Databricks Lakebase Tackles Marketing Costs

Databricks is pushing its Lakehouse architecture into the transactional database space with Lakebase Postgres, aiming to streamline marketing operations. The core innovation lies in its serverless OLTP database capabilities, designed to handle the spiky demand of personalized marketing campaigns more efficiently and cost-effectively. This approach tackles common industry inefficiencies, such as paying for idle database resources between campaign pushes.

StartupHub data

Companies working on this

Profiles of the companies named in this story, with funding and a one-liner from our database.

Databricks

$190.0B

A unified data analytics and AI platform built on the lakehouse architecture.

SAP

Global leader in enterprise software, specializing in ERP, cloud solutions, and business intelligence.

Traditionally, customer segments are prepared in a data lake and then pushed to OLTP databases for marketing tools. This process often leads to data teams building and maintaining complex synchronization pipelines. Lakebase aims to eliminate this operational burden.

Serverless Scaling for Bursty Workloads

Lakebase Postgres leverages serverless autoscaling, allowing compute resources to scale down to zero when not in use. This is particularly beneficial for marketing campaigns that experience significant traffic spikes followed by long periods of low activity. Costs align directly with usage, ensuring that companies only pay for what they consume.

The architecture separates storage from compute. This allows for a growing breadth and depth of customer attributes without a linear increase in compute costs. This flexibility is crucial for richer personalization efforts.

Streamlined Data Integration

A key feature is the native integration with the Databricks Lakehouse via Synced Tables. This managed synchronization removes the need for manual pipeline development and maintenance. Marketing teams can reportedly make new customer segments available to platforms like SAP Engagement Cloud in just a few clicks, accelerating time-to-market.

The system is optimized for high-concurrency point lookups and short OLTP queries, distinguishing it from traditional OLAP systems.

Performance and Optimization

As Lakebase is built on Postgres, standard optimization techniques apply. Creating indexes is highlighted as a primary method for improving query performance, especially for filtering customer IDs in marketing campaigns. Databricks also provides tools within its UI to monitor query performance, including specific metrics like PREFETCH and FILECACHE relevant to Lakebase.

The company suggests tuning parameters like work_mem and autovacuum_vacuum_scale_factor for further optimization based on observed performance bottlenecks.

Databricks positions Lakebase Postgres as a solution to significantly reduce TCO and accelerate campaign delivery by combining serverless OLTP with its Lakehouse platform, as detailed in their blog post.