How Snowflake's Connector for Google Analytics Centralizes Marketing Data

‍

Nothing is worse than when I can’t get at all my data. Which is why the Snowflake Connector for Google Analytics has been a lifesaver for me as an analytics engineer at Sigma. 

In the fast-paced realm of digital marketing, understanding your customers' journey is like navigating a complex maze of data points, each representing a user interaction or specific action on your website. And just when you have it figured out…the data points change and so does the entire maze. When news broke that Universal Analytics would cease data processing after July 1st, 2023, it sparked questions about how marketers would analyze the same data now processed differently. They also faced the challenge of data migration. For many years, Google Analytics has been the leading web analytics tool of choice, with an impressive 88% market share among web analytics tools. Google Analytics 4 (GA) is the latest version of Google's analytics platform, offering enhanced user-centric measurement and event-based tracking, while Universal Analytics (UA) represents the older version, focusing on session-based tracking and pageviews. 

The GA Migration

The key divergence between the two platforms lies in their data structures. While Universal Analytics relies on sessions and pageviews, GA4 takes a different route, built on events and parameters. As an Analytics Engineer at Sigma, I led the initiative to ensure our company seamlessly adapted to these changes, solving for accurate data and deciphering website performance with precision. This role demanded strategic problem-solving to navigate the intricacies of the transition, and the outcome has been a data-driven approach that aligns seamlessly with the evolving landscape of Google Analytics.

Another twist: GA4 Standard introduced higher data sampling and thresholding rates when producing reports, which can lead to incomplete and less accurate data analysis. These practices involve limiting the amount of data processed or substituting actual data with approximations, potentially hindering the precision of insights derived from the analytics platform. This change posed a challenge for our stakeholders as it impacted their ability to answer crucial questions. 

Ex: Performance of less trafficked, mid-funnel pages were completely buried below the surface reporting.  

The increased data sampling and thresholding rates in Google Analytics 4 meant that the analysis wasn't dependent on 100% of the raw data available. This limitation became evident when stakeholders found themselves unable to extract detailed insights into specific user behaviors, assess campaign performance with granularity, and make decisions based on the most accurate representation of their data. Recognizing the significance of these constraints, the data team at Sigma was tasked with finding a better solution to ensure our stakeholders could continue to extract meaningful and reliable insights. 

Getting the Raw Data

Next, we had to figure out how to migrate the data - a problem few companies have yet to solve. 

If you Google how to get Google Analytics data into Snowflake, most blogs either point to a no-code connector that pulls data from the Google Analytics API (which is not great because of data sampling) or a long (tl;dr) and complicated technical blog of how to get Snowflake to read parquet files from GCS. Hence, innovation became paramount.

While GA4 allows raw exports into BigQuery, Sigma uses Snowflake as our primary internal cloud data platform. Unifying this level of raw data with Salesforce data was critical for us, as it provides the granularity and depth needed to power sophisticated marketing campaigns. The Snowflake Connector for Google Analytics Raw Data was the perfect fit, enabling seamless transfer and transformation of GA4 data. With no-code dbt models pulling in both GA4 and Salesforce data, we now have a comprehensive single source of truth for website performance. This unified data landscape not only surpasses the limitations of the GA4 UI but also unlocks significant potential for AI/ML. Having raw data readily available within Snowflake empowers us to natively apply cutting-edge AI/ML capabilities without data movement or external dependencies. This allows us to build smarter, more personalized campaigns that deliver exceptional results.

Sigma’s Google Analytics 4 Template

While the internal data team was trying to solve this problem, we realized that other companies using Snowflake as their data platform must be going through the same issue. So we came up with an effective solution for stakeholders to get the insights they need faster. The result is a template that connects to their Google Analytics data in Snowflake and empowers users to access their GA4 data in two minutes or less. While Sigma’s data team spent time navigating how to solve this issue, we wanted to make sure that other data teams can use this solution to get insights for their stakeholders within two minutes.

This template isn't just about saving time; it's all about being smart with how you access and use your data. We wanted to give users a quick and easy way to dive into their GA4 data, getting rid of any unnecessary hassles or delays. Basically, this template makes exploring data a breeze, whether you're a seasoned analyst or a marketing pro. It's like a user-friendly shortcut to unleash the full potential of your GA4 data in the Snowflake world.

Moreover, with Sigma, you're not limited to surface-level insights. You can delve into every page's performance, going beyond the tip of the iceberg of your website's performance. Sigma empowers you to explore your data at the lowest level of granularity, providing a comprehensive view of user behavior and site navigation. It's not just about getting answers to common GA4 questions; it's about gaining a deep understanding of your website's dynamics and optimizing strategies based on detailed performance metrics. The template acts as your analytical ally, delivering straightforward answers to the classic GA4 questions that stakeholders commonly ask and allowing you to see the complete picture of your website's performance.

Sneak Peek into Sigma on Sigma: GA4 (inspiration to dive deeper into Sigma)

Internally at Sigma, the only thing left after transforming all that data was to turn it into cool insights. The workbook created by the data team is used every day cross-functionally to understand how our website is doing. By using some of the most beautiful and functional visualizations as well as workbook actions , our stakeholders can answer questions like, “How many Marketing Qualified Leads were generated through the website in the past 90 days?” or even “What is the most popular customer journey that users from the financial services industry typically follow?” as well as more standard questions around Engaged Sessions and Conversions.

Drilling to a Conclusion 

In navigating the GA4 frontier, what unfolded wasn't just a technical solution, but an adaptation to digital evolution. GA4 and Snowflake emerged as a practical duo, allowing us to decode the digital maze efficiently. The once intricate path of data analysis now seems clearer, with fewer twists and turns to navigate. It's as if we've illuminated the way, making the journey through the intricacies of digital analytics more straightforward. With GA4 and Snowflake, the end of the maze is almost in sight, offering marketers and analysts a more streamlined and insightful approach to understanding and leveraging data in the dynamic world of digital marketing. The journey continues, uncovering new possibilities armed with tools that redefine the way we understand and leverage data in the dynamic world of digital marketing.

‍