Does splitting an Apex batch class with multiple actions into multiple batch Executables slow performance a lot?
No. Splitting a batch job into modular, rules-driven Executables improves clarity and maintainability without materially affecting performance. Runtime is driven by the number of DML operations executed and the automation those DMLs invoke (Flows, Triggers, Processes), not by whether logic runs in one batch or multiple Executables.
Benchmark (same workload, three implementations)
Each implementation retrieves 30k Accounts, then updates those Accounts 3
times per batch (no field changes), with all automation disabled.
| Implementation | Batch Size | Bulk API | Total Time |
|---|---|---|---|
1) Single Apex batch class; 3 updates in execute method
|
200 | No | 62 min |
| 2) DSP Pipeline; 3 identical batch Executables | 200 | No | 26 min |
| 3) DSP Pipeline; 3 identical batch Executables | 2000 | Yes | 4 min |
