A nonprofit admin notices the nightly NPSP batch jobs are suddenly taking significantly longer to...

Nightly batch jobs in NPSP, particularly the Rollup Donors batch, process thousands or millions of records. Any increase in the "workload" per record will result in a measurable increase in total processing time.

Two Primary Performance Factors:

New Flows (A): Salesforce Flows (especially Record-Triggered Flows) are highly powerful but resource-intensive. If a new Flow was activated on the Opportunity or Account object, it will fire every time the NPSP batch job updates a record. If the batch job updates 100,000 Accounts, the Flow runs 100,000 times. If the Flow contains "In-Loop" queries or complex logic, it can drastically slow down the batch execution time or even cause it to hit governor limits.

New Customizable Rollups (C): Every time you add a new Customizable Rollup in NPSP Settings, you are adding another calculation that the batch job must perform for every record. If the new rollup has complex filters (e.g., "Total Gifts from 3 years ago excluding In-Kind and Grants"), the system must query more data and perform more evaluations per record, extending the batch window.

Why other options are less likely:

Standard Roll-up Summary (Option B): These are calculated by the Salesforce platform kernel and are generally more efficient than Apex-based rollups. While they add some overhead, they usually do not cause the "significant" slowdowns seen with custom logic or complex NPSP settings.

Role Hierarchy (Option D): While changes to the hierarchy trigger sharing recalculations, this typically affects record visibility and sharing performance rather than the execution speed of a data-processing batch job.