Comment définir des politiques de qualité des données ?

ROI assessment on DQ engagements is something that I've seen missing in most of the Data Management Programs. DQ rules might get obsolete, lose their relevance over time or might not be tuned to Business KPIs. A proper dollar figure needs to be attached with each Business/Functional KPIs or usecases to understand the impact of bad data in terms of financial loss, only then we can do informed investment with tailored solutions in DQ programs which will bring an allover positive ROI to the organization from an Enterprise Data Management standpoint

It is imperative to understand and introspect the current state of data before implementing a DQ program. There may still be some DQ happening in the current processes and systems, but whether it is sufficient and relevant to business context, brings value and aligns to the organisational goals and objectives is the point of concern here. Data Quality is an integral part of the Data Governance program and should align with the strategic vision.

In data governance, assessing quality needs isn't a one-time task. Business goals and technology evolve; your data quality strategy should too. That's why I advocate for quarterly 'Pulse Checks.' A good data governance leader or team should revisit their data sources, consumers, and key processes. Is everything aligned with your current objectives? Are there new variables you haven’t considered? Doing these checks adds a dynamic layer to your static data assessment. Your goal is not to just check a boxes. You really want to ensure that you're in tune with business needs. Adaptability is as vital as accuracy. So don't just assess your data quality needs—continuously reassess them.

To set a data quality policy, define objectives, stakeholders, metrics, standards, data profiling, cleansing, documentation, governance, monitoring, improvement, security, compliance, training, communication, reporting, and regular adaptation.

In our data-driven age, data as an organizational asset helps build the foundation of trust and success, and so ensuring the quality of your data is paramount for any organization. Here's a concise guide: 1. Define data quality objectives: Pinpoint what it means for your organization. 2. Assess existing data: Identify issues and inconsistencies. 3. Data Governance Framework: Assign roles and responsibilities for data quality. 4. Data Quality Policies: Craft clear policies that maintain high standards. 5. Training and Awareness: Educate your teams on the importance of data quality. 6. Continuous Monitoring: Regularly audit data for improvements. Remember, it's an ongoing journey. Refining policies to adapt to changing needs is key.

Identify relevant dimensions top-down: from business strategy, through process KPIs, to the most critical data defects. This approach ensures that each data quality dimension directly supports your business objectives. For transparency, it is crucial to have clear methodologies and definitions that everyone understands. This clarity allows for effective communication and alignment across the organization, ensuring that everyone works towards the same data quality targets. It is not just about having policies, it is about having policies that are lived and breathed by the entire team.

Clearly state the objectives of your data quality policies. What do you want to achieve in terms of data accuracy, completeness, consistency, and timeliness?

Defining data quality dimensions is a critical step in any data management strategy and aligning them to your specific business objectives is the key. It's the foundation upon which you build trust in your data and make informed decisions. Moreover, setting clear thresholds and objectives is crucial. It not only guides data collection and processing but also helps you measure progress, facilitate remediation of bad data and determine the effectiveness of data quality initiatives. In today's data-driven world, where data is at the heart of decision-making, defining and consistently monitoring data quality dimensions is a fundamental part of ensuring that the data you rely on is not just accurate but also actionable.

Absolutely, the strategic importance of clearly defining and prioritizing data quality dimensions cannot be overstated. By setting specific targets and thresholds, organizations not only ensure the reliability of their data but also align their data management practices with their strategic objectives.

The dimensions have to measurable characteristics of data and may change from one organisation to another. The most used and common are accuracy, completeness, uniqueness, validity and timelines.

Todo dado deve passar por algum tipo de chancela, isso trás mais confiança para todo setor de Analytics, hoje praticamos uma avaliação em duas camadas, negócios e técnica e esse alinhamento gera resultados incríveis. Podemos até nos basear em modelos pré existentes, mas somente alinha com nossas necessidades internas chegaremos ao patamar ideal de qualidade.

Establishing data roles, defining their responsibilities, and their interplay is crucial. Especially, the concept of ownership for data and its quality is key: thus, the role of a "data owner" is mission-critical. While roles are important, their introduction needs to be handled with care: at best, there is no new position established but rather existing structures, roles, and boards are enhanced with data quality responsibilities. This "non-invasive" approache ensures a lean data organization, which is closely intertwined with the existing organization - leading to a high understanding of the business needs and a quick roll-out.

In any organization, establishing data quality roles is crucial and is one of the foundational building blocks for successful data governance. Data owners, stewards, custodians, users, and quality managers all contribute to data quality. Owners are responsible for the quality and integrity of data within their respective domains. Data Stewards play a hands-on role in managing and curating data while being responsible for day-to-day data quality tasks. Custodians protect data, users provide feedback acting as good data citizens, and quality managers oversee the process. A framework like RACI streamlines responsibilities, promoting informed decision-making. and compliance in an increasingly data-centric business landscape.

Utilizing a framework like RACI simplifies the understanding of responsibilities and ensures accountability across the board. Clarifying the scope and authority of roles like data stewards and quality managers helps in creating an environment where data quality is a collective responsibility. This structured approach not only enhances data integrity but also fosters a culture of proactive data management. Great insights on building a robust data governance structure! #DataQuality #RACIFramework #DataManagement

I think this is where a lot of us struggle when standing up our Data Governance Organization or in a relatively immature DG Organization. Initially we tried to incorporate data quality roles within our data management organization and quickly found out that DQ roles sit better in governance.

I don’t think it is necessary to have a separate policy on data quality. Data quality can be covered by an umbrella data governance policy just fine. From my experience, it is important that, through established communication lines for data governance and constant work on raising data literacy and awareness, dedicated data experts from different teams work on definitions and rules. These should then be regularly monitored, and issues effectively addressed, strongly supported by the management of the organization.

Quando falamos de qualidade de dados, criar um próprio modelo pode ser algo trabalhoso e levar tempo. Acredito que o melhor caminho é apoiar-se em estudos como DmBOK . O ponto de atenção aqui é entender que nem sempre é necessário usa-lo por completo, destaque aquilo que mais faz sentido pra você em seu momento com dados.

It helps to have a formal data quality plan. The plan outlines: 1️⃣ Formal quality thresholds and measures 2️⃣ Descriptions of data controls in place 3️⃣ Roles involved in maintaining quality

A clear and concise documentation serves as a blueprint for all stakeholders, ensuring everyone understands their roles and responsibilities. Including specific dimensions, weights, and thresholds helps in setting concrete targets and standards for data quality. The use of templates like those from DAMA International can significantly streamline this process, providing a structured approach to policy development. This documentation acts as a guiding light for ongoing data quality assessment, improvement, and reporting. Great advice for organizations aiming to establish a strong foundation in data quality management! #DataQuality #DataGovernance #DataManagement

Aqui está um outro ponto crucial quando falamos de dados em geral, não somente sobre qualidade. Precisamos entender os níveis que temos em nossa corporação e comunicar de forma adequada com cada uma dela. Criar um modelo de maturidade em dados ajuda nesse processo e pode nos ajudar a entender como evoluir passo a passo em cada unidade de negócio.

Deploying tools like data quality software and dashboards is just as important as ensuring all stakeholders are well-trained and aligned with these policies. The ADKAR model is a fantastic framework for managing this change effectively, focusing on awareness, desire, knowledge, ability, and reinforcement. Creating a culture where data quality is valued, understood, and actively pursued by everyone in the organization is crucial. Excellent approach to making data quality a cornerstone of organizational practice! #DataQuality #ChangeManagement #ADKAR #DataGovernance

Successfully implementing and communicating data quality policies across an organization is crucial for enhancing data management practices. Developing data quality solutions and ensuring that all stakeholders are well-informed about their roles and responsibilities has been a key aspect of our strategy. By integrating data quality software and dashboards, and through comprehensive training sessions, we equip our team with the necessary tools and knowledge to uphold data standards. Additionally, fostering a culture that values data quality through incentives and clear feedback mechanisms supports ongoing compliance and improvement, ensuring that our data management efforts are both effective and sustainable.

Data Quality Policies Data quality policies encompass a comprehensive set of guidelines, procedures, and standards established by an organization to ensure the accuracy, completeness, reliability, and consistency of its data. These policies are pivotal in upholding the integrity of data and facilitating well-informed decision-making. Here are the key components typically included in data quality policies: 1. Data Definition and Classification: • Clearly define what constitutes data for the organization. • Classify data based on sensitivity, criticality, and purpose, specifying which components are Personally Identifiable Information (PII) and critical.

One should review and update data quality policies when the organization's business or data landscape changes, or when stakeholders raise concerns. Review annually or biannually to ensure policies are up-to-date.

The dynamic nature of data and business environments means that data quality policies need regular reviews and updates. Implementing a robust process for this, as you've suggested, is crucial. Collecting and analyzing data quality metrics provides a data-driven foundation for policy refinement. Moreover, incorporating stakeholder feedback ensures that policies remain relevant and effective across different areas of the business. The PDCA model is an excellent framework for this continuous improvement cycle, promoting proactive adaptation to new challenges and opportunities in data management. Great insights! #DataQuality #ContinuousImprovement #PDCA #DataGovernance

2. Data Ownership and Stewardship: • Explicitly assign responsibilities for data quality. • Define roles such as data owners (responsible for the accuracy and integrity of specific data sets) and data stewards (responsible for overseeing the overall quality of data within the organization). 3. Data Collection and Entry Standards: • Establish standards for how data is collected, entered, and recorded. • This may include guidelines for data formats, validation rules, and data entry procedures. It's important to strike a balance to maintain validation rules without unduly slowing down the process.

1. Document and Socialize your Data policy with both business and technical stakeholders 2. Incorporate the policy in Architecture 3. Enforce the policy as code in Implementation 4. Monitor /Audit policy in a given period of time

This article seems to say everything about data quality instead of just focusing on a policy. If your company already has policies, it's good to have a policy around data and follow the company template. Policies often need a high-level executive approval, so if you get too detailed in the policy, you're going to be constantly updating it and constantly sending it for approval. If the company doesn't have policies, you probably don't need a data policy. Getting the detail you need in processes can be very effective. If there's no policy template and they really want one, I focus on why - why do we care about data. Also think if you need a data quality policy, or if you really need a more general data policy.

Set Data Quality Standard Define and set data standard for each data quality metric. For example, if you're setting a standard for data accuracy, specify what level of accuracy is acceptable (e.g., 98% accuracy).

Beyond the foundational steps, consider the evolving nature of data and technology. Integration of advanced analytics and AI can further enhance data quality management. Additionally, it's vital to foster a data-centric culture where every team member understands the importance of data quality. This involves regular training and awareness programs. Also, consider the impact of external factors such as regulatory changes and industry trends on your data quality framework. Finally, ensure scalability in your policies to accommodate future growth and technological advancements. #DataQuality #DataGovernance #DataCulture