To deliver timely and relevant communications to customers, businesses must effectively integrate data across multiple platforms. While this challenge isn't new, it has evolved considerably over the years. Historically, companies turned to custom API development for their data integration needs. However, in recent years, many have begun adopting Customer Data Platforms (CDPs) like Segment or reverse ETL tools such as Hightouch to synchronize data into necessary applications more efficiently.
The Rise of Warehouse Connectors: A New Era in Data Activation
However, we are now observing a new trend. SaaS platforms like Braze, Mixpanel, Amplitude, and Iterable (among others) are introducing native data warehouse connector features. These connectors enable businesses to bypass the traditional intermediary processes and link directly to their central source of truth—the data warehouse - to push enriched data into SaaS tools. This approach of treating the data warehouse as a direct source for customer data activation has the potential to drive greater efficiency and flexibility in data-driven engagement strategies.
Preparing for Integration: Navigating the Challenges Ahead
As with all technological advances, businesses must be well-prepared to make use of newly available capabilities. Integrating data between a data warehouse and a SaaS platform spans multiple functions and teams across an organization. This leads to more stakeholders, heightened communication demands, a greater need for alignment, and, at times, additional bottlenecks
Key Considerations:
- Data Quality: The Cornerstone of Effective Integration
The effectiveness of native connectors is only as good as the quality of the data they handle. Issues like missing data, incomplete records, or outdated information can undermine the entire process. Implementing proper data cleansing and validation processes is essential. - Bridging Skill Gaps: The Need for Specialized Knowledge
Using native connectors often requires specialized knowledge of both the SaaS tools and the data warehouses involved. This expertise is typically spread among marketing, product, and engineering teams. Effective collaboration across these teams is necessary to bridge knowledge gaps and ensure a smooth implementation. - Schema Alignment: Overcoming Data Formatting Challenges
Different tools often require data to be formatted in specific ways. Misalignments in schema or data structure between the warehouse and the destination tool can create friction, necessitating additional transformations that can be time-consuming or error-prone. Teams may lack the necessary expertise to format data to align with each tool, highlighting the need for careful planning and possibly additional training. - Evaluating Connector Capabilities: One Size Doesn't Fit All
Not all native connectors are created equal. Some connectors might lack the full feature set or robustness required for more advanced use cases, so it is important to evaluate the capabilities of each connector to ensure it supports the business need before proceeding.
Case Study: Reducing Churn for a Leading Media and Entertainment Subscription Business
We recently worked with one of our top clients, a leading media and entertainment subscription business, to implement a solution using a data warehouse connector.
The Challenge: Combating Involuntary Churn
As a subscription business, mitigating churn was a top concern for our client. Specifically, involuntary churn—when users are lost due to payment issues such as expired credit cards—was presenting a significant business challenge. We needed to develop a solution that would automate communications any time a subscriber triggered a payment failure event, with the goal of recovering these subscribers within a three-day window before they were officially flagged as involuntarily churned.
Crafting the Solution: Our Strategic Approach
Assessing Data Requirements and Exploring Options
First, Mammoth Growth spent time determining the data requirements and exploring potential options for automating the dunning campaigns. The key requirement was to integrate transactional data from their third-party payment processor into Braze to trigger the automated campaigns and re-engage subscribers.
In their current setup, a third-party payment processor provided daily file dumps containing transactional data to an AWS S3 bucket. Snowflake was able to ingest these S3 files, but there was still a gap in making this data available to Braze, where the marketing automation would take place.
Selecting the Optimal Solution: Braze Cloud Data Integration
With this in mind, Mammoth Growth had to architect a solution to package the data residing in Snowflake as custom events in Braze. We explored several potential solutions, including integrating with Segment CDP, using Hightouch, and leveraging Braze’s Cloud Data Integration (CDI).
Implementing Reverse ETL: Automating Dunning Campaigns with Braze CDI
Ultimately, we opted for Braze Cloud Data Integration due to its simplicity and robust native integration capabilities.
What is Braze Cloud Data Ingestion (CDI)?
Braze CDI is a feature within Braze that allows businesses to integrate and sync data directly from their data warehouses—such as Snowflake, Amazon Redshift, Google BigQuery, Databricks, and others—into Braze. This enables companies to use their existing data infrastructure to send custom events, update user profiles, and trigger marketing campaigns in real-time without needing intermediary platforms.
How Mammoth Growth's Architecture Accelerated the Solution
Mammoth Growth’s Medallion Architecture is a structured approach to data management that organizes data into distinct tiers to optimize processing, governance, and accessibility:
- Bronze Tier: Serves as the initial collection point for raw data, ensuring data integrity and supporting the early stages of data processing.
- Silver Tier: Transforms raw data into structured insights through rigorous processing and the application of business rules, preparing it for strategic use.
- Gold Tier: Optimizes data for end-user consumption, ensuring that insights are reliable, timely, and actionable for decision-making.
Since this architectural framework was already in place for the client, we were able to bypass common data management challenges and deploy the solution rapidly. The centralized business logic and efficient data presentation allowed us to seamlessly send high-quality, consistent, and accurate data into Braze.
Key Insights from Our Implementation
- Mastering Data Formatting: Constructing dbt Models for Compatibility
Braze allows several types of data to be passed in, such as user attributes, subscription statuses, custom events, purchase events, catalog items, and user deletion requests. However, Braze requires this data to be sent in a specific format. We carefully constructed dbt models to match Braze’s API request format, primarily involving creating JSON objects to ensure data compatibility. - Synchronizing Data Refreshes: Ensuring Data Integrity
Data refreshes from dbt loads and Braze CDI syncs require careful synchronization to maintain data integrity across systems. Although Braze CDI supports syncs running as frequently as every 15 minutes, we implemented a daily sync to align with the client's needs and resources.
After successfully configuring the integration, dunning campaigns were triggered automatically based on the custom events in Braze. This allowed the client to re-engage with customers before they were flagged as churned, improving retention rates and reducing involuntary churn.
The Power of Collaboration: Leveraging Marketing and Data Expertise for Successful Deployment
Implementing this solution required seamless collaboration between marketing and data subject matter experts. Success demanded in-depth knowledge of Braze's capabilities, campaign strategies, and customer engagement best practices, alongside expertise in data warehouse integrations and building dbt models for the data pipeline. If the teams had worked independently, they would not have reached a solution as quickly or effectively. The marketing team lacked the technical skills to manage data formatting, synchronization, and integration complexities, while the analytics engineers did not have a complete understanding of the marketing objectives, campaign triggers, or the nuances of customer communication strategies. Mammoth Growth bridged this gap, aligning both technical implementation and marketing goals to ensure a successful deployment.
Mammoth Growth: Your Partner in Seamless Data Integration
At Mammoth Growth, we create solutions that leverage modern, innovative technology to build repeatable data patterns that are sophisticated yet not overly complex. Navigating the complexities of modern data integration requires not just the right technology but also the right team and strategy.
Our team of data engineers, growth marketers, technologists, and consultants works at the intersection of data, technology, people, and process. This unique blend enables us to help businesses harness emerging trends and drive innovation.
