BigQuery in Practice: How MacPaw Achieved 10x Faster Analytics and Built a Unified DataHub

The Challenge: The Need for Fast Analytics

MacPaw processes large volumes of data every day, including user behavior, marketing performance, and analysis of new product releases. This work is a daily responsibility for business analysts and often requires significant time and resources.

Initially, the company relied on Google Analytics and small in-house solutions. It soon became clear that this approach was insufficient. Google Analytics limited access to raw data due to sampling, and combining its metrics with other sources—ad platforms, payment systems, internal products, and backend services—was not possible.

To gain a complete and manageable data set, the team built an analytics platform on Redshift. Over time, Redshift no longer met requirements for speed and flexibility, which led to the search for a new technical foundation. As a result, the team chose Google BigQuery—a decision that became the backbone of a new DataHub.

BigQuery as the Foundation of DataHub

Once the team migrated to BigQuery, it became clear that this was more than just another data warehouse. It was a different way of thinking about analytics.

Today, the team uses BigQuery as the core platform for data storage and processing and has built its own internal product—DataHub. It aggregates data from MacPaw websites and additional systems and is actively used for analytics and reporting. This is where everyone—from analysts to marketers—finds answers.

Why BigQuery?

The MacPaw team highlighted several decisive advantages that made BigQuery the right choice:

• Fast data processing that enables complex calculations with results in seconds rather than hours.
• A high level of security, essential for storing business-critical data.
• Usage-based pricing tied to the volume of processed data, suitable for both large and small businesses.
• An easy-to-use interface that lets analysts query data without writing code or relying on engineers.
• Built-in ML capabilities that enable the use of Google Cloud machine learning and AI services.
• Data visualization in Looker Studio (formerly Google Data Studio).

Since then, the company has significantly expanded its Google Cloud architecture, increased data volumes, and launched its first AI pipelines. Below is how MacPaw’s data ecosystem has evolved.

How It Works Today: Current Architecture and Scale

Since 2020, MacPaw’s use of Google Cloud for data workloads has grown more than threefold. This growth resulted from a shift to in-house analytics solutions and ecosystem expansion.

Today, MacPaw operates a flexible Google Cloud architecture that covers every key stage of the data lifecycle and is built around BigQuery as the central core. Processing and automation run on Google Kubernetes Engine (GKE) using Airflow and Spark.

Pub/Sub is used for event ingestion, while all data—from raw to processed—is stored in Google Cloud Storage (GCS) in Delta Lake format. BigQuery serves as the primary data warehouse, combining managed tables with BigLake to analyze data directly from GCS.

Results are visualized in Looker Studio, and the Data Science team trains and deploys models on Vertex AI.

A Unified Ecosystem for the Business

This architecture allows the team to launch new products quickly and test ideas without worrying about resource constraints. All data lives in a shared repository with clearly defined access controls via IAM. Dozens of datasets are no longer a challenge—they are the norm.

As a result, MacPaw gained a single platform that provides:

• a stable infrastructure for product operations;
• fast and secure access to analytics;
• the ability to scale AI initiatives without investing in on-prem infrastructure.

The Evolution of BigQuery: From Initial Datasets to Dozens

MacPaw considers increased team productivity and efficiency when working with large data volumes to be the main business benefit. BigQuery reduces wait times for calculations and results, which directly speeds up analytics and overall company operations.

The platform scales easily and is flexible to manage, simplifying data integration for data engineers and allowing the company to host growing volumes of data without purchasing additional servers.

• At the start, DataHub worked with 1–2 datasets.
• Today—dozens of datasets with access controls via IAM and Authorized Views.
• Used BigQuery streaming ingestion for near-real-time analytics.
• Added BigLake for SQL queries directly on data stored in GCS.

Vertex AI: When Analytics Turn Into Predictions

The next step was moving beyond analysis to building predictions. To achieve this, MacPaw integrated Vertex AI.

Vertex AI serves as the foundation for machine learning at MacPaw. It is integrated with the entire Google Cloud ecosystem, helping teams run experiments quickly and reuse training pipelines across projects , including work with large language models and deep learning—without additional infrastructure.

The outcome: faster model development cycles, scalable AI projects, and transparent cost control.

Business Impact

MacPaw built a unified data ecosystem where every data-driven decision starts in BigQuery and ends in Looker Studio dashboards. The company optimized costs while enabling secure data access at scale and reducing time-to-insight as the company expands.

Key Results:

What’s Next: AI as the Next Priority

MacPaw continues to expand its AI initiatives, focusing on Vertex AI and Google Cloud services that allow teams to launch predictive models quickly, experiment freely, and control infrastructure costs.

Cloudfresh as a Trusted Partner for MacPaw

To ensure everything runs reliably, MacPaw needed more than just the right technology—they needed a team to advise, support, and step in when needed. That partner is Cloudfresh. As a Google Cloud Premier Partner, Cloudfresh provided end-to-end consulting on BigQuery and helped integrate it into MacPaw’s ecosystem. Google Cloud experts at Cloudfresh continue to support MacPaw on technical matters.

What did Cloudfresh’s role include?

1. Data architecture consulting: optimizing BigQuery dataset structures, IAM setup, security, and scalability.
2. Rapid support in critical situations: assisting with service recovery and minimizing downtime risks.
3. Direct communication with Google: acting as a reliable intermediary to remove blockers through direct vendor engagement.
4. AI and ML expertise: advising on the use of Vertex AI to accelerate model deployment and scaling.

Our partnership with Cloudfresh continues — we are still here to strengthen the MacPaw team with expertise, support, and new Google Cloud capabilities.