Why High Leverage Analytics is a Cross-Functional Team Sport

Summary

In the domains of movies and real-life data analytics, the process is a group task rather than an individual endeavor. The conversation led by Steph Olavstadir, CEO and co-founder of Avo, explores in detail the difficulties and solutions related to managing data quality and analytics in the development of products. Steph, having a decade of experience across diverse fields, stresses the need for cross-functional collaboration and treating data as a product to create better user experiences. She references her experiences at QuizUp, where issues related to data quality led to the creation of internal tools, and emphasizes the need to incorporate analytics into the cycle of product release to prevent data errors and enhance decision-making. The discussion further elaborates on the transformation from the chaotic state of data management to a self-serve governance model, stressing the need to align stakeholders and create a culture where analytics is a group task. By applying data mesh principles and creating a shared language across teams, organizations can achieve high-impact analytics cultures that empower teams to use data effectively.

Key Takeaways:

• Effective analytics requires cross-functional collaboration and communication.
• Considering data as a product can improve data quality and decision-making.
• Incorporating analytics into the product release cycle prevents data errors.
• Global and local standards in data governance are important for efficiency.
• A data-driven culture requires proactive data management, not reactive solutions.

Deep Dives

Cross-Functional Collaboration

Steph Olavstadir indicates that analytics is a group task, involving product managers, developers, and data scientists working together. The secret to success is in aligning these stakeholders through purpose meetings, where goals, metrics, and data structures are defined collaboratively. This method not only improves data design but also allows team members to use data effectively. "Analytics is the most fundamental tool that developers have to build better user experiences," Steph mentions, stressing the shift in developers' perspectives when they actively participate in data-driven decision-making. By creating a shared language and understanding across teams, organizations can ensure that analytics is smoothly integrated into the product development process.

Data as a Product

Considering data as a product is a novel concept that Steph advocates for enhancing data quality. This perspective involves considering usability, feasibility, desirability, and business viability when creating data. By treating data like a product, teams are encouraged to include stakeholders in the design process, ensuring that the data meets the needs of its users. This method aligns with the principles of data mesh, which promotes decentralized data ownership and a shared self-serve data infrastructure. Through this perspective, data becomes a strategic asset that drives better decision-making and user experiences.

Incorporating Analytics in Product Release Cycle

Including analytics as an integral part of the product release cycle is important for preventing data errors and enhancing decision-making. Rather than being an afterthought, analytics should be incorporated from the start, with data kickoffs aligning with product sprint kickoffs. This proactive approach creates a culture where high-quality data is prioritized, and teams are motivated to implement effective analytics. Steph shares, "It's not often we have a webinar where the solution to everything appears to be more meetings," but highlights the value of a single, well-structured meeting to align on data goals and metrics as a preventive measure against future issues.

Striking a Balance between Global and Local Data Governance

Effective data governance requires striking a balance between global standards and the flexibility for local teams to innovate. Steph discusses the importance of establishing global naming conventions and casing standards to prevent duplicate entries and ensure clarity in data analysis. By implementing these standards, organizations can facilitate cross-team data usage while allowing domain-specific teams to create their own relevant data events. This balance supports a self-serve governance model, where teams can efficiently use data for decision-making without the bottleneck of centralized control, ultimately leading to quicker and more informed product development.