Your First Prospecting Search

Your First Prospecting Search

Prospecting used to mean opening five browser tabs — Candid, GrantStation, Instrumentl, your state’s grant portal, maybe a foundation’s 990-PF filings on ProPublica — and tab-hopping until something stuck. You’d read a funder description, squint at their recent grants, decide whether to keep going, then open another tab. The searching itself was most of the work.

Grantable flips that. You describe what you’re looking for, and the AI runs the search for you — across its own proprietary funder dataset, GrantGraph™, federal opportunity databases, and the open web, all in one conversation. The skill you bring to prospecting isn’t searching databases anymore. It’s knowing what to ask for.

What Happens Behind One Prompt

When you ask Grantable for funders, it’s not running one search — it’s running several in parallel and merging them.

All four tracks run, their results get deduplicated, and the AI scores each candidate against your organization’s profile. What you see in chat is the merged, ranked list.

Running a Search

You can start a prospecting search two ways, and both work equally well.

Plain chat. Just describe what you need:

“Find family foundations that fund youth mentoring programs in the Mountain West, grant sizes $25K–$100K, that have given in the last two years.”

Grantable recognizes this as a prospecting request and runs the search. No command needed.

The /prospecting skill. Type / in the chat input and pick Prospecting from the skill list, or type /prospecting directly. This invokes the skill — a set of instructions that walks Grantable through a thorough prospecting workflow: scoping the search, running complementary queries, writing structured briefs, and presenting results in an interactive table.

The skill’s job is to give you a repeatable, high-quality prospecting pass. Plain chat is faster for exploratory questions. Use whichever matches the moment.

What Comes Back

A /prospecting search produces three things that all tie together:

1. A search vector folder under /Prospecting/ in your workspace library — something like /Prospecting/Youth Mentoring - Mountain West/. This folder is the search itself, saved. You can come back to it, rerun it, or schedule it to refresh.
2. A Prospect Brief for each top match inside that folder — one file per funder, with fit classification, why it’s a match, key data points (giving history, typical grant size, geographic focus), notable grantees similar to yours, any red flags, and cited sources.
3. An interactive prospect table in chat — sortable by fit, clickable rows that open each funder’s brief, and Accept/Dismiss buttons so you can narrow a long list into a shortlist without leaving the conversation.

Because the briefs live as real files in your library, everything the AI does next — fit scoring, opportunity briefs, writing outreach letters — can pull from them. A prospecting search isn’t a one-shot report. It’s the start of a research thread your team can build on.

Skills and Templates Are Shortcuts You Own

The /prospecting skill and the Prospect Brief template both live in your workspace library, at /Skills/prospecting/. They ship pre-loaded — but they’re files, not hardcoded behavior. You can open them and edit them.

Steering the Search

Specificity is the single biggest lever. A vague request produces a vague result. Be concrete about:

• Program area — “K-12 STEM education for girls” beats “education.”
• Geography — “funders giving in Pennsylvania” beats “U.S. funders.”
• Grant size range — “$25K–$100K” filters out the Gates Foundation grants that’ll never come to you.
• Funder type preference — “family foundations only” or “avoid corporate funders” shapes results.
• Language the funder would use — “positive youth development” and “youth mentoring” overlap but surface different funders. Describing your work the way the sector describes itself expands coverage.

You can also:

• Add material mid-search. Upload an RFP you’re studying, or @mention a file from your /Library/ — a past proposal, an annual report, a board deck. Grantable will use it to shape what “good fit” means.
• Emphasize a vector. Say “focus on federal opportunities only” or “skip our current funders, we want new ones.” Grantable will bias the search.
• Iterate. A single pass is rarely enough. Run two or three variations — different focus wording, different geographic scopes — and compare what each surfaces.

The First Pass Is a Shortlist, Not a Final Answer

Prospecting works in rounds. A /prospecting run hands you a ranked shortlist — enough to decide which funders deserve a closer look. From there, you drill in by asking Grantable follow-up questions about the ones that caught your eye:

“Tell me more about Meyer Memorial Trust’s recent giving to youth programs.”

“Which of these three has the strongest history with organizations our size?”

“Show me the last five grants from Oregon Community Foundation and flag anything similar to ours.”

Grantable pulls deeper research on the funders you point it at — using the briefs it just wrote as context, plus any additional 990 data, web sources, and org-specific analysis. Your first query doesn’t have to be perfect. Start broad enough to see the landscape, then go narrow on what matters.

Next Lesson

Next we’ll look closer at GrantGraph itself — how it connects organizations to their funders, and how to read what it’s telling you when a funder shows up in your results.