Data Quality ROI: The ‘Why’ and ‘Who’ of Data Quality with Gaurav Patole

Executive Summary

This webinar outlines the key concepts from the book “Data Quality ROI,” tracing Gaurav Patole’s journey in the realm of Data Quality. Guarav Patole begins with the genesis of the book and defines Data Quality as “fit for purpose.” He then explores the complexities of Data Quality through a three-way struggle, challenging common myths surrounding it. Incorporating Newton’s laws, the book emphasises the importance of action (First Law), partnership (Second Law), clear communication (Third Law), and engagement strategies (Fourth Law) in fostering a robust Data Quality culture. It also discusses the transformative power of dashboards by aligning metrics with business language and impact. Furthermore, the webinar highlights the role of AI in enhancing Data Quality and advocates for the use of balanced scorecards to connect data initiatives with business outcomes. Lastly, it concludes with an invitation to explore the book for comprehensive insights and actionable strategies.

Webinar Details

Title: Data Quality ROI: The ‘Why’ and ‘Who’ of Data Quality with Gaurav Patole
Date: 2025-11-10
Presenter: Gaurav Patole
Meetup Group: DAMA SA User Group Meeting
Write-up Author: Howard Diesel

Data Quality ROI and Gaurav Patole’s Journey

Gaurav Patole serves as a Principal Data Strategist at ThoughtWorks in Amsterdam, bringing over a decade of consulting experience in data-related fields. His expertise encompasses Data Governance, Data Quality, and Data Strategy, reflecting a comprehensive understanding of the challenges organisations face in managing their data assets effectively. Gaurav’s journey into writing began five years ago, initially through blogs and expanding to LinkedIn, where he has received notable recognition from the data community.

Through his diverse work with multiple clients on short-term projects, Gaurav has gained valuable insights into varying data cultures. He has identified common challenges, particularly those related to Data Quality, ownership, and implementation. His experience highlights the critical need for organisations to adopt robust data strategies that not only address these challenges but also foster a culture of accountability in data management. Overall, Gaurav’s contributions and observations continue to inspire improvements in data practices across different industries.

Gaurav shared that his consultancy background exposed him to diverse industries and company structures, providing unique insights into how different organisations struggle with similar Data Quality issues. This cross-industry perspective became the foundation for his book, which addresses not just the “how” of Data Quality, but more importantly, the “why” and “who.”

Figure 1 Data Quality ROI

The Genesis of the Book: Why Data Quality ROI?

Gaurav’s motivation for writing the book stems from his observation of a recurring issue in organisations regarding Data Quality initiatives. Often, these efforts exist in silos, where business units assume that IT is solely responsible for managing Data Quality. Conversely, IT teams feel that they cannot take ownership of data initiatives without a solid understanding of the business context. This disconnect highlights a significant challenge in fostering effective data management practices within organisations.

Despite attending numerous conferences where industry experts delve into the intricacies of Data Quality, Gaurav noted that many discussions primarily focused on the “how”—covering frameworks, technologies, rules, and dashboards. However, these conversations frequently overlooked crucial foundational questions concerning the ownership and purpose of Data Quality initiatives. By addressing these gaps, Gaurav aims to encourage a more collaborative approach between business and IT, fostering a shared sense of responsibility and purpose in Data Quality management.

The inspiration for Data Quality ROI’s framework struck him while watching a documentary on Newton’s laws of motion. Gaurav realised that just as Newton created laws applicable to the physical world, there could be corresponding laws for the data world. This creative framework became the structure of his book, making complex Data Quality concepts accessible and memorable.

Gaurav emphasised that his book represents collective industry wisdom, shaped by conversations with clients, colleagues, and the broader data community—making it truly “our book” rather than his alone. Additionally, the book‘s unique approach aims to be practical and digestible, enabling readers to understand any chapter independently without needing to read it from start to finish.

Defining Data Quality: Fit for Purpose

Gaurav offers a nuanced perspective on Data Quality, defining it as “the degree to which we can create or deliver data that is accurate, usable or reusable, trusted, and solves the needs of consumers—essentially fit for their purpose.” This definition emphasises that Data Quality goes beyond mere metrics; it should reflect the specific needs of its end users. Rather than adopting a one-size-fits-all approach, organisations should consider how well the data aligns with the intended consumption context.

By framing Data Quality from a fit-for-consumption perspective, Gaurav challenges traditional views that treat Data Quality as a universal standard across the organisation. This shift in focus encourages businesses to assess data based on its relevance and applicability to specific consumer requirements. Ultimately, prioritising the needs of consumers in evaluating Data Quality can lead to more effective decision-making and enhanced overall satisfaction.

The concept of “fit for consumption” plays a crucial role in determining the necessary standards for specific datasets. Gaurav clarified that this term primarily addresses the volume aspect, emphasising that quality parameters should be determined based on actual use cases. For example, if a dataset consists of hundreds of fields, but only ten are essential for a particular application, then it is necessary to select only those ten fields. The focus of Data Quality efforts should be directed toward those ten critical fields.

This idea highlights the importance of prioritising Data Quality efforts to ensure that only the most relevant information is refined and maintained. By concentrating on what truly matters for specific use cases, organisations can allocate resources more effectively and enhance overall data usability. Ultimately, recognising the nuances within the “fit for consumption” framework allows for more strategic data management practices that align closely with practical needs.

Gaurav elaborated that accuracy ensures the data is correct, usability means it doesn’t require weeks of transformation before use, and trust indicates that someone in the organisation has certified the data’s reliability. This targeted approach prevents organisations from wasting resources trying to achieve 100% Data Quality across all data when only specific elements impact business outcomes.

Figure 2 Definition of Data Quality

The Data Quality Complication: A Three-Way Struggle

Data Quality issues often arise from the disconnect between business teams and IT departments, creating significant challenges within organisations. Business teams typically prioritise delivering results over addressing Data Quality, believing that data management falls squarely within the technical realm of IT. Consequently, they often fail to acknowledge their role in ensuring data integrity, viewing it as someone else’s responsibility.

Conversely, IT and data engineering teams recognise the importance of Data Quality but struggle to take ownership due to the lack of business context. While they can pinpoint technical errors such as missing values and formatting issues, they are uncertain about the relevance of these issues to business objectives. This misalignment underscores the need for collaboration between business and IT teams to establish a shared understanding of Data Quality responsibilities, ultimately enhancing both data integrity and business outcomes.

Third, the Data Governance team (CDO office) creates policies and mandates specific technologies; however, neither business nor IT knows how to apply these policies or utilise the prescribed tools. This creates a significant Data Quality gap—business leaders wonder what to do, IT thinks they should implement a technical solution, and the CDO team makes assumptions based on limited information.

The result is extensive meetings, siloed discussions, and “coffee cup meetings” with finger-pointing but no progress. This communication gap, not technical capability, represents the fundamental problem in Data Quality initiatives. Organisations become stuck in endless debates about ownership and responsibility, rather than addressing actual data issues that impact business outcomes.

Figure 3 The Data Quality Complication

Challenging Data Quality Beliefs and Myths

Gaurav highlighted several prevalent beliefs regarding Data Quality that were often taken for granted across various industries and conferences. One such belief was that “Data Quality can be achieved 100%,” which raised a critical question: was perfection a realistic goal, or even necessary for effective data management? Another common belief is that “Data Quality is a one-time cleansing activity.” This prompted further inquiry into the sustainability of data cleanliness over time; if data were cleaned today, would it remain clean tomorrow?

By critically examining these widely accepted beliefs, Gaurav encouraged a more nuanced understanding of Data Quality. He challenged the notion of absolute perfection and emphasised the need for ongoing efforts in data maintenance. Ultimately, these discussions aimed to foster a more realistic perspective on Data Quality, highlighting the importance of continuous improvement rather than viewing data cleansing as a finite task.

The third belief that “Data Quality is IT’s responsibility” led to a crucial question: Can IT really improve Data Quality without understanding what the data means? For example, IT might identify 10 active and 10 inactive statuses in a status field, but they can’t explain the business meaning of “active” versus “inactive.”

The “shift left” trend raises important questions about the responsibility of source systems in Data Quality management. Gaurav shared his concerns regarding whether source systems, primarily designed to meet operational needs, should also be tasked with addressing Data Quality issues for analytical purposes. These systems may lack the necessary skills, resources, and leadership support to handle this dual responsibility effectively. Ultimately, while the shift left approach emphasises tackling Data Quality at the source, it is essential to evaluate whether source systems are adequately equipped to fulfil this crucial role.

Finally, the belief that “buying a Data Quality tool” solves everything prompted the question: Is a tool really the answer? Will spending half a million dollars on fancy technology solve the communication challenges and organisational alignment issues previously discussed? Gaurav encouraged the audience to challenge such beliefs with critical thinking across all data management areas.

Figure 4 “They say … But ask yourself…”

Newton’s First Law: The Law of Data Inaction

Drawing inspiration from Newton’s Law of Inertia, Gaurav introduced a significant principle in Data Management. He asserts that just as an object at rest will remain in its current state unless acted upon by an external force, bad data will not rectify itself without proactive intervention from individuals within the organisation. This analogy emphasises the necessity of taking action to address Data Quality issues rather than assuming they will resolve themselves.

To drive this point home, clear communication about the importance of addressing bad data is essential. It is crucial to clearly convey the reasons behind the call to action and effectively illustrate the benefits to the business. By effectively communicating “what’s in it for them,” organisations can foster a culture of accountability and initiative, encouraging teams to take responsibility for maintaining high Data Quality standards.

This principle underpins the entire data ownership conversation happening daily in organisations. Why should a business team own anything related to data? Unless organisations communicate this in language that business leaders understand, they won’t fully engage. There’s a human reason for this resistance: most business leaders come from non-technical backgrounds—such as finance, chemistry, or other industry specialisations—not IT or software engineering. They’ve led their verticals for years with a particular mindset about their business.

Merging the technical mindset of IT/software teams with the business mindset requires communicating in a language both understand. Gaurav offered a banking example: start with the customer complaints department. This reveals use cases that demonstrate how poor Data Quality affects customer experience in mobile apps and during in-person bank visits. Making business leaders curious about how Data Quality ownership benefits them is key. It’s not just about improving business decisions—it drives overall efficiency when business contributes to organisational Data Quality management.

Figure 5 Newton’s First Law: the Law of Inertia

Figure 6 How does Data Quality Influence Areas of Business?

Newton’s Second Law: The Data Quality Partnership Matrix

Gaurav’s introduction of the second law underscores the importance of collaborative efforts in achieving significant improvements in Data Quality. He firmly emphasised that no single team—be it data, product, business, or IT—can accomplish this independently. Instead, a successful approach demands a partnership that integrates diverse perspectives and expertise.

To illustrate this concept, Gaurav presented the Data Quality Partnership Matrix, which includes two critical axes: business engagement and support. These axes represent the comprehensive involvement required from all data governance teams to effectively enhance Data Quality. In conclusion, fostering collaboration across teams is essential for driving improvements in Data Quality, demonstrating that teamwork is the cornerstone of success in this endeavour.

The matrix reveals four quadrants: When business engagement is high, but support is low, organisations experience “total chaos”—initiatives and talk, but no actual delivery. When business engagement is low, but support is high, there’s “unclear ownership”—data governance teams have budgets and tools, but business sponsorship is missing, resulting in wasted data potential. This is the most common scenario.

When both business engagement and support are low, the result is “talk and no show.” When business engagement is high, but support remains insufficient, organisations achieve “non-scalable Data Quality”—investments in certain projects or initiatives, but in siloed fashion. Each team pursues its own governance and quality efforts (such as multiple data catalogue initiatives across the organisation) without centralised guidance.

The ideal state is the upper-right quadrant: high business engagement plus high support equals “strategic Data Quality.” This occurs when business-sponsored use cases exist with clear ownership, alongside a centralised data team providing policies, technology, education, literacy support, and help with issue alignment. When both work in partnership, organisations yield results in Data Quality and data governance.

Figure 7 Newton’s Second Law: Force

Figure 8 Data Quality Partnership Matrix

Newton’s Third Law: Communication Over Technical Jargon

Gaurav highlighted a crucial principle in Data Management, paralleling it with Newton’s Third Law: “For every action, there is an equal and opposite reaction.” He emphasised that when technical jargon is excessively pushed onto business stakeholders, it often leads to confusion, frustration, and an influx of questions from them. This phenomenon highlights the importance of clarity and effective communication in data governance.

To further illustrate his point, Gaurav painted a vivid scenario of becoming a data owner overnight. He depicted the overwhelming experience of sifting through a multitude of Confluence or SharePoint documents laden with complex terminology—such as Data Quality rules, dimensions, accuracy, timeliness, lineage, federated computational governance, Collibra, and Informatica. This scenario underscores the importance of translating technical concepts into accessible language to ensure that all stakeholders are aligned and empowered in their understanding of Data Quality.

The natural reaction is often resistance, as business leaders often lack an understanding of this world. They don’t know why they should do it or what benefits they’ll receive. Gaurav displayed a typical Data Quality dashboard he had seen consistently over the past 12 years: total rules passed, total rules failed, rules fixed in the past 3-6 months, and dimensions such as accuracy, consistency, timeliness, and freshness. When shown to businesspeople unfamiliar with this world, their first reaction is, “So what should I do about it?”

For a customer data owner, does this mean the app will stop functioning tomorrow, or will customers unsubscribe? They don’t care because they don’t understand the message. On the surface, they’ll say “good work,” then return to their actual jobs. The core message is that Data Quality struggles not from a technical gap, but from a communication gap between multiple organisational actors. This insight became central to the book’s approach—transforming technical metrics into business-relevant insights.

Figure 9 Newton’s Third Law: Action Vs. Reaction

Transforming Dashboards: Speaking Business Language

A technical Data Quality dashboard can be transformed into a more accessible format for business stakeholders. By presenting information in clear terms, such as “70% decision confidence” with “accuracy” included in brackets for technical experts, he ensures that critical insights are easily understood across different audiences. Furthermore, the revised dashboard emphasises potential impacts, detailing the implications of a 70% score on key financial processes, including forecast accuracy, month-end close times, and delays in expense reimbursements.

This innovative approach not only bridges the gap between technical data and business language but also enhances decision-making capabilities within the organisation. By focusing on direct impacts, stakeholders can better grasp the significance of Data Quality metrics and their consequences on operational efficiency.

Understanding the relationship between system consistency and customer satisfaction is crucial for business leaders to achieve optimal results. For instance, when Gaurav pointed out that a mere 56% cross-system consistency can negatively impact customer Net Promoter Scores (NPS), it immediately piqued their interest.

Research indicates that when mismatches exceed 20%, the likelihood of a decline in NPS increases significantly. This statistic serves as a critical insight for decision-makers. Consequently, once business leaders grasp the implications of low consistency scores, their focus shifts to actionable solutions, prompting questions about how they can enhance these metrics. Ultimately, recognising the link between system performance and customer sentiment encourages proactive measures for improvement.

Technical teams can then explain that there may be a problem with the source system or that the transformations are not being captured correctly. They can request business help to prioritise this work in the coming weeks. The business leader is now on board as a data owner, recognising this metric’s importance and providing necessary investment, budget, or part-time resources. The key is transforming issues into business language, understanding their quality perception and important KPIs, then translating them into data points.

Figure 10 Finance Data Quality Monitoring Dashboard

Figure 11 Quality Assessment against Key KPIs

Bridging Metrics to Business Impact

Data Quality rules play a crucial role in illustrating the connection to business impacts. A participant underscored the importance of having approved Metadata that details which business processes are compromised when Data Quality rules fail. This documentation is vital, as it enables stakeholders to clearly understand the implications of these failures.

However, challenges arise when attempting to link failed Data Quality rules directly to business KPIs. Business leaders often seek explanations for discrepancies, but without documented connections, any conclusions drawn can seem purely speculative. Therefore, establishing clear connections between Data Quality failures and business outcomes through metadata is essential for informed decision-making and accountability.

Gaurav agreed, explaining there’s a layered approach not fully shown in the presentation due to time constraints. The “36 failed rules” break down into specific indicators: 15% missing records and 13% data dropouts. The next layer asks: if records are missing, what does it mean? It means potential impact on forecast accuracy. But to what level?

This requires business analysts or data stewards to have conversations with business leaders: “What’s your ideal threshold for forecast accuracy? When do you see a problem?” Business might respond that 80-100% is acceptable, but below that threshold requires attention. These conversations inform KPI calculations—it’s a layer-by-layer approach, not a direct jump.

Cost impact is a crucial consideration in discussions about dashboard effectiveness, particularly regarding error costs that are not typically displayed. One participant highlighted this concern, while Gaurav emphasised that cost can indeed function as a Key Performance Indicator (KPI) for certain business units. For example, in compliance scenarios, the extended time required to achieve GDPR or BCBS compliance—which can increase from three to six months—has significant financial implications. Additionally, the conversation underscored the vital role of data catalogues, which serve as a centralised hub for quality Metadata, notifications, and communication between business and IT. Ultimately, understanding these factors can enhance organisational decision-making and performance.

Implementation: Roles, Responsibilities, and Starting Points

Clearly defining roles and responsibilities is essential for effective Data Quality management, as emphasised by a participant who advocated for the use of RACI (Responsible, Accountable, Consulted, Informed) matrices. They highlighted that in their organisation, the responsibility for Data Quality initiatives often shifts too easily, resulting in a lack of accountability among team members. Additionally, while governance structures are vital for oversight, they frequently fall short in ensuring that Data Quality measures are implemented at the source. Therefore, establishing clear definitions and strong governance is crucial to fostering a culture of accountability and enhancing Data Quality initiatives.

Ultimately, a robust governance framework, which includes the proper delineation of roles and responsibilities, is essential for improving Data Quality. By addressing these foundational challenges, organisations can foster a culture of accountability and effectively enhance the integrity of their data systems. This approach not only mitigates issues related to ownership but also empowers teams to take proactive steps in maintaining high standards of Data Quality right from its origin.

Gaurav acknowledged that while he mentions “business team” or “IT team” for a broader understanding, organisations must filter this into their specific structure and hierarchy. He personally prefers RACI matrices for clarity. However, the trend of pushing ownership to business is only correct if that responsibility is crystal clear—not just a one-page document stating, “You’re responsible for Data Quality.”

Establishing a clear RACI (Responsible, Accountable, Consulted, Informed) framework is crucial for effective project management and collaboration. This framework should be detailed and accompanied by comprehensive education, support, and enablement to ensure all participants understand their roles. Specifically, it is essential to educate individuals designated in the RACI so they can clearly grasp their responsibilities and expectations. By prioritising this understanding, teams can enhance their efficiency and accountability, ultimately leading to successful project outcomes.

A participant asked whether to select roles first or fix quality with IT first. Gaurav emphasised that either approach must be linked to a specific use case or business outcome. Data Quality shouldn’t be a siloed exercise—it must connect to business outcomes. Whether creating roles or cleansing data, there must be an underlying business reason. When that outcome is clear (e.g., CRM cleansing, analytical use cases), you naturally break it down into steps, identifying necessary roles, understanding needs, enabling requirements, policies, and technology.

Newton’s Fourth Law & Engagement Mantras

Gaurav introduced a fundamental principle of Data Quality that highlights the intrinsic link between attractive forces in physics and organisational data practices: “All objects with mass attract one another.” This principle emphasises that when organisations effectively implement the foundational laws of Data Quality, they foster an environment where Data Quality initiatives become a natural focus rather than a burdensome obligation. Consequently, instead of facing resistance driven by management’s demands, the pursuit of high-quality data evolves into an inherent motivation for the business, leading to more sustainable and effective practices.

The shift from a push to a pull perspective in Data Quality initiatives signifies a more organic integration of these efforts within the organisation. As businesses cultivate an environment that prioritises Data Quality through these frameworks, they not only enhance operational efficiency but also foster a culture that values high-quality data. Ultimately, this transformation leads to more robust decision-making processes and improved overall performance.

In the realm of effective business engagement, embracing key strategies fosters a collaborative and innovative environment. Firstly, organisations should focus on creating a culture of pull rather than push, encouraging teams to seek solutions rather than feeling pressured. This involves avoiding finger-pointing and public shaming, as no one enjoys being singled out among peers. Instead, Data Quality issues should be viewed as opportunities to enhance efficiency, profitability, and the customer experience. Educating rather than dictating is crucial, as data governance and IT teams must empower business owners to take charge and support them throughout the onboarding process.

Moreover, fostering collaboration is essential for successful outcomes. Bringing together producers, transformers, and consumers through event storming workshops can facilitate valuable discussions and generate insightful findings. Identifying the “number twos,” or key individuals each leader relies on for data conversations, can also bridge significant gaps in understanding. Revising technical jargon into business-friendly language can enhance clarity, while transparently sharing feedback incentivises active participation and highlights success stories. Finally, leveraging AI thoughtfully enables automation that simplifies operations without compromising human oversight, resulting in a more cohesive and efficient team. oversight

Figure 12 Newton’s Law of Universal Gravitation

Figure 13 Some Key Engagement Mantras

The Role of AI in Data Quality

Gaurav highlighted the crucial impact of artificial intelligence on improving Data Quality management in his work. His book includes two dedicated chapters that explore this relationship: “Data Quality for AI” and “AI for Data Quality.” These chapters explore how AI technologies can enhance Data Quality processes by automating tasks and identifying anomalies, while also emphasising the importance of high-quality data for the successful implementation of AI systems.

In addition to providing insights into the capabilities of AI, Gaurav’s chapters emphasise a reciprocal relationship between Data Quality and artificial intelligence. By illustrating how effective Data Quality management enhances AI performance, he makes a compelling case for the symbiosis of these two fields. Ultimately, his examination reveals that prioritising Data Quality is essential not only for the capabilities of AI but also for optimising overall outcomes in data-driven environments.

To effectively implement AI in Data Quality, human oversight is crucial to ensure accuracy and relevance. While a human-in-the-loop approach is necessary for certain tasks, many routine responsibilities currently handled by Data Quality consultants and analysts can be efficiently automated. Ultimately, leveraging AI in this way not only streamlines operations but also empowers professionals to focus on more strategic initiatives.

AI plays a crucial role in enhancing communication and efficiency in Data Management by providing valuable tools and insights that improve Data Management. It offers Data Quality rule suggestions, generates and suggests metadata, and improves dashboards with more effective visualisations. Additionally, AI enables co-pilot experiences for natural language inquiries, allowing users to ask questions in plain language and receive immediate, impactful answers.

Maintaining a balance between automation and human oversight is crucial, especially in the realms of privacy and governance. A participant highlighted the importance of this balance, noting that while AI applications excel in pinpointing areas of quality, identifying root causes, and offering actionable insights, they cannot replace the need for human involvement. Gaurav emphasised that AI should serve as an enabler and accelerator for Data Quality initiatives, enhancing human judgment and providing essential business context. Ultimately, leveraging the strengths of both AI and human expertise leads to more effective and responsible Data Quality management.

Figure 14 Closing Remarks

Balanced Scorecards and Business Outcomes

Engaging department managers effectively hinges on addressing the urgency of their performance metrics. By encouraging them to review their balanced scorecards and assess the quality of data that underlies specific metrics, managers become more invested in the discussion. When prompted to consider how poor Data Quality may lead to distorted scorecard outcomes or failures, their interest often peaks, prompting them to seek assistance from business or process analysts.

This approach not only fosters deeper engagement but also highlights the tangible connection between Data Quality and overall performance metrics, compelling managers to recognise the importance of investing in data integrity for improved results. Ultimately, aligning balanced scorecards with Data Quality creates a strong rationale for continuous support and investment in Data Governance.

Gaurav emphasised the crucial connection between Data Quality, Governance, and business outcomes, highlighting that the ultimate goal is to enhance overall performance. He illustrated that achieving a significant return on investment (ROI) from data governance and quality necessitates ongoing effort; organisations do not generate a $25 million impact from a single action, but rather through a series of initiatives that streamline operations and gradually enhance results. Gaurav reinforced the notion that Data Quality is an ongoing journey, rather than a final destination, requiring continuous engagement and incremental improvements that align with the business’s priorities.

Book Information and Closing Remarks

Gaurav highlighted the importance of “Data Quality ROI,” which is now available on Amazon, as a valuable resource for the entire data community. He encouraged readers to share their feedback, emphasising that such input is essential for enhancing the collective wisdom captured in the book.

With his LinkedIn profile and email provided for further questions and connections, Gaurav reinforced the idea that this book belongs to everyone, drawing from the insights of clients, colleagues, and data professionals alike. In conclusion, he believes that engaging with the community will foster growth and improvement in Data Quality practices.