Maximizing Business Value with Data Products

Organizations struggle to deliver usable data promptly, leaving businesses with the dilemma of how to make data-driven decisions. Companies have worked with centralized data teams to build datasets, reports, and dashboards for many years. The thought was that having centralized data skills and knowledge would be the most productive way to be structured and deliver data quickly. Over time, and with more data from more sources, tremendous pressure has been put on these teams to produce more. Scaling the data teams appeared to be the only way to keep up with the organization's needs, but at what cost?

Data leaders are becoming increasingly aware of the complexity and efforts required to deliver. According to a 2023 Gartner survey, “less than half of data and analytics (D&A) leaders (44%) reported that their team is effective in providing value to their organization.”¹ So, what are the critical items keeping teams from delivering data and insights sooner?

One crucial point is that it is difficult for centralized data teams to keep pace with requests due to the ramp-up time it takes to obtain business domain knowledge for each project. Most organizations can have tens to hundreds of domains, making it impossible for somebody to understand all of them. But what if you were to align data teams around the business areas they serve so the data teams would have a better understanding of the business needs? What if you treated data like software products, with product owners accountable for their creation?

Zhamak Dehghani had the same questions and developed a strategy to address them. In 2018, she unveiled the data mesh methodology that detailed the fundamental principles for creating data products.

What is a data product?

A straightforward explanation of a data product is the packaging of data assets (raw or transformed) to solve a particular business need that is discoverable, understandable, and trustworthy. Examples include datasets, dashboards, reports, and APIs.

On the surface, this is nothing new. According to Dehghani, the principles of domain ownership, data as a product, self-serve data infrastructure, and federated data governance have been the missing components needed to enhance the ability to deliver the right information quickly to the data users. A data product owner will then understand the needs of the data consumers (i.e., business users, analysts, data scientists, etc.) and develop data products to meet those needs.

Before embarking on your data products journey, assessing your data culture maturity is essential to ensure alignment with the business on how you will work towards corporate goals to drive the business forward. Two key components that must be in place are a data governance program and stewardship.

It is critical early on to promote to business leaders why a product mindset for data is needed. This cultural change can be more difficult for many organizations than process or technology changes. The data product journey is an evolution that will take time. You are more likely to achieve success if you start with the following principles: emphasizing ownership, creating better collaboration, and enabling continuous improvement.

These principles, along with others, help data product owners positively impact their data users by creating data products that:

• Start with business value, 

• Ensure discoverability and understandability, and 

• Increase the trustworthiness of the data.

Start with business value

When working from a product mindset, data producers better understand priorities aligning to deliver what is truly important instead of guessing. Too often in the past, solutions were created that missed the mark, leaving business users wondering how they would ever obtain the data they needed to make critical decisions. 

The production of data products emphasizes the importance of domain knowledge and understanding business needs firsthand. In addition, having well-curated data assets helps the data producers to have a better understanding. For Alation customers, data producers can quickly create, package, and surface their data products in the Alation Data Intelligence Platform for consumption, providing faster time to value.

By having a data intelligence platform with self-service capabilities like Compose and Alation Connected Sheets, users can achieve faster time to insight by analyzing data products directly from the point of discovery.

As business needs change, data products can evolve and inspire new ones. Business users can provide direct feedback through Conversations on the platform with the data product owner. In addition, Alation Analytics provides key usage metrics that are crucial when making decisions on investments or how to allocate resources.

Ensure discoverability and understandability

To drive the adoption and use of data products, data product owners must provide the means to make them easy to find and understand. Product registries allow business users to find pertinent data products by organizing the registries by domain, department, or other categories to streamline access to trusted products.

The intimate domain knowledge of the data producers delivers a deep understanding of the data products to enable use. Important facets of this knowledge include examples, underlying data assets, and limitations so that the business user can determine if the product is fit for purpose. When questions arise, business users can collaborate with the data product owners for a better customer service experience.

Increase the trustworthiness of the data

The most critical item for data products is that business users must trust them. Data product owners must deliver trust by providing transparency to the product's service-level objectives, including frequency of change, accuracy, and completeness.

Data quality in the past has been a challenge for centralized data teams. The teams could identify the issues, but often, the quality checks they created existed in the data warehouse with no means to fix the data. By shifting data quality left, the accountability would align with the data product development team and allow quality checks to be introduced sooner and issues to be resolved faster.

With the Alation Open Data Quality Framework, Alation partners can help with data quality and observability issues from within the data pipelines to the final product. Data health and trust check flags can be used to communicate the product's trustworthiness.

Another essential way to promote trust is for the data products to adhere to federated governance policies. Properly handling data starts with organizational policies while allowing producers to add domain-specific rules that provide the right governance level for data quality, privacy, and sensitivity.

"Data has become a crucial form of capital, essential for the survival and growth of organizations,” said Stewart Bond, Vice President of Data Intelligence and Integration Software at IDC. “Alation’s Data Products capabilities can help organizations unlock value from data by empowering domain experts to create high-quality data products by leveraging the breadth of the Alation Data Intelligence Platform. This approach ensures that teams utilizing data products can efficiently find relevant and high-quality data, reducing time spent on data discovery."

A new mindset

The data product movement mirrors what application engineering teams have accomplished over the years by embracing agile principles to deliver domain-specific data products faster and with better quality. 

Dedicated efforts to instill a product mindset and ensure data producers gain domain knowledge are needed to accelerate the creation of data products. The Alation Data Intelligence Platform is where data product owners can deliver packaged data products that are easy to find, understand, and trust. 

Alation's Data Intelligence Platform helps businesses realize value from their data products faster than ever before. Organizations can invest, build, and mature their data culture through collaboration and by understanding the usage of products to evolve them and create new ones, enabling a truly data-driven approach to achieving strategic business outcomes.

^{1  Source: Gartner - https://www.gartner.com/en/newsroom/press-releases/03-21-2023-gartner-survey-reveals-less-than-half-of-data-and-analytics-teams-effectively-provide-value-to-the-organization}