Your Donor Database Is Full of Answers — Here’s How to Find Them
Most fundraising teams aren’t losing donors because they lack data. They’re losing ground because the data they already have isn’t usable at the moment it matters. Contact records sit in one place, giving history in another, event attendance somewhere else entirely. A staff member who knows what they’re doing can piece it together, but doing that for thousands of records every week isn’t a strategy — it’s just labor.
The good news is that the data problem is usually solvable without replacing your CRM or starting over. For most organizations, the answer is already inside the system they’re using. It just needs to be organized, connected, and surfaced in a way that actually helps someone make a decision.
This guide covers what a functional donor prioritization system looks like, how to build one inside HubSpot, and what it takes to get there.
Why Most Donor Databases Don’t Work the Way Teams Think They Do
When a CRM is first set up, it makes sense. Contacts go in, properties get filled out, a few workflows run. But over time, as the organization grows and staff turns over and new objects get added, the data spreads out. What started as a unified system becomes a collection of related records that require effort to connect.
For a fundraising team, this creates a specific problem. The information that matters most — how many times someone has given, whether they attended a campus event, what their relationship is to a current student, whether they’ve asked not to be contacted — is all technically in the system. It’s just not visible from a single place. A development officer looking at a contact record has to click through multiple associated records, cross-reference notes, and rely on memory to form a complete picture.
At small scale, that’s manageable. At 100,000 records, it’s not.
The result is that teams end up defaulting to whatever is easiest to see. They reach out to recent donors. They work the same lists. They miss the person who gave consistently fifteen years ago, stopped, and now has significant capacity to give again — because nothing in the system flagged it.
That’s not a CRM failure. It’s a data organization and surfacing problem, and it’s one that AI is well-suited to help solve. You can read more about what AI actually does for fundraising teams before diving into implementation if you want more context on where AI fits.
What a Donor Prioritization System Actually Does
The goal of a donor prioritization system isn’t to replace fundraiser judgment. It’s to apply that judgment at a scale no human can match manually.
Think about your best development officer. When they look at a contact, they’re synthesizing a lot: giving history, recency, frequency, event engagement, family connections, capacity signals, whether the relationship has been cultivated recently or gone cold. They do this instinctively because they’ve been doing it for years. The problem is that expertise lives in their head, not in the system.
A donor prioritization system takes that logic — the criteria your senior staff actually use to decide who to call — and applies it automatically across every record in the database. The output doesn’t have to be complicated. In most cases, a simple high, medium, low score is enough. High means reach out now. Medium means cultivate. Low means hold.
That output, attached to every contact record and updated regularly, changes how a team operates. Instead of spending time figuring out who to call, they spend time making calls. Instead of management guessing at where to direct staff energy, they have a clear, data-driven view. Instead of donors slipping through because no one got around to reviewing their record, the system surfaces them automatically.
If you’re wondering whether your HubSpot data is in good enough shape to support this kind of system, that’s a reasonable place to start before building anything.
How HubSpot’s Data Agent Makes This Possible
HubSpot’s data agent is part of Data Hub, a set of data processing and visualization capabilities that allows you to aggregate information from across your HubSpot instance and do something useful with it. The data agent specifically allows you to create prompts, hydrate those prompts with contact data and related record data, and produce an output — all inside HubSpot.
For donor scoring, this means you can pull together everything that matters about a contact: their giving history, event attendance, school affiliation, family relationships, recurring donation patterns, communication preferences — and feed all of that into a structured prompt that returns a score.
The key capability here is that the data agent isn’t limited to what’s on the contact record itself. It can pull aggregate data from related records and include that full context in the prompt. That’s what makes it powerful. A contact record alone might tell you someone is an alum from 2003. The data agent can tell you they’ve given three times, attended two campus events in the last five years, have a child currently enrolled, and haven’t been contacted in eighteen months. Those are very different inputs, and they produce a very different output.
The score gets written back to the contact as a property. From there, it becomes a filter, a list segment, a workflow trigger. The output of the AI process integrates directly into how the team already works.
A recent episode of AI in Action covers exactly how this was built for a university fundraising team managing over 100,000 records, including the specific data points that went into the scoring prompt and how the output was structured.
Building the Scoring Logic
The scoring prompt is where your organization’s expertise gets systematized. This is also where most teams need to slow down and think carefully, because a vague prompt produces a vague score.
Good scoring logic starts with a clear question: what does a high-priority donor actually look like? Not in general terms, but specifically. What giving history qualifies someone as high? What recency threshold matters? How does event attendance factor in? Is a lapsed major donor who hasn’t given in five years a high or a medium?
Your senior fundraisers probably have answers to all of these questions. The work is getting those answers out of their heads and into a structured set of instructions that the data agent can apply consistently.
A few things worth building into the scoring logic:
Recency and frequency of giving. A donor who gave last year is different from one who gave three years ago and different again from one who gave consistently for a decade and then stopped. The system should distinguish between these profiles, not treat all past donors the same.
Event and program engagement. Attendance at campus events, participation in giving campaigns, and engagement with institutional communications are all signals. They indicate connection to the mission, which correlates with likelihood to give.
Family and household relationships. A student isn’t likely to be your best donor today, but their parents might be, and the student may become a significant donor in ten years. Tracking those relationships and scoring them appropriately requires intentional data structure. How to Build a Donor Scoring System Inside HubSpot goes into more detail on how to structure this.
Suppression signals. Anyone who has asked not to be contacted, or who has specific communication restrictions, should be filtered out before the scoring runs, not after. Build this into the logic early.
Capacity indicators. If your organization tracks wealth screening data or engagement with major gift programs, those signals belong in the prompt. They’re often the difference between a medium and a high.
What Changes When the System Is Running
The operational shift that comes from a working donor prioritization system is significant, and it shows up in a few specific ways.
The first is time. Fundraisers stop spending the first part of their day figuring out who to call and start spending it making calls. The research that used to take an hour per contact happens automatically, across thousands of records, before anyone sits down at their desk.
The second is coverage. Manual review processes miss people. They favor whoever is easiest to find — recent donors, contacts that have been touched recently, names that come up in conversation. A scoring system doesn’t have a recency bias. It looks at every record and surfaces the ones that qualify, including the ones that would have been overlooked. Why fundraising teams lose ground when donor prioritization is manual explains this pattern in more detail.
The third is management clarity. When scores are attached to contact records and updated regularly, leadership has a real picture of the pipeline. How many high-priority prospects are in the system? How many have been contacted this quarter? Where is staff time actually going? These questions become answerable without building a separate reporting process.
The fourth is scalability. As the organization grows and the database expands, the system grows with it. Adding 10,000 records doesn’t require adding a staff member to sort through them.
Getting Started Without Overcomplicating It
The most common mistake organizations make when approaching something like this is trying to build the perfect system before building anything. They want every data point included, every edge case handled, every score to be defensible before launch.
That approach tends to stall. A simpler starting point usually works better: identify the five to seven data points your best fundraiser uses most often, build a prompt around those, and run it. The first version won’t be perfect. It will still be useful. And once the team sees it working, the appetite for refinement grows naturally.
Can HubSpot automate donor prospecting? is a good read before you commit to a specific build approach — it covers what the platform can handle on its own and where configuration is required.
The other thing worth noting is that this kind of system is not a one-time project. Donor relationships change. Giving patterns shift. New signals become relevant. The scoring logic should be reviewed periodically and updated as the team learns what the output is actually predicting well and where it’s missing.
How FSM Approaches This Work
When Fast Slow Motion works with fundraising organizations on donor prioritization, the starting point is always the same: understanding what your best people already know. That expertise exists. The goal is to get it into a system that can apply it at scale.
From there, the work is about data structure — making sure the information that matters is stored in a way the scoring system can access — and prompt design — translating institutional knowledge into instructions that produce a useful output.
The result is a system that runs inside HubSpot, updates automatically, and gives every member of the development team a clear view of where to focus. How a university fundraising team got a unified donor view in HubSpot is a good example of what that looks like in practice.
If your organization is sitting on a large donor database and the team is still spending significant time on manual research, there’s a real opportunity to change that. The data is almost certainly there. The question is whether it’s organized and surfaced in a way that makes it usable.
Listen to the full podcast episode here.