October 2022 Summaries

Why are vampires bad at art? They're only able to draw blood. The company Census has partnered with Google Cloud to make BigQuery data more actionable, and also announced partnerships with Prefect and other companies such as Figma and Google Ads. Census is introducing new features including business metric alerting, workspace organization, and model activity tracking, and will be hosting an event on building a Composable CDP with Google Cloud. The company is also hiring, and invites users to provide feedback or join the OA Club on Slack.

Ergest Xheblati recently hosted a hands-on SQL workshop where he shared his expertise on query decomposition patterns, query maintainability patterns, and query performance patterns. Through examples and real-life scenarios, Ergest demonstrated how to break down complex queries into smaller, independent pieces using Common Table Expressions (CTEs), making them more readable and maintainable. He also discussed the importance of following rules for query performance, such as avoiding sort operations until the final SELECT statement, reducing data before joining it, and keeping WHERE clauses simple. The workshop covered various questions from SQL practitioners worldwide, with Ergest providing valuable insights on topics like CTE usage, commenting in SQL, and the differences between CTEs and temp tables.

Workspaces are the next level of governance for your Census account, enabling granular permission management for different teams, projects, or use cases. This allows customers to organize syncs and models however they wish, giving them more control over sensitive data access while reducing compute costs and minimizing mistakes. With a range of roles available, including Owner, Editor, Operator, and Viewer, users can be assigned specific permissions within each Workspace, ensuring only those with appropriate access can view or modify sensitive information.

Account scoring is an analytics concept that helps determine the level of engagement of each account, sorting prospects from most to least valuable. It allows understanding each account's overall company health, prioritizing sales team targets, and tracking customer adoption. The key components include a stream of activities, a scoring mechanism, and a strong understanding of the ideal customer profile. Building an account scoring matrix requires listing important activities, creating a user table, an account table, and a scoring table. It is essential to start simple, gather data from various sources, and normalize aggregation by total users associated with the account. Account scoring can solve business problems in customer-related areas such as preventing churn, identifying power users, and building case studies. By combining account-level data with tools like Census, teams can create concise communication, analyze interactions, and automate with confidence, ultimately shifting their mindset to an account-oriented approach.

You want to unload data from Snowflake to use it in other places, such as building dashboards or merging with other data in your CRM. Before unloading data, consider whether you'll need the data in your tools more than once and evaluate whether automation is necessary. You should also backup your data to ensure its safety. The `COPY INTO` command enables you to copy an entire table or a query result to a Snowflake stage, from where you can download the data to your local system or unload it to external storage locations such as Amazon S3, Google Cloud Storage, or Microsoft Azure. You need to specify the destination, source, storage integration, partitioning, file format, and copy options when using the `COPY INTO` command. Additionally, consider which file type is the best option for your data and ensure that it can be handled by your destination platform. If you're overwhelmed with manual or automatic unloading setup, you may want to consider using Census reverse ETL instead, as it offers over 85 destination integrations and sync speeds of over 4,000 records per second.

The Census team has made significant improvements to the Model Activity tab, providing increased transparency and collaboration for users. Users can now view a detailed log of specific changes to each model, including who made those changes. This feature is available for all SQL models except those created through dbt or Looker integrations. The updated model history logs are not only available for future changes but also for past changes, with activity history found for models that are over 3 years old.

Customer data is constantly changing, but business teams often struggle to get real-time updates on customer activity. To solve this challenge, Census has introduced a new feature called Notifications, which allows users to set up business metric alerting using their single source of truth, without code. This feature uses a flexible visual builder to customize message content and automatically triggers messages to Slack or Microsoft Teams whenever data changes. With Notifications, any team can stay updated and take action on customer activity, enabling personalized outreach, identifying opportunities, and preventing churn. The new feature is available today for all Census users on all plans, offering a better way to trigger messages and leverage data compared to point-to-point integration models like Zapier.

If you’re searching for a data integration tool to connect your business apps, you’ll probably narrow your search down to two tools: iPaaS and Reverse ETL. Think about how your data might be used throughout the business and consider exactly how your data will be used, as inaccuracies in that data can have significant implications. The key difference between iPaaS and Reverse ETL is that iPaaS integrates data by creating point-to-point connections, while Reverse ETL syncs a single source of truth to GTM tools via a hub-and-spoke method, making your data more reliable and accessible. The biggest benefit of Reverse ETL is standardizing on a single source of truth - your analytics warehouse. Choosing Reverse ETL over iPaaS allows for growth without the messiness of adding extra connections for every app. SQL is a crucial business skill for RevOps professionals, as it helps work faster, make well-rounded business decisions, and mold a fulfilling career path. Developing key "star" metrics to unify your team can help iron out product roadmaps and create a more focused roadmap. RevOps acts as the organization's layer of defense, acting as a middleman between departments to keep the entire organization running smoothly. Don’t try to boil the ocean by trying to do everything at once; start small and build towards it. Open communication across the organization is essential for getting everyone on the same page and driving impact.

Operational leaders at growing companies are struggling with large-scale data integration due to using iPaaS tools that weren't designed for such tasks. These tools, like Zapier and Workato, have a low entry barrier but can quickly become unwieldy as the number of integrations increases, leading to messy data and a lack of visibility. As workflows grow more complex, users need detailed knowledge of how APIs work, defeating the purpose of a "no-code" tool. iPaaS tools also lack scalability for large-volume batch updates, such as product usage data, which can become very expensive quickly. Furthermore, they complicate data governance and quality by allowing easy integration of sensitive data with low-security data, leading to potential security risks. In contrast, Reverse ETL provides a single source of truth by delivering customer data from a data warehouse right to the tools used by frontline business teams, enabling high-quality data models and self-serve data syncing. Ultimately, companies should consider using both iPaaS and Reverse ETL depending on their specific needs, with Reverse ETL being more suitable for larger organizations that require visible pipelines inside and out of a single source of truth.

RevOps teams can leverage their data ecosystem to maximize revenue outcomes by implementing product-led growth strategies, which involve using customer product usage data to drive personalized marketing and sales efforts. This requires building a comprehensive 360° view of product usage, enriching customer accounts with context from internal data, and empowering product-led growth through authentic conversations with customers. However, RevOps teams face challenges in deciding what product data to use, as there is no standardized approach, and siloed data can lead to a lack of confidence in the data. To overcome these challenges, teams need to centralize their data in a single source of truth, model it together, and focus on metrics that matter for their business. Additionally, RevOps leaders must adopt an agile product approach to activating their data, prioritize projects based on sponsor input, and iterate to get exactly what they need, while also ensuring data quality and trustworthiness. By doing so, teams can drive better outcomes and make the most of their operational tools.

Census helps companies centralize customer information in data warehouses like BigQuery, making it actionable in business tools. Reverse ETL is the missing piece of the data stack, enabling teams to deliver trustworthy first-party data or fine-grained metrics for personalization and scoring in minutes. Census achieves Google Cloud Ready – BigQuery designation, proving its integration meets core functionality and interoperability requirements. The platform connects seamlessly with other tools in the Google ecosystem, empowering business teams to self-serve, scale the impact of the data team, and multiply the ROI from existing stack tools like dbt, Airflow, and Looker.

The text discusses the adoption of dbt, an open-source ETL tool, by N26, a tech company. The data analytics team at N26 found that dbt allowed them to set up complex production-grade data pipelines using SQL SELECT statements, freeing up time for data engineers to focus on administrative tasks. Before adopting dbt, the team used a light ETL microservice that was cumbersome to use and had several bottlenecks in tooling and data governance. The introduction of dbt revolutionized how the team works, allowing them to easily add business logic to raw data sources without taking precious time from the DWH team. The team now has over 800 data models running daily or weekly, with well-organized projects and a containerized application on a Kubernetes cluster. Despite some challenges, dbt has had a significant impact on the team's productivity and scope, and the company plans to continue using it as they expand their business.