Building a Flexible and Scalable Customer Data Platform
Exploring the Modern Data Stack Approach: how to avoid CDP vendor lock-in, drive down data costs and integrate their CRM at best with their Data Warehouse.
- Author
- Ruben Burdin · Founder & CEO
- Published
- November 1, 2024
- Read time
- 3 min read
I talk with a lot of clients on how to avoid CDP vendor lock-in, drive down data costs and integrate their CRM at best with their Data Warehouse. I decided to write a blog post to share what many have found valuable.
In this short blog, we deep dive into architecting a composable customer data platform (CDP) that scales for any type of company size.
The strategy I recommend is centered around the "Modern Data Stack," a concept that has been gaining significant traction in the data world for the past few years.
Step-by-Step Data Journey
1. Data Ingestion
Begin by gathering all your data from various sources using Airbyte and funneling it into Snowflake, your central data warehouse. This initial step is crucial in consolidating your data landscape. All your tool’s data are consolidated into a central place, the central source of truth.
2. Data Warehousing and Metrics
Within Snowflake, you have the freedom to develop comprehensive metrics, data views, dbt models, and aggregations across different data sources. This is particularly useful for identity resolution, allowing you to create a complete 360-degree customer profile directly within your data warehouse.
3. CRM Integration with Stacksync
Stacksync's two-way sync enables you to seamlessly read from and write to your CRM directly from your data warehouse. This eliminates the need for additional tools and simplifies the complexities of integrating your CRM. Any data written into Snowflake is instantly updated in Salesforce (and vice-versa), providing real-time synchronization.
Because the CRM is now embedded in the Data Warehouse with real-time and two-way sync, you won’t need to manage custom script, authentication, refreshing security tokens, data pagination, error handling,... and so on. All that is fully managed by Stacksync so you just have to update Snowflake tables with SQL queries to actually write data to your CRM instantly.
4. Business Intelligence Tool Flexibility
Choose from a range of BI tools like Tableau, Looker Studio (free from google), Power BI... The beauty of this setup is that all these tools can directly plug into Snowflake, giving you the liberty to switch tools as per your requirements.
5. Marketing Tools Integration
If your tech stack includes other marketing tools, consider using custom scripts or platforms like Hightouch/Census/Rudderstack for reverse ETL from your data warehouse to your marketing tools. While these are not real-time and require some setup, they offer a cost-effective and simpler alternative to more expensive options like Segment or Salesforce CDP. Not to mention these two boys (Segment or Salesforce CDP), they often trap customers into massive bills, lock-in with yearly contracts and are often limited in the numbers of connectors they offer. This means that sending your custom data into these CDPs is at your own cost. We meet more people willing to move out from these tools, than people willing to use these.
Why This Approach?
The Modern Data Stack stands out because it empowers companies to avoid vendor lock-in, scale data operations affordably, and interchange tools at will. It lays a solid foundation for long term data strategy, ensuring that you're equipped to handle the dynamic demands of today's data-driven landscape.
Conclusion
Embracing the Modern Data Stack approach for your company’s CDP is a strategic move that promises flexibility, scalability, and cost-effectiveness. By following this approach, you can look forward to a robust and adaptable data platform that supports the ever-evolving needs of the business.
If you have any feedback or would like to learn more and chat about real-time and two-way CRM data sync, feel free to reach out at ruben@stacksync.com or book a call with me.
FAQ
Frequently asked questions
Related articles
