A brownfield environment begins with a system the organization already depends on. Moving an existing data warehouse or data lake to a lakehouse means accommodating historical data, legacy scripts, established reports, and daily marketing work. Unlike a greenfield project, the team must plan not only what to build, but how to retire the old system safely. section 5.4.1 section 5.4.2
A practical migration proceeds in phases:
- Inventory existing feeds, reports, and scripts; estimate the cost and duration of running two environments. section 5.4.1 section 5.4.2
- Move historical records in passes while piloting new uses of fresh data. Run old and new systems side by side, comparing measures such as total orders, cart value, and campaign ROI. section 5.4.1 section 5.4.3
- Log migration steps and reconcile outputs before cutover. Set explicit goals, timelines, and exit criteria so temporary duplicate storage and re-engineering costs do not continue indefinitely. section 5.4.3 section 5.4.4
Retail warehouse modernization
An apparel retailer has historical sales in an on-premises warehouse and email activity in a separate SaaS platform. It inventories its feeds and budgets for parallel operation, then migrates some sales history while piloting campaigns with fresh Shopify orders. Comparing old and new results protects existing reporting as the team works toward real-time segmentation. section 5.4.1 section 5.4.2
Table 5.3 Summarizing key challenges
| Aspect |
Brownfield |
Greenfield |
| Parallel runs |
Often crucial to compare new versus old metrics; doubles costs temporarily |
Usually not required unless you’re connecting multiple SaaS tools at once |
| Legacy scripts |
Must be reverse-engineered or retired |
Minimal or no legacy code, but risk of hidden logic in SaaS tools |
| Scalability |
Must migrate data gradually to avoid disruption |
Can scale from day one, but must define naming conventions early |
| AI readiness |
Potentially delayed if old system can’t handle large ML demands |
Easier to integrate AI if the architecture is planned for it |
Compare brownfield and greenfield challenges before setting the migration timeline: parallel runs, legacy scripts, gradual scaling, and AI readiness carry different costs and risks. section 5.4.6
A successful cutover is more than matching numbers. Sensitive marketing data needs security, compliance, auditing, and quality controls; teams also need training so they can use the new platform. Once the lakehouse is stable, a composable CDP can activate its data for campaigns. Track campaign setup time, conversion gains, retired-system costs, and data-team productivity to test whether the transition delivers value. section 5.4.5 section 5.4.6
