What is Data as a Product?
Understanding data as both commodity and concept.
The phrase “data as a product” can be thrown around to vaguely signal “hey, we think this data is important and you should too,” but simply collecting or using data is not sufficient to reach product status. Data can exist in an organization without generating any value, and tools built on top of data may never have been thought of with customer needs, design principles, or iteration in mind.
Instead, data as a product can take two forms. As a commodity, it is used to generate revenue from outside your organization. A a concept, data as product will generate value within your organization.
Here, I’ll describe these two forms data can take as a product and provide specific examples to help you spot them in the wild, determine if you need to adopt a better product mindset in your own work, and identify new opportunities to create revenue streams.
As a commodity, data is a raw material or resource that can be bought and sold, just like other tangible goods. Data can be collected from various sources, such as sensors, social media, or websites, and can be packaged, priced and marketed to its target market. As data is becoming more and more valuable, it can be considered as a tradable commodity, similar to oil, gold, or wheat. Let’s review some more detailed examples.
A data set can be considered a data product if it is collected, cleaned, processed, and packaged with the intent of being sold or licensed to other organizations or individuals. A data set can include a wide range of information, such as demographic data, financial data, customer data, sensor data and so on.
For example, a company that specializes in retail data may collect data on consumer purchasing habits and product trends. They can then package this data into a data set and sell it to other companies that want to improve their own marketing and sales efforts. Example: Experian credit scores.
A metric can be considered a data product if it is created by analyzing data and is then sold or licensed to other organizations or individuals. For example, a company that specializes in web analytics may collect data on website traffic, user behavior, and conversion rates. They can then create metrics based on this data, such as the “bounce rate” or “time on site,” and sell access to this information to other companies who want to improve their own website performance. Example: NPS scores.
An algorithm, likewise, can be sold or licensed to other organizations or individuals. For example, a company that specializes in machine learning may develop an algorithm that can predict customer behavior based on their past actions. They can then sell access to this algorithm to other businesses that want to improve their own customer targeting and marketing efforts. The algorithm in this case is a product that is created from data and can be used to automate some business process, such as fraud detection, personalized recommendation, image recognition and so on. It can be sold as a standalone product or as a service, for example, via an API. Example: Chat GPT.
Reporting and insights are also routinely packed up, marketed, and sold. For example, a company that specializes in data analytics may collect data from various sources such as website traffic, social media, financial transactions and so on. They can then use this data to generate detailed reports and insights, such as customer demographics, purchase patterns, and website performance.
The reporting and insights can be delivered in various formats such as dashboards, visualizations, alert notifications and so on. They can be used to support various business processes such as marketing, sales, finance, and operations. Example: Neilsen reports detailing media viewership.
A data platform can be considered a product if it is designed to collect, store, and manage large amounts of data and make it easily accessible to other organizations or individuals. For example, a company may develop a data platform that can handle large amounts of customer data, including things like purchase history, browsing behavior, and demographic information. They can then sell access to this platform to other businesses that want to use this data to improve their own marketing and sales efforts.
A data platform may be a product that can be licensed and installed on customer’s premises or can be a cloud-based service. It can also include features such as data visualization, data governance, data integration, data quality, data security and so on. It can be used to support various business process such as reporting, analytics, and machine learning.
In summary, a data platform as a product is a software-based solution that can be used to store, manage, and analyze large amounts of data. It provides businesses with the ability to easily access and use data to improve their operations, generate new revenue streams, and make better decisions. Example: AWS, or adtech platforms like DMPs.
Not all data sets or applications are sold externally, however — they are sometimes meant to drive value within your organization. This is where the concept of data as a product emerges. This concept refers to identifying data as a valuable resource that can improve business operations, and investing in its development lifecycle accordingly. It is a paradigm shift that helps organizations better manage, protect, and activate on their own data.
Just because a data artifact like a dashboard may exist in your organization does not mean it was treated as an actual product. How many dashboards have you tried to use that don’t update with new data on time, are missing metrics, or have almost no other views? These were built and distributed, but not with a product mindset.
Data product development matches any new product lifecycle — identifying the target market (even internally!), developing a marketing or socialization plan, determining appropriate pricing and packaging (prioritization), delivering and supporting the product, and continuously monitoring and improving based on customer feedback. It is the adoption of this concept to ideate, build, and iterate in data that demonstrates that you are thinking of data like a product. Next, let’s get more specific about what this mindset looks like in delivering high quality data sets, and in building tools like dashboards with customers in mind.
An organization that implements a data mesh is treating their data as a product. The data mesh is an architectural paradigm that aims to provide a more decentralized, scalable, and resilient way to handle data. This approach advocates for a product-centric data architecture, where data products are built and managed independently, by domain owners, rather than centralized teams. It enables organizations to improve the way they collect, process, store, and analyze data, making it more accessible and usable for different teams. Additionally, it emphasizes data governance, as security and privacy as part of the product development lifecycle. Overall, the data mesh concept is closely related to data as a product, as it helps organizations to better leverage and monetize their data by making it more accessible, usable and valuable.
Data visualizations and dashboards can be, and are best, developed like other software by following a process of design, development, testing, deployment, maintenance, and iteration. Data analysts may fail to build with a product mindset by not understanding customer needs or the business, not thinking about scalability, not addressing plans for maintenance, and not thinking about data governance. By gathering customer or stakeholder pain points and asks, developing initial views, iterating, socializing, and measuring engagement over time, analysts, engineers, and data scientists can adopt a data product mindset and deliver outputs that are valuable and actionable to their organization.
Conclusion
Data as a product is both a commodity and a concept. As a commodity, data can be bought and sold, just like other tangible goods. It can be packaged, priced and marketed to its target market. As a concept, data as a product refers to the idea of treating data as a valuable resource that can be used to improve business operations. The concept of data as a product also includes understanding the potential uses and benefits of data and identifying new opportunities to use it. It is a paradigm shift that helps organizations to better manage and protect their data.
