top of page

Stop treating data as a Service and treat it as a Product

The conventional handling of data typically leans towards a service-based approach. This perspective, while functional, often relegates data to a supportive yet slightly underutilised role within organisations. I advocate for treating data not as a service but as a product. This shift represents a significant rethinking in the way data is perceived, managed, and utilised, particularly for data analytics and data science teams across various sectors.

The essence of this shift lies in recognising data as a core asset, pivotal not just in operational efficiency but in fuelling business growth and innovation. It moves away from the notion of data as a subsidiary outcome of business processes, presenting it instead as a primary driver of strategic development. The thesis of this blog post centres on the idea that a product-focused approach to data can open doors to new opportunities for business expansion, innovative customer engagement, and direct contributions to revenue growth. In this role, data teams transition from being internal service units to key players in product management, responsible for crafting, enhancing, and delivering data products that have a tangible impact on the business’s top line.

In the ensuing sections, we will delve into the concept of data as a product, the advantages of this approach, and the necessary frameworks, metrics, and strategies for effective implementation. This discussion aims to shed light on how a shift from a service-centric to a product-centric view of data can revolutionise data management practices, aligning them more closely with the strategic ambitions of the organisation.

The journey from viewing data as a service to treating it as a product is more than a semantic adjustment; it’s a strategic reorientation in the role of data within an organisation. Let’s explore how this reorientation can be a catalyst for growth and innovation in the way organisations harness the power of their data.

The Shift in Perspective: Data as a Product

The concept of data as a product entails a fundamental reevaluation of its role within an organisation. Unlike the traditional service model where data is a reactive, request-driven entity, treating data as a product places it at the forefront of business strategy and innovation. It becomes an asset developed proactively, with a clear purpose and target audience in mind, akin to any other product offering.

In a product-centric model, data is not just an output of various business processes; it is a crafted asset designed to meet specific business needs. This approach necessitates a deep understanding of the end-users — whether they are internal teams like marketing and sales or external customers. By focusing on user requirements, data products are developed to provide actionable insights that directly contribute to business growth, such as identifying new market opportunities or enhancing customer engagement strategies.

One of the key advantages of this approach is the ability to capture and capitalise on opportunities for business expansion. Data products, when well-designed, can reveal untapped market segments, customer needs, and potential areas for innovation. Additionally, the product mindset ensures that data initiatives are always aligned with strategic business goals, rather than being mere responses to ad-hoc requests. This alignment means that the impact of data on the business’s top line is not only more significant but also more measurable. The capacity to quantify Return on Investment (ROI) becomes more straightforward and effective, as each data product has clear objectives, usage metrics, and impact assessments.

Adopting a product-centric model for data ushers in a transformative approach for organisations, where data is not merely a byproduct but a catalyst for business growth. This paradigm realigns data initiatives with core business objectives, transforming data from a background utility into a strategic asset. It fosters an environment where data-driven insights are not just for optimising existing processes, but for pioneering new avenues in market engagement and revenue generation. The productisation of data, therefore, is not just an operational shift; it is a strategic move towards harnessing data as a key driver of innovation and competitive advantage.

Applying Product Management Frameworks to Data

Incorporating product management frameworks into data management signifies a strategic evolution in how data is conceptualised and delivered within an organisation. These frameworks, drawn from the well-established domain of traditional product management, are adapted to enhance the value, relevance, and impact of data.

Implementing go-to-market strategies for data products requires a nuanced understanding of the business environment. This process involves a series of steps designed to ensure the data product effectively addresses the needs of the business and is adopted widely. Initially, it’s essential to conduct a thorough market analysis to identify gaps, opportunities, and key stakeholders within the organisation. Following this, a positioning strategy is developed, defining how the data product will be perceived within the organisational context. This involves tailoring the messaging to resonate with the specific needs and challenges of different departments or teams.

The final stage is the strategic launch of the data product, aligning its release with organisational priorities and ensuring that support mechanisms are in place for user adoption and feedback. This methodical approach aims to maximise the impact and uptake of the data product, ensuring it delivers tangible value to the organisation. But the job is not done after delivery; it’s critically important to track the success of the product to iterate on it where required (based on user feedback) and to learn lessons for future products.

Evaluating the success and impact of data products requires a set of well-defined metrics. These metrics provide insights into how effectively the data product is meeting its objectives and how it’s being received by its users.

Adoption Rate: This metric measures the percentage of the target audience that starts using the data product, calculated as the number of new users in a period divided by the total target audience.

Churn Rate: Churn rate indicates the rate at which users stop using the data product, calculated by dividing the number of users lost in a period by the initial number of users at the start of that period.

Customer Satisfaction (CSAT): This assesses user satisfaction with the data product, typically calculated through user surveys and feedback, often on a scale of 1–5.

Time to Value (TTV): Time to Value measures the time taken for users to derive value from the data product, calculated from the initial usage to the point where significant value is reported.

Net Promoter Score (NPS): NPS gauges user loyalty and the likelihood of recommending the data product, calculated based on responses to the likelihood of recommending the product on a scale of 0–10.

These metrics serve as a critical tool for internal data products, offering an analytical framework to measure their effectiveness and alignment with business objectives. They allow data teams to track and optimise the performance of data products, ensuring they continue to meet and adapt to the evolving needs of the business.

This transition to data as a product necessitates a deep understanding of the business environment and a structured methodology in delivering data products. Enhancing the relevance and impact of data in this way ensures it acts as a dynamic tool for driving business growth, positioning data teams as strategic partners in the organisation’s development and fostering an environment where data is a key driver of innovation and market competitiveness.

Comparing the two approaches

Exploring the divergent outcomes of treating data as a product versus as a service provides valuable insights into their respective impacts on an organisation’s data strategy.

Data as a Product: Selective Focus with Expansive Benefits

In the data-as-a-product approach, the initial focus is on delivering quick, quantifiable value in specific areas of the business. This targeted approach allows for the development of highly effective, custom data products that address particular business needs one at a time. Each successful data product not only delivers immediate value but also acts as a catalyst for infrastructure improvement, tailored to support its deployment and functionality.

As these data products demonstrate their value, they grow in popularity within the organisation. Functions that were not initially the focus of data product development start to recognise the benefits and, consequently, express interest in having similar data-driven solutions. This demand drives a cultural shift towards valuing data-driven decision-making. In this way, the value, infrastructure, and culture around data grow in tandem, creating a reinforcing cycle of innovation and strategic alignment.

Data as a Service: Broad Efforts with Diffuse Impact

Conversely, the data-as-a-service approach aims for broad, albeit less targeted, improvements across the organisation. While this model strives for a comprehensive infrastructure build-out, the improvements tend to be slow and often unquantifiable in terms of efficiency gains. This approach carries the risk of perpetuating legacy architecture, which may not align optimally with emerging business needs.

A significant challenge in this model is the burden it places on individual teams. As the data team works to push forward the self-service and data mesh paradigms, teams across the organisation often find themselves too preoccupied with their core responsibilities to fully engage with and leverage these new tools. This leads to a situation where the data team is continually striving to promote data utilisation against the headwinds of existing workloads and priorities.

The comparison between data as a product and data as a service reveals deeper implications for long-term organisational growth and adaptability. Treating data as a product, with its focused and strategic deployment, not only delivers immediate, measurable benefits but also sets the stage for a more adaptable and innovative data infrastructure.

This approach, over time, nurtures a proactive data culture, where departments increasingly seek data-driven solutions, recognising their value in driving business objectives. In contrast, the service-oriented approach, while comprehensive, may lack the dynamism needed to foster a robust, data-centric mindset across the organisation. The key insight here is that a product-centric data strategy, by successfully embedding data into the strategic fabric of the organisation, paves the way for a more agile, forward-looking approach to data utilisation and decision-making. This strategic foresight positions the organisation to better navigate and capitalize on future market dynamics and technological advancements.

Challenges in Transitioning to a Data Product Model

Transitioning to a data-as-a-product model introduces significant challenges across organisational, technological, and cultural domains. Successfully navigating these challenges is key to adopting this transformative approach.

Organisational Challenges

A major organisational challenge lies in upskilling data professionals to align with the product-centric model. Data analysts, traditionally focused on data processing and insights generation, must now embrace aspects of product management. This shift involves understanding product lifecycle, user engagement, and strategic product positioning. Similarly, data engineers need to adopt design-centric development practices, focusing on building data products that are not only technically robust but also user-centric, working back from customer-focused goals to ensure relevance and effectiveness.

Technological Challenges

On the technological front, the transition requires more than just general IT upgrades. For data science products, robust MLops platforms are essential to efficiently manage the lifecycle of machine learning models, from development to deployment and monitoring. These platforms must support version control, testing, and scalable deployment, ensuring that machine learning models are effectively integrated into data products. Additionally, there’s a need for advanced analytics tools and data governance systems that can handle the complexities and scale of product-oriented data management.

Cultural Challenges

Culturally, one of the hurdles is expanding the awareness and understanding of data products among executives and decision-makers. Many leaders may not fully grasp the scope of what is possible with data products, which can lead to underinvestment or misalignment of expectations. Educating and enlightening these stakeholders about the potential and strategic value of data products is crucial for securing support and resources for the transition.

Strategies for Overcoming Challenges

Leadership and Alignment

Clear communication of vision and benefits from the leadership is crucial for the transition. This involves aligning the strategic importance of the data-as-a-product model across the organisation.

Cross-Functional Collaboration

Collaboration between data teams, IT, and business units ensures that data products are developed with a comprehensive understanding of both business needs and technological capabilities.

Investing in Technology and Skills

Prioritising investments in specific technologies like MLops platforms and upskilling teams in areas like product management and design-centric development are key.

Iterative Approach

Employing an iterative approach in the development of data products allows for continuous adaptation and refinement based on user feedback and business evolution.

Fostering a Data-Driven Culture

Cultivating a culture that values data-driven decision-making involves ongoing efforts in communication, engagement, and showcasing success stories of data products.

Six Steps to Transition to Data as a Product

1 Assessment and Planning

Begin with a thorough assessment of the current data landscape and identify areas where a product-centric approach can add value. Develop a strategic plan that outlines the objectives, scope, and roadmap for the transition.

2 Leadership Buy-in and Vision Communication

Secure commitment and support from the top management. Clearly communicate the vision and benefits of the data-as-a-product model across all levels of the organisation.

3 Restructuring Teams and Roles

Restructure data teams to align with a product management approach. This involves defining new roles and responsibilities that focus on product development, user engagement, and lifecycle management.

4 Upskilling and Training

Invest in training and upskilling initiatives to equip data analysts with product management skills and data engineers with design-centric development expertise.

5 Technology Upgrade and Integration

Implement necessary technological upgrades, such as adopting robust MLops platforms and advanced analytics tools. Ensure these technologies are integrated seamlessly with existing systems.

6 Iterative Development and Feedback Loops

Adopt an iterative approach in developing data products. Establish feedback mechanisms to continuously refine these products based on user input and evolving business needs.

Final Thoughts

In conclusion, treating data as a product is not just a trend but a strategic imperative for organisations looking to leverage data for meaningful business transformation. This approach ensures that data initiatives are tightly aligned with business objectives, delivering tangible and measurable value. As organisations embark on this journey, the focus should be on developing a cohesive strategy that encompasses technological innovation, cultural change, and organisational alignment. By doing so, data can truly become a cornerstone of strategic decision-making, driving innovation and long-term success in today’s data-driven business landscape.


bottom of page