July 2025 Summaries

An effective data governance framework is essential for managing and protecting data within an organization, establishing clear roles, accountability, and processes that ensure data quality, security, and compliance with regulations like GDPR and HIPAA. Various models, such as top-down, bottom-up, center-out, silo-in, and hybrid, can be adapted based on an organization’s structure, data maturity, and regulatory requirements. The framework should incorporate components like data quality standards, security strategies, and change management to facilitate consistent and trustworthy data usage. A collaborative approach involving data stewards, IT professionals, business analysts, and executives is crucial for shared ownership and successful governance implementation. Challenges such as overcomplicating processes or failing to communicate the value of governance can be mitigated by tying governance goals to measurable outcomes and integrating governance within existing workflows. Tools that support automation, collaboration, and flexible policy enforcement can enhance the effectiveness of data governance, enabling organizations to make better decisions, manage risks, and foster data-driven innovation.

The guide explores the creation of efficient data science workflows, emphasizing the importance of Agile methodologies and CI/CD practices to enhance productivity and collaboration. It breaks down the data science process into key phases such as problem definition, data gathering, exploration, modeling, evaluation, and deployment, highlighting the necessity of the right tools to streamline these steps. Tools like Hex are recommended for integrating Python, SQL, version control, and visualization into a cohesive platform, reducing the inefficiencies of switching between different applications. The guide also outlines popular workflow frameworks like CRISP-DM, SEMMA, and OSEMN, and addresses common challenges such as resource management, version control, and maintaining workflow simplicity and modularity. By prioritizing these elements, data teams can optimize their workflows, ultimately improving their ability to deliver actionable insights and drive better business decisions.

Predictive analytics, an evolving field leveraging data analytics, statistical modeling, and artificial intelligence, is increasingly adopted by companies to anticipate future trends and make informed decisions. It utilizes various models like classification, time series, regression, clustering, and neural networks to cater to different types of data and business needs. Real-world applications span industries such as retail, where Amazon's recommendation algorithm drives significant sales, healthcare, with predictive models optimizing resource allocation, and human resources, where analytics forecast employee performance. Success in predictive analytics depends on factors such as data quality, model interpretability, technical infrastructure, and continuous monitoring. For businesses aiming to implement predictive analytics, it's crucial to align models with company objectives, ensure data readiness, choose appropriate tools, select suitable models, and iteratively refine them. Hex Technologies offers a platform facilitating the development and sharing of interactive data products, underscoring the transformative potential of predictive analytics in enhancing business outcomes.

Data leaders from HubSpot, Justworks, and The Movement Cooperative discuss the evolving role of data governance and semantic layers in creating trustworthy, AI-ready data systems. While 84% of data leaders prioritize data quality and reliability, only 50% have invested in semantic layers, with many considering changes due to skepticism about their effectiveness. Tony Avino from HubSpot emphasizes the importance of centralized data governance for AI and LLMs, while Mona Khalil from Justworks focuses on minimizing cognitive load and ensuring consistency for easier accessibility. Bella Wang from The Movement Cooperative highlights the challenges of data sharing and compliance among diverse members. The discussion reveals that while semantic layers are seen as tools for enhancing AI, there is caution about their utility without a solid foundation in data governance. The leaders agree that AI's success depends on high-quality data governance, and investments in data quality benefit AI models by providing the necessary documentation and control.

The blog post explores the current and future landscape of self-serve analytics among data teams, highlighting findings from the State of Data Teams 2025 report, which reveals that while 70% of data leaders find self-serve a valuable goal, 53% are dissatisfied with their BI platforms. It features insights from leaders at Settle, Doximity, and Kaplan, who emphasize the need for dashboards to evolve beyond traditional roles to deliver real business value. The discussion reveals a spectrum of definitions for self-serve and addresses the challenges of democratization, data governance, and balancing centralization with decentralization. The leaders express a vision where AI plays a significant role in automating report generation and enhancing data accessibility while maintaining the human element for nuanced analysis. The conversation underscores the importance of a collaborative approach in data management, where self-serve tools are part of a larger cycle of inquiry and resolution, fostering trust and consensus around data usage.

Hex recently hosted its inaugural NYC hackathon in collaboration with Snowflake, where 60 data professionals across 20 teams tackled various challenges to create impactful data applications. The event featured diverse projects, from evaluating the feasibility of introducing Waymo in NYC to developing an app for finding safe and convenient restaurants, and analyzing the effects of congestion pricing on city traffic. Participants had just two hours to develop and present their applications to a panel of five judges. The top prize was awarded to a team that proposed a phased introduction of Waymo in NYC, while other notable projects included a guide for agoraphobes and a parking app. The event showcased the potential of Hex's platform in facilitating the creation and sharing of interactive data products.

Cohort analysis is a method used to track and analyze the behavior of groups of users who share common characteristics over specific time periods, providing deeper insights into customer behavior and lifecycle. Unlike general customer segmentation, which groups users by shared traits regardless of time, cohort analysis focuses on temporal patterns, revealing insights such as when customers are most active or likely to churn. Businesses benefit from this analysis by crafting personalized marketing strategies, reducing churn, and enhancing fundraising efforts through more nuanced data stories. There are various types of cohort analysis, including time-based, behavior-based, size-based, and event-based cohorts, each offering unique insights into user behavior. The process can be executed using tools like SQL for structuring data, Python libraries such as Pandas and Matplotlib for in-depth analysis and visualization, and R packages for streamlined cohort creation and visualization. Additionally, platforms like Hex enhance cohort analysis by integrating SQL, Python, and visualization capabilities in a collaborative environment, simplifying the process and making it accessible even to non-technical team members through AI-driven tools.

Hex's Account 360 is a comprehensive customer health dashboard designed to streamline customer data management by integrating information from various sources such as Salesforce, Stripe, and internal support systems. Initially developed as a prototype to address the inefficiencies faced by the Hex data team in answering customer inquiries, it has evolved into an indispensable tool that enables instant access to crucial customer information. The app organizes data into tabs displaying customer health scores, user activity, feature usage, revenue, and support stats, allowing the customer team to have informed conversations and prioritize accounts effectively. By unifying siloed data and providing actionable insights, Account 360 minimizes the time spent on data wrangling, enhances proactive customer engagement, and reduces operational bottlenecks. It empowers all team members, regardless of technical expertise, to swiftly answer customer questions and make strategic decisions, contributing to healthier customer relationships and allowing data teams to focus on broader business insights.

Hex Technologies has launched a new suite of Admin REST APIs designed to help Hex Admins automate, audit, and scale workspace management more efficiently. These APIs enable organizations to handle administrative tasks like user and group management, project collections, and data connection credentials programmatically, reducing the manual overhead that grows with team size. Available on Team and Enterprise plans, these APIs allow for seamless integration with existing IT frameworks and ensure consistent access policies across teams. Additionally, an advanced Observability API, exclusive to Enterprise customers, provides insights into project data usage, aiding in audits and impact analysis. The company is also considering introducing a Model Context Protocol server to enhance user and permission management capabilities.

Despite 77% of data leaders expressing excitement about AI's potential, only 3% prioritize it due to its integration being more behind the scenes than as headline initiatives, according to insights from data leaders at HubSpot, Wellthy, and Harvest Hosts. Stephen Moseley from HubSpot notes that AI accelerates development work through tools like Cursor and Claude Code, while Kelly Burdine from Wellthy emphasizes its use in generating YAML files and exploring codebases. Brent Driscoll from Harvest Hosts highlights AI's role in classification and extraction tasks. The discussion also touches on using AI to support business users, the importance of a semantic model, and the challenges of balancing AI priorities with other tasks. Leaders share advice on integrating AI, emphasizing the need for leadership approval and demonstrating value through small successes. They express excitement over projects like real-time exploration of unstructured data and automating analytics engineering, underscoring AI's transformative impact on development workflows and data exploration.

Data teams often face repetitive and evolving customer inquiries that consume significant time and resources, but leading teams overcome this by creating interactive data apps that anticipate question evolution and empower stakeholders to explore data independently. Traditional Customer 360 dashboards are inadequate due to their static nature, often requiring either granting broad SQL access or overloading data teams with constant requests. Instead, data apps designed with common customer success questions and guided exploration features allow stakeholders to investigate using built-in parameters and business context, transforming the investigative process into an intuitive experience. This shift reduces the need for repeated queries, enabling data teams to focus on developing scalable tools rather than handling individual requests, thus evolving from service desks to strategic platform builders. At Hex, the focus is on building and sharing these interactive data products, which enhance team efficiency and autonomy, allowing customer success teams to independently find answers and make informed decisions.

Hex Technologies revolutionized its lead conversion strategy by employing data analysis and artificial intelligence to refine its Marketing Qualified Leads (MQL) approach, transitioning from a gut-feel model to a data-driven one. Initially, Hex assigned scores to leads based on aspects like persona and engagement, but it was still reliant on intuition. By utilizing regression analysis, Hex developed a dynamic app that identifies significant traits and behaviors predicting conversions across different segments. This approach revealed insights like the universal importance of demo requests and varying significance of seniority based on company size. The app, residing in Hex, allows for continuous updates, offering real-time, actionable insights tailored to specific market segments, thereby enhancing campaign specificity and scoring strategies.

Choosing the right product metrics is crucial for making informed decisions about product performance and strategic direction. Unlike Key Performance Indicators (KPIs), which are aligned with overarching business objectives, product metrics are more granular and focus on specific features and user interactions. They are instrumental in identifying what aspects of a product are effective or need improvement, thereby driving better product decisions and supporting strategic presentations with data-driven insights. Selecting appropriate metrics involves clarifying problem statements, choosing suitable frameworks like the HEART or Pirate metrics, and avoiding common pitfalls such as vanity metrics or outdated measurements. Effective product metric strategies incorporate both quantitative data and qualitative feedback to paint a comprehensive picture of user experience and product value, ensuring that teams can prioritize effectively and maintain a competitive edge. Tools like Hex facilitate the exploration and sharing of these insights across teams, promoting a unified approach to data-driven decision-making.