#6 Treat Data as a Product
By Pacific Northwest Data Analytics Leadership Board members Nellie McBride, Rohith Maneyapanda, and Gina Nuss.
How do you ensure that you maximize the usability and value of your enterprise data? And how do you embed practices that encourage adoption? Year in and out we continue to learn more about various approaches to data management and the best architecture to choose for each business scenario – should we use a data warehouse or a data lake? What about a centralized approach vs. distributed such as a data mesh? Do we need to leverage a data marketplace? Regardless of how you answer these questions, one common thread that will help maximize usability, value, and adoption is treating data as a product. By applying the relevant best practices from product development philosophies and principles, we can bring improved focus, vision, and relevancy to data solutions.
Treating data as a product requires thinking about the value of data in its various forms and how your end users use data every day. Who are the consumers of your data? How are those consumers interacting with the data and for what purposes? What output format is most helpful? What are the expected business outcomes? Applying product development practices to how you design, develop, deliver, and manage your data will improve your solutions. Benefits of this approach include:
A sharpened focus with iterative delivery ensuring your value is directly aligned to what your users need every day.
A value-centric data product road map that complements business roadmaps, ensuring that data solutions are aligned to key business strategies and will evolve appropriately over time through evaluation and adjustments that continue to provide a strong value proposition.
The ability to drive the organization forward with a data-driven focus that considers people, process, and technology.
End user interaction and feedback woven throughout the data product lifecycle encouraging adoption and maintaining engagement at all phases.
What is a data product? Data products can make finding and consuming organizational data as easy as shopping on Amazon! A data product is a reusable data asset, engineered to deliver a trusted dataset, for a specific business and end user-aligned purpose. It most often integrates data from relevant source systems, processes the data, ensures that it’s compliant, and makes it instantly accessible to anyone with the right credentials. Data products can be raw data, derived data, data marts, algorithms, decision support and automated decision-making, dashboards and other business intelligence (BI) solutions, and more.
In order to successfully treat data as a product to improve focus, vision, and adoption, there are several best practices to consider:
Create Data Product Roadmaps that Complement Business Strategy Roadmaps
Develop Data Product Value Propositions
Apply a User Focus with User Personas and Usage / Usability Assessments
Create Data Product Roadmaps that Complement Business Strategy Roadmaps
Strong data product roadmaps make sure that business needs and architecture plans for technical solutions complement each other and align to end user needs, expected business outcomes, and value delivered. Doing the legwork to determine your strategic roadmap for data and technical solutions in alignment with business will help move the organization away from delivering piecemeal point-designed data solutions as requests come in and towards a focused, integrated approach where common and reusable elements are leveraged to improve development efficiency, end user discoverability, and one source of truth. Taking a holistic approach ensures additional elements such as data governance and data management needs over time are considered. Meeting your stakeholders where they are in their business journey with their critical strategic goals and aligning both the content and structure of data products to those goals will also organically improve adoption.
For example, Ro Maneyapanda a Director of Business Intelligence and Analytics, found success with this approach. His large technology company with a digital advertising platform sought to increase advertising sales revenue while reducing third party advertiser churn. His central advertising data analytics team was tasked with providing insights to sales, marketing, operations, strategy, and partnerships teams to help drive growth through insights. By writing business value-informed project briefs and then getting signoff from the internal customers both alignment on what was needed to produce value and commitment to use the analytics increased. Once a project reached the briefing stage the product owner pitched the value and feasibility so that it could be prioritized and added to the team’s roadmap which then naturally corresponded with the overall business strategy and roadmap. To learn more, visit the case study at Ro Maneyapanda Technology Case Study (pacnwdataanalytics.net).
Develop Data Product Value Propositions
How often have you seen days, weeks, or months spent developing a data analytics solution that gets used for a short period of time then put on a shelf to collect dust, quickly becoming obsolete either as the business shifts, or as end users realize it didn’t truly meet their needs? Oftentimes requirements for a solution are gathered upfront but quickly focus on how a solution will be implemented or what components it will include, rather than what the expected business outcome is and why that provides business value. Driving requirements and priority based on the value proposition of the request not only decreases the risk of a solution missing the mark, but it also takes the conversation of what we’re building and why to a different place – what are the key use cases for this solution and how does it contribute to key business KPIs? This approach can also help facilitate tradeoffs as the investment and resulting business value are already captured and can be compared. Understanding how the ROI will be calculated, measured, and improved or maintained over time supports the bottom line and contributes to healthy adoption and ongoing engagement.
Apply a User Focus with User Personas and Usage / Usability Assessments
Prioritizing a focus on end users of data products from requirements gathering through implementation, monitoring and maintaining solutions is another way to stay closely aligned with, and ensure adoption from, your stakeholders. Understanding the user persona as part of requirements gathering is key – this representation of the typical goals and characteristics of the people who will be interfacing with a data product is critical to ensuring a business-relevant solution. Taking time to fully define this persona will help the analytics team understand the challenges and opportunities the end users face and allow them to fine-tune solutions accordingly. Ro Maneyapanda’s team at the large technology company used persona analysis to inform their project briefs including articulation of the problems, hypotheses, data signals, solutions, and expected outcomes. The understanding of the user personas was key to achieving substantial increases in the use of analytics by the sales, marketing, operations, strategy, and partnerships functions. The insights helped contribute to increases in sales revenue as selected functions could better perform activities such as optimizing ad placement, targeting advertisers with pricing and offers to reduce churn, and running the advertising platform more efficiently.
Conducting usage and usability assessments over the life of products closes the loop and supports ongoing alignment with the established roadmap. What was the original product plan, how is it impacting my users today, how am I delivering on that value, and how am I improving over time? These recurring assessments will allow you to continually inform, adjust, and improve the product roadmap and assets over time. If products are no longer providing a strong value proposition, the team can proactively assess whether to adjust or retire the product accordingly.
Pitfalls to Watch Out For
In order to truly treat data as a product in your organization, it’s critical to bring on resources that have the appropriate experience, skills, and training to be product managers in a data-driven environment. Shifting data engineers into this role may work but can also be a challenge bringing them up out of the weeds enough to build an appropriate road map and to drive user experience and adoption. Bringing on a true product manager or providing training and communities of practice to help employees shift their mindset, can help offset these challenges.
Software product development typically follows an Agile approach to development and delivery, and data product development benefits from this iterative approach as well. More often business teams follow a Waterfall methodology. This difference can pose challenges when it comes to determining when and how the two teams engage with each other. Communicate early and often and set clear expectations for engagement – for example in an Agile framework the analytics team may need end user validation and acceptance testing as an iterative engagement rather than a task at the end of a larger project. Data teams will also benefit from including innovation and experimentation support in their roadmaps to minimize the risk of becoming overly rigid. Identify periodic sprints for experimentation to encourage innovation and foster collaboration.
Conclusion
Treating data as a product provides an inherent focus on end users and value propositions that goes hand in hand with increasing adoption and ongoing engagement with data analytics solutions. A culture shift is required to make this happen – treating data like a product requires us (both within data teams and across the enterprise) to ask different questions and include different people to focus on where the business is going, what data is needed when, and how we enable solutions technically. This shift is worth the effort! When business priorities are understood and built into a data product roadmap along with data governance and data management considerations, the appropriate end users are identified and participating in the process, and the team includes skilled product managers and training of other key members this approach can make a substantial impact on adoption, ongoing engagement, and meeting strategic business goals.