Data governance: The key to retail excellence

For decades, the retail landscape has been defined by a tug-of-war between the tangible allure of brick-and-mortar stores and the analytical precision of digital platforms. The latter offers a more convenient and personalized shopping experience, while the former boasts a sensory-rich experience, with expansive inventory and unparalleled adaptability. However, contemporary shoppers seek a synergistic blend of both worlds. In this evolving marketplace, savvy retailers recognize that leveraging data is not just beneficial—it is imperative for success.

But how can this data be utilized without significant investment? How can big retailers continue to leverage their valuable data to optimize inventories, streamline logistics, target consumers, ensure adequate staffing, and enhance the customer experience? How can merchandisers use data to present the right products to the most valuable consumers at the optimal times? And how can retail IT teams ensure data remains accurate and trustworthy, data sources remain integrated and accessible, and the business retains easy access to timely, high-quality data? This is where data governance (DG) comes into play. Poor data quality and inconsistent data formats can lead to delays, affecting services like customer support, sales, and marketing outreach and eventually to erosion of consumer trust.

A McKinsey publication highlighted that leading firms have eliminated millions of dollars in cost from their data ecosystems and enabled digital and analytics use cases worth millions or even billions of dollars. Data governance is one of the top three differences between firms that capture this value and firms that don’t.

A large retail organization implemented data governance with supply chain digitization to optimize its supply chain operations. By standardizing data across its vast network of suppliers, distribution centers, and stores, the retail organization enhanced inventory management, reduced stockouts, and improved overall supply chain efficiency.

In the realm of large-scale retail, the complexity of data and the multiplicity of data sources present significant challenges. Data governance in retail differs from that in other industries such as finance due to the unique challenges and objectives of each sector:

1. Customer-centric vs. Regulation-centric: Retail data governance is often driven by the need to understand and serve customers better, focusing on personalization and customer experience. In finance, data governance is heavily influenced by risk management and reporting requirements, which leads to a higher level of compliance tracking.
   
   
2. Data types and usage: Retailers deal with a high volume of transactional data, customer interactions, and inventory details, which require governance strategies that support real-time decision-making and marketing initiatives. Financial institutions manage a variety of data types, including sensitive personal financial information, which demands stringent governance for security and confidentiality.
   
   
3. Technology and integration: Retailers prioritize governance frameworks that support the integration of diverse data sources, such as point-of-sale systems and e-commerce platforms, to create a unified view of the customer. Financial institutions often focus on governance frameworks that ensure data integrity and accuracy for financial reporting and risk assessment, along with more integrated data sources to manage their customer data.
   
   
4. Speed of innovation: Retail tends to be more dynamic, with rapid changes in consumer behavior driving the need for agile data governance that quickly adapts to innovative and adaptable marketing strategies and technologies. Finance, while also evolving, has a more deliberate pace of innovation due to the need for more stability and trust in financial systems.
   
   
5. Data volume and velocity: The volume and velocity of data in retail is significantly higher due to daily transactions and customer interactions across multiple channels. This requires governance processes that can handle large-scale data analytics. Finance also deals with large volumes of data but focuses more on the veracity and validation of data for accurate financial reporting.

Understanding these differences is crucial for implementing effective data governance strategies tailored to the specific needs and goals of each industry.

Data governance is a pivotal aspect for any organization, but it becomes even more significant when dealing with one of the world’s largest footwear and apparel retailers. Such a retailer faces unique challenges and opportunities that make data governance not just a necessity, but a strategic asset.

Here is an enhanced deep dive into the various facets of data governance that Cognizant applied to a leading global retailer in the footwear and apparel industry.
 

Building governed glossaries with metadata management

Metadata management is a critical practice for building a data-driven business. Understanding what data exists, where it resides and what it means for business in context of analytical products and data products in big retail organizations, helps harness full potential of the data. Caught between the complexity and numerosity of data sources, big retail companies today struggle to strike the right balance with data-driven decision making. Hence, with more distributed data, business glossaries become crucial for creating a shared language around data in retail organizations. Business stakeholders are vital and are the sole providers of semantical definitions since data is not IT-focused, but handling data is. Key components of a business glossary include:

1. Business terms and definitions
2. Reference data assets and data models
3. Data governance policies

As business glossaries are captured and standardized across the organization, subject matter experts, data owners, and business users play a key role in capturing tribal knowledge and context around data and terminology. It is surprising to see how a simple retail concept like a “stock-keeping unit” or a pricing term like MSRP—Manufacturer's Suggested Retail Price—is misaligned between teams, with different understanding and logic of what they mean.

Cognizant DG framework and organization focuses not only on DG strategy but also on technology and culture-related aspects like top management commitment and sponsorship-promoted governance culture that should be business owned and technology executed. By documenting a business glossary with over 400 key performance indicators (KPIs) and critical data elements, Cognizant ensured trustworthiness and clarity of data with business teams. More effective structures for data management led to improved data processing, more succinct data reports and efficient use of resources to increase value of data assets.
 

Maintaining data quality

As retailers continue to invest in digital transformation initiatives and omnichannel strategies, data quality (DQ) gains more value in the overarching data governance strategy. The integrity of data directly influences critical business decisions, ranging from inventory management to customer engagement strategies. In a complex data landscape spread throughout different geographies, precision and timeliness are crucial; ensuring that data is accurate, consistent, and reliable becomes non-negotiable.

In current retail businesses, both business and technology teams struggle to trust the data they work with. Multiple data sources, diverse tools, and a lack of standardization and alignment between global and local geographies results in duplicated efforts and inefficiencies. Companies like ASICS have optimized user experience with governed data, leading to a growth of over 5% in conversion and an impressive one million in additional revenue each year.

Cognizant created and executed a common DQ framework (see below) that was scalable to global teams. The application of a data quality framework includes data profiling, definition of data quality rules and thresholds, data quality dimensions, a well-structured issue remediation process and monitoring & reporting, focusing on:

• Whether all pertinent data is present and refreshed correctly
• If business rules are applied accurately
• If the product is of the highest quality and thoroughly checked for issues
• Who is responsible for resolving data issues

Figure 1: Cognizant’s data quality framework

Tying data governance and data quality together led to a product maturity score growth by 2.7 points (1.26 to 4 out of 5) by implementing extensive data validation checks, proactively monitoring and ensuring action on the flagged data issues. Efficiencies gained by reducing the effort and time it takes to identify data meaning, data source, data flow and impact during an analytical product development, amounted to a 15% reduction in time and resources.

Data quality practices and implementations for big retail have historically lagged than those of more mature industries such as the financial and pharmaceutical sectors, in which stringent regulations and heavy fines prompt a higher standard of data quality and proactive implementations. Proactive measures by such frontrunners ensure that building a data driven culture is more necessary than ever.

Using data lineage to build transparency

Data undergoes multiple aggregations and transformations within various complex systems and layers before it is consumed and visualized. In retail, that data could be transactions, inventory records, or customer or product information. Data lineage documents this detailed journey from a source to visualization layer.

In retail, this data journey becomes even more complex, owing to the expansive data landscape and decentralized data sources. Tracking data lineage becomes crucial for different stakeholders. For middle management, it reaps more strategic benefits like cost-saving opportunities, enhanced user experience, and increased trustworthiness. For analysts and engineers, day-to-day analysis and execution becomes simplified with more efficient debugging and impact analysis, automated documentation, and easier cross-teams communication.

Complexity of data landscape, limited integration capabilities and internal resistance were some of the key challenges faced in implementation of data lineage at most retail clients. Close collaboration with the change management team is crucial to manage changes in ways of working for stakeholders involved.

Cognizant undertook a use case-driven approach to guide and enable relevant stakeholders with visibility into the data lineage, by navigating the complex data landscape in retail and ensuring:

1. Enhanced data reliability
   Tracking data at every stage of its journey, tying together with the DQ approach, enabling the retailer to take proactive measures, and ensuring that critical business decisions are based on high-quality data ensures enhanced data reliability.
   
   
2. Improved data traceability
   Retail companies are subject to GDPR and CCPA regulations. Therefore, in case of an audit, enabling and capturing data lineage ensures compliance with the regulatory requirements.
   
   
3. Efficient data management and process optimization
   Leveraging data lineage to identify inefficiencies in data management and opportunities for cost-saving ensures efficient data management and process optimization.

Process perspective

Figure 2: Data lineage for analytical product

Foundational pillars and key considerations

Some of the key considerations while implementing robust data management for retail could differ for those in a financial institution. Some of the key areas to include:

• Roles and organization
  Organization includes a Business Owner, Master Data Governance Office (Data Governance Owner, Data Lead, Data Quality Lead), Data Operations Lead, and IT Owner and there is a clear division on decision rights related to each role.
  
  
• Data strategy
  There is a formalized and approved data governance blueprint and roadmap with priorities and program objectives aligned with corporate strategy. There is also a defined data governance playbook (Policies, Data Domain Taxonomy, Data Dictionary, Data Catalog) with:
  
  - Value of data: What does it deliver, what does it earn and what do I need it for?
  - Storage of data: When, how often, how quickly and where is it needed?
  
  
• Policies & standards
  Definition and implementation of key policies: data standards, data quality, data security, data retention, data cleansing.
  
  
• Architecture
  Clear definitions & integrations of data models for source and target systems is available and critical data elements are defined, tracked, and monitored which allows clarity of systems and “what/where” to track.
  
  
• Compliance
  Identification of regulatory, statutory, Sarbanes–Oxley Act (SOX), and internal compliance requirements. Business rules are defined and managed with clear roles, responsibilities, and segregation of duties.
  
  
• Issue management
  An issue remediation process is in place with defined actors and actions, tracking escalation procedures for data issues and reporting on issue status in a predefined cadence.
  
  
• Data asset valuation
  There is a clear method of valuation of data assets (costs, potential growth, returns, risks). Data-driven companies outperform their competitors by 6% in profitability and 5% in productivity.
  
  
• Communication
  Regular communications and adoption plan for stakeholders with notification matrixes along with a training plan and resources to support DG processes.

With the rise in non-tech organizations such as big retailers shifting focus to become data-driven, data governance and robust data management play a pivotal role in enabling retailers to stay agile and competitive in an ever-evolving market landscape. Business glossary management, data quality implementations and visibility over data lineage predominantly remain the focus. Robust data governance frameworks centered around data quality serve as the base for providing clarity around data landscapes, while data lineage capabilities can elevate enterprise data governance by fostering innovation.

To conclude, change management and adoption remain crucial to the successful governance of an organization’s data. Cognizant helps streamline requirement identification, making it effortless to delineate needs and execute them with precision, ensuring a seamless transition from conception to realization. Contact us to learn more.