for data literacy means knowing what you want to accomplish in data quality and in data governance and in metadata management. One tool may be able to address aspects of each of these functions, but the tool choice should come after the evaluation of roles / responsibilities / challenges and after developing basic data standards and a few policies.

Develop comprehensive documentation that serves as a tool for users seeking to understand data quality processes and standards. This documentation should be accessible and include guidelines, policies, and best practices regarding data quality. Consider establishing a centralized repository where all data quality resources are stored, such as case studies, FAQs, and troubleshooting guides. Regular updates and clear, practical examples will make this document a valuable tool for improving data literacy.

Business units / Federated Data Offices must ensure appropriate change management processes are in place to ensure the Data Quality Inventory is periodically refreshed and current with the corresponding implementation in the Data Quality tool. Consumers of the data / business process owners should be engaged in reviewing the rule results (Breach Reports / Dashboards) and thereby opining on rule enhancements or proposing additional rules that might be needed to improve the overall data quality.

Data quality documentation processes should be part of the organization's overall approach to data management - and the processes should be part of the organization's enterprise data quality and data governance program(s). There are different kinds of data quality documentation and each organization should review the different kinds and choose which ones address their needs now and in the future. Good documentation helps every data creator and consumer meet the organization's needs for higher quality data.

The pursuit of a culture that values and prioritizes data quality is essential for improving data literacy. This involves more than just training and tools; it requires a shift in mindset throughout the organization. Celebrate successes in data quality enhancements and make data quality a key component of performance metrics and reviews. By incorporating data quality into the organizational culture, you can ensure that it becomes a natural part of daily work tasks.

We need to go above and beyond just data quality when it comes to data literacy. Data, and curation of that data is more than just the quality aspext. Think about metadata, lineage, master data, reference data - also, the usage of that data. Many times we hear that "the data is bad" but in reality, the data is correct, but it's not fit for the purpose of that person's use case/query. Long story short, data literacy is so much more than just data quality.

Actually, developing a data quality culture should be part of the process from the beginning - not just "step five". Data literacy is more than data quality - and the need for the organization to have higher data literacy is dependent on many other factors. Data governance policies, the use of trained data stewards, the collection and availability of metadata (especially business data definitions), and so much more. A data culture values all the aspects of data management throughout the enterprise.

I've seen some analysts & engineers get frustrated when users surface data quality issues. While I am understanding of this, it's important to note that improving overall data quality is like a team sport. There's no change of maintaining high, or even mediocre standards of data quality with only a few people checking for issues. Data teams need to be champions for improving data quality and express gratitude, excitement, encouragement, etc. to those who surface those issues - positive reinforcement goes a long way. Not reinforcing the surfacing of data quality issues also risks the culture not being reinforced throughout the company. Not to mention the mental health benefits of expressing gratitude and having a positive attitude!

Don't forget the need to assess / evaluate, analyze, and improve METADATA quality as well as the quality of the actual data values. Most experts feel that over 70% of DQ issues arise from missing or incomplete or erroneous metadata (business and technical). Metadata management should be part of every DQ effort.

Data literacy should be more than just data quality management. The purpose of data literacy should be to help data users and consumers get value out of data and promote self-service data usage. There are many other foundational management capabilities that will be required to enhance data literacy. For example: 1. Metadata Management to provide visibility of what data is available and where to get the data from. 2. Data Governance to ensure the right people, and processes are in place for data decision making, issues resolution and to support data consumers

In my experience, there we often overlook "last mile" data quality. This can encompass embedding rules in platforms where business code is deployed all the way through intent to execution in spreadsheets, reporting and BI.