If you’re running Nonprofit Success Pack, you’ve likely felt the shift coming. Salesforce has stopped issuing new NPSP licenses through the Power of Us Program, rebranded Nonprofit Cloud as Agentforce Nonprofit, and made clear where its investment is going. The question for most of our clients isn’t whether to move — it’s how to do it without breaking the systems their fundraising and program teams depend on every day.
We’ve guided nonprofits through more than 1,250 Salesforce projects, including dozens of NPSP-to-Nonprofit-Cloud and legacy-CRM migrations. This guide lays out the framework we actually use with clients: why the move matters, the step-by-step approach, where migrations typically go wrong, and how to tell if now is the right time for your organization.
Why Nonprofits Are Making the Move
NPSP was built on top of standard Salesforce Sales Cloud objects — Contacts, Accounts, and Opportunities repurposed to track donors and gifts. It worked well for over a decade, but it was always a workaround. Nonprofit Cloud (now Agentforce Nonprofit) is built from the ground up on a data model designed specifically for mission-driven work: donor and constituent relationships, program delivery, case management, and grants, all as first-class objects rather than retrofitted ones.
Figure 1 — NPSP vs. Agentforce Nonprofit (formerly Nonprofit Cloud) at a glance
In practice, that means:
- A standardized, scalable data model that doesn’t buckle under complex household or organizational giving relationships
- Native support for programs, benefits, and case management — no more bolting program tracking onto custom objects
- Automation built on modern Flow, not legacy triggers and workflow rules
- Deeper reporting, plus the on-ramp to Agentforce AI features like the donor support and participant management agents
The Belmar Migration Framework
Every migration is different, but the sequence rarely changes. Here’s the seven-stage approach we run with clients, from initial assessment through post-launch support.
1. Discovery & Assessment
Before touching anything, we map what you actually have: standard objects in use, custom Apex and Flow, third-party integrations (payment processors, email/marketing tools, event platforms), and reporting dependencies. Just as important, we sit down with fundraising, programs, and finance staff to understand what’s actually working today — migrations fail when they optimize for the data model and ignore the people using it.
Outcome: a scoped migration roadmap and a realistic timeline.
2. Data Model Mapping
This is the step organizations most often underestimate. NPSP and Nonprofit Cloud don’t map one-to-one, so every custom field, relationship, and record type needs a deliberate translation plan before a single record moves.
For each mapping, we define the field-level translation, any transformation logic (e.g., how soft credits or matching gifts carry over), and deduplication rules so you’re not migrating years of duplicate household records into a clean new system.
3. Data Cleansing & Preparation
We clean before we move, not after. That means removing duplicates, filling gaps in critical fields, standardizing formats (names, addresses, emails), and archiving records that no longer need to live in production. Clean data going in is the single biggest predictor of a smooth migration.
4. Build & Configure Nonprofit Cloud
We configure the target org — Programs, Benefits, Cases, security model — and rebuild automation natively in Flow rather than lifting and shifting old workflow rules. This is also the point where we push back on “just recreate what we had”: the goal is a system aligned to how your team should work going forward, not a copy of legacy workarounds.
5. Phased Data Migration
We move data in a deliberate order: master records first (constituents, accounts), then transactional data (gifts, commitments), then relationships and dependencies — validating after each phase rather than waiting until the end to discover a problem.
6. Testing & Validation
We run data validation (record counts, relationship integrity), functional testing on every automated process, and structured user acceptance testing with your team. We also run reports side-by-side in NPSP and Nonprofit Cloud to confirm the numbers your board and funders see don’t shift unexpectedly.
7. Go-Live & Post-Migration Support
Launch is the beginning, not the finish line. We train your team, monitor performance and data accuracy in the first weeks, and stay close to resolve issues fast — then fold feedback into ongoing optimization.
Where Migrations Go Wrong (and How to Avoid It)
NPSP vs. Agentforce Nonprofit: Key Differences
Data Security & Compliance
Donor and program data is sensitive by nature, and migration is exactly the moment security gets overlooked in favor of speed. Before go-live, revisit access controls and field-level security in the new org, confirm encryption and backup practices, and check migration handling against any regulations that apply to your organization — GDPR for international donors, state-level privacy laws, and funder-specific data agreements among them.
Change Management & User Adoption
We’ve seen technically flawless migrations fail at adoption because staff weren’t brought along. Training shouldn’t be a single session the week before launch — it works better as role-specific walkthroughs (what changes for gift processing vs. program staff vs. finance), paired with quick-reference guides and a clear escalation path for the first few weeks of real use.
When Should You Migrate?
Migration isn’t urgent for every org on day one, but a few signals suggest it’s time to start planning:
- NPSP is no longer available to new applicants under the Power of Us Program, and Salesforce has stopped shipping new NPSP features
- Your current data model is straining under complex household, corporate, or grant relationships
- You need program or case management and are duct-taping it onto custom objects
- You want access to Agentforce AI tools for donor support and participant management
- Your reporting needs have outgrown what NPSP’s data model can support cleanly
Our take: if you’re actively evaluating this, that’s usually the signal to start a discovery engagement — not necessarily to migrate immediately. A short assessment tells you your real scope and timeline before you commit to a date.