Scoring 800+ communities to find the ones that actually qualify

Most grant research fails in the same way: a long list of "possible" opportunities that nobody can act on. This is how I do it instead — criteria first, data-verified, ranked by real eligibility.

The problem

When I started this practice my own working database listed federal programs and nothing else. It missed the real question: which specific communities are eligible, and which ones score high enough to actually win? Choosing targets by gut feeling wastes weeks on applications that never had a chance. So I rebuilt it, and this page is that rebuild — my method on my own data, not a client engagement.

What I did

• Locked the criteria first. Before searching, I defined eligibility against the official USDA rules (7 CFR Part 3570): population, median household income, and project type.
• Pulled primary data, not second-hand lists. I queried the US Census API for population and median-income figures for every place in the state — the source of truth, not a scraped estimate.
• Scored every community against the eligibility thresholds and priority-point rules, so each town got a clear, comparable score instead of a vague "maybe."
• Grew the dataset from ~450 to 800+ records by sourcing the missing places directly from the primary data, verifying every field.
• Delivered a clean, sortable spreadsheet — filter by score, income band, or population, ready to act on.

Sample of the deliverable

Illustrative rows (structure of the actual spreadsheet). Every value in the real file is verified against Census / USDA sources.

Why it works

Because the shortlist is built on official rules + primary data, you stop guessing and start applying where the odds are real. The same method transfers to any funding research: define the criteria, pull from the source, score, deliver something sortable and honest.