How to rank unserved census blocks for rural build priority

Every operator and regulator doing rural build planning hits the same wall. The FCC broadband map, a state broadband office's eligible locations file, or a subsidy program's eligibility list tells you which census blocks count as unserved or underserved. None of it tells you which ones are worth building first.

Ranking is a different job than eligibility. A block can qualify for subsidy and still be a bad first build if the households in it are three scattered farmhouses two miles apart, or a good first build if what the census data calls one sparse block is actually a tight cluster of forty homes along a single road. The ranking question is about what's really on the ground, not what the boundary says.

Why census-block ranking goes sideways

Census blocks were drawn for counting people, not for planning fiber or fixed wireless drops. A single block can run for miles in open country and barely cover a city street downtown. When you rank by block-level unserved counts, you're ranking by an artifact of how the boundary was drawn as much as by real demand.

Population grid products have the same problem from a different angle. They distribute population across a grid using proxies like nighttime lights or land cover, which smooths out exactly the thing a build priority list needs: where the settlement sits, how big it is, and whether it's one cluster or several. A grid cell showing moderate density might be one real town or five spread-out homesteads. For a subsidy bid or a multi-year build order, that distinction changes which drop gets funded this cycle and which waits.

The result shows up in build plans that get challenged during the comment period, subsidy applications that undercount serviceable locations in the areas that matter most, and crews sent to coordinates that don't match what's standing there.

Building a rank order that matches the ground

A build priority list that holds up starts from actual settlement clusters, not block or grid boundaries. For each cluster you want three things: where it sits, how big it is, and whether it's residential, mixed, or something else entirely. Rank on cluster size first, since a 60-home cluster three miles off the trunk line usually beats a 6-home cluster that happens to sit in a technically more remote block. Then layer in distance to existing infrastructure, terrain, and whatever weighting your subsidy program requires.

The practical sequence most planning teams land on:

• Pull current imagery for the target region and extract settlement clusters, not grid cells.
• Classify each cluster by rough size and type so you can tell a hamlet from a single farmstead from a mixed residential and light-commercial strip.
• Cross-reference clusters against the eligibility file to confirm unserved or underserved status.
• Rank clusters by size within eligibility tier, then adjust for engineering distance and program-specific scoring.
• Rebuild the list on a yearly cycle, since rural settlement grows and shifts faster than redistricting does.

That last point matters more than it sounds. A build priority list drawn once at the start of a grant program is stale by year two in any area with real growth. New clusters form, existing ones expand past what the original eligibility pass counted, and the list that looked right in year one misses those new builds by year three.

This is the gap Settlement Mapping is built to close: an annual high-resolution pass over your target region that turns the image into settlement clusters, segmented and classified by size and location, instead of population-grid guesses stitched to outdated boundaries.

A note on subsidy bids specifically

If you're building a rank order to support a specific subsidy round, keep the settlement layer separate from the scoring formula. The program's rules will tell you how to weight cost per location, speed tier, or local match. Settlement data answers a narrower, upstream question: how many real locations sit in this area and how are they grouped. Get that wrong and every downstream score is built on a bad denominator.

If your current build priority list is still running off the eligibility file alone, it's worth checking what a settlement layer would change before the next bid round closes.