Empowering Growth with Customer Lifetime Value - A Data Product Journey
At Axel Springer’s German subsidiary, Axel Springer National Media & Tech, the Customer Intelligence team, comprising data scientists and data engineers, is on a mission to enhance the marketing and customer relationship management for our media outlets. Their goal is to understand and drive the monetary growth of these outlets’ subscriber bases. Central to this mission is a powerful metric known as Customer Lifetime Value (CLV).
Customers represent the most important assets of a company. CLV allows assessing their current and future value and aligning the customer relationship management strategy and marketing resource allocation based on this metric. But what makes one subscriber more valuable than another? The answer how to measure this lies in CLV. Imagine knowing precisely which subscribers will bring the most value over time. This insight is not just useful; it’s crucial for steering your company’s focus towards nurturing relationships that promise the highest returns. Let’s explore how the team is transforming this concept into a full-fledged data product that reshapes our approach to customer acquisition and customer retention.
What is Customer Lifetime Value (CLV)
Customer Lifetime Value (CLV) is the estimated value that a customer will bring to a business over a defined time-frame.
To understand CLV, think of it as identifying which subscribers will continue their subscriptions over time and contribute the most revenue. Each subscriber generates a steady flow of monthly subscription fees. The key is determining which subscribers will maintain and grow their engagement over time, consistently renewing their subscriptions each month. Media companies use CLV to forecast which subscribers are most worth the investment, ensuring ongoing revenue and minimizing the risk of cancellation.
There are various ways to calculate CLV. One common method is to multiply a customer’s average monthly recurring revenue by the sum of monthly probabilities that the customer will continue their subscription over time. The core of this calculation is a predictive model that forecasts the probability of a customer staying for future months up to a specific point in time.
Given its importance, the team decided to put CLV as a Data Product on the roadmap and make it accessible to the company in form of a CLV dashboard.
Why CLV Matters for Media Outlets
Understanding Customer Lifetime Value (CLV) is crucial for our media outlets. CLV provides insights into the long-term value of each subscriber, enabling media companies to make informed decisions about resource allocation, marketing strategies, and customer relationship management. By identifying high-value subscribers, media outlets can tailor their content and offers to enhance engagement and retention. This targeted approach not only maximizes revenue but also optimizes marketing spend, ensuring efforts are focused on subscribers who contribute the most to the business’s bottom line. Ultimately, leveraging CLV helps media outlets achieve sustainable growth and maintain a competitive edge in the industry.
From Data Projects to Data Products
When looking at data initiatives, the distinction between a “Data Product” and a “Data Project” is crucial yet often misunderstood. A “Data Project” typically refers to a one-time endeavor with a specific, limited scope and a defined endpoint. It is aimed at solving a particular problem or answering a specific question. Once the goal is achieved, the project is considered complete.
There are advantages worth mentioning resulting from considering the CLV topic as a Data Product opposed to a Data Project:
A “Data Product” is a living entity, designed to evolve and continuously provide value. It integrates seamlessly into the operational fabric of a business, offering ongoing services and updates. Data Products are constructed with scalability and sustainability in mind, allowing them to adapt to new requirements and insights over time. This approach not only extends the lifespan of the data’s utility but also transforms it into a core business asset that drives decision-making and strategic initiatives on an ongoing basis.
For instance, in the context of a Customer Lifetime Value (CLV) dashboard, treating it as a Data Product means continuously refining and updating the user interface and algorithms based on new customer data and business requirements. This ongoing process helps maintain the relevance and accuracy of the insights provided, unlike a data project where the analysis might become outdated once the project concludes.
Benefits of CLV as a Data Product
As stated previously, the CLV is based on a predictive model built using machine learning (ML). Especially AI and ML-related work has traits that make it benefit from treating it as a data product [Source: Data Product Management]:
- Adaptability to Change: In AI and ML development, organizations often can’t fully determine the potential outcomes of a product before its development is complete. Uncertainty is usually quite high. Typically, aspects like specific outputs, deadlines, and budgets can only be partially defined in advance. Therefore, adopting a product mindset is crucial, where planning involves conceptualizing the product using discovery methods and making informed assumptions rather than setting fixed parameters from the outset.
- Continuous Development: AI solutions require ongoing development and management similar to a product lifecycle, as they are never truly “complete” or “finally delivered”. Organizations should establish dedicated teams responsible for specific AI products, mirroring the approach taken with regular products. These teams handle continuous development, integration, and adoption of the data products within the organization.
- Integration into Larger Ecosystems: To maximize the value of AI and ML data products, they need to be integrated into a larger ecosystem, such as a digital product or platform. This integration of features into broader ecosystems is a common responsibility in data product management
In summary, considering CLV as a data product rather than a one-time data project leads to a more adaptable, continuously improving, and integrated solution. This approach not only maximizes the value of AI and ML efforts but also ensures that the insights derived remain relevant and actionable over time.
Developing the CLV Dashboards
In an era where data is king, delivering data projects in the media must evolve. Embracing this, we adopted a data product methodology for the development of Customer Lifetime Value (CLV) dashboards for our media outlets. This approach emphasizes anticipating business value from the start, ensuring that every technical feature aligns with user needs and business goals.
User Requirements Gathering
The initial phase involved deep dives into gathering user requirements through workshops and feedback sessions. This helped us identify the critical metrics that drive customer retention and acquisition — essential for any media outlet aiming to thrive in a competitive landscape.
Data Transformation Process
We then forwarded these requirements to the Customer Intelligence team, which operates in SCRUM mode. They began the data transformation process with the goal of streamlining and automating data flows, ensuring high data quality and speedy retrieval.
Rigorous Software & Data Pipeline Engineering
Developing the dashboards required rigorous data pipeline engineering, crucial for supporting data updates and ensuring scalability for new features. One significant challenge was switching the customer master data source during the main development phase. Another challenge was developing a single code base for several user interfaces, i.e. one dashboard for for each media outlet. Despite this challenge, we maintained focus on usability and interactivity.
Iterative Refinement
The dashboard initially featured a Cohorts Tab visualising the CLV for each cohort in a chart. Then followed the Executive Summary overview. Based on user feedback, we added the option to see the retention rates, and a marketing offer tab for deeper insights into CLVs. We continuously refined the design and the underlying model, addressed user feedback and eliminated bugs. The end result is a robust dashboard that not only met but exceeded user expectations.
Continuously Providing Value Through CLV Dashboards
During the presentation of the CLV dashboards, the impact was immediately evident. A senior executive exclaimed, “Wow, we finally left the stone age!” This reaction underscores the transformative effect of the CLV dashboards on our business processes. Both CLV dashboards have become an important tool in our customer-focussed business strategy, eagerly supporting our customer retention and customer acquisition efforts.
For example, customer acquisition precisely knows which amount of money to spend on an individual user to keep up profitability and can further optimize for the latter!
The dashboards provide monthly, fully automated updates, which keep stakeholders informed with the latest data insights. This democratization of data has empowered teams across the organization to make informed decisions quickly, fostering a culture of agility and responsiveness.
Future Directions and Continuous Improvement
As we continue to enhance the functionality of our CLV data product, our goal is to keep democratizing contained insights across the organization. This is especially true since our described data products are dashboards, which inherently rely on the user’s need to watch them and were developed with that in mind providing valuable lessons from our customer bases.
We are committed to refining our data product portfolio, ensuring they not only meet the current needs but also adapt to the evolving media landscape. By doing so, we aim to maintain our leadership in innovation and customer understanding, driving our media outlets towards greater successes.
Conclusion
In summary, the development of CLV dashboards for our media outlets represents a pivotal advancement in our data-driven strategy. These dashboards not only streamline decision-making but also foster a culture of agility and responsiveness and were developed with a data product mindset. As we continue to refine our data products, we are committed to staying at the forefront of innovation, ensuring sustained growth and success for our media outlets.
About the author: Justin Neumann is a Data Scientist and MS in Predictive Analytics striving to achieve great things by addressing technology to the needs of human beings, and implementing viable solutions altogether. He works at Axel Springer National Media & Tech.
