A Databricks data engineering team is designing a multi-task job to process customer orders. The pipeline includes two tasks:

OrderValidation: Checks each order for completeness and correctness, marking invalid orders as such.

OrderFulfillment: Processes and ships only valid orders.

The team configures the job so that OrderFulfillment only starts after OrderValidation has completed successfully. Why is this task dependency essential in the pipeline design?