Church planting site selection: compare neighborhoods before you commit
A tract-by-tract way to narrow a city to a few candidate neighborhoods, then go and walk them.
Use the data to replace a windshield survey with a defensible shortlist. Use prayer, elders and your own feet to make the call.
How should a church choose a planting site?
Church planting site selection with GospelGraph is a three-stage narrowing: choose a metro by counting the tracts where evangelical Gospel access is thinnest, shortlist 8–12 candidate census tracts using the Gospel Desert Index with its domain breakdown, then assess each candidate on the ground with a field report and a visit. The data does the first two stages. Accountable church leadership does the third, after prayer, interviews and walking the streets.
That division matters. A planting commitment costs a family years. No map should commission a planter. What a map can do is keep a team from mistaking a well-served neighborhood for an unreached one, or from overlooking the subdivision nobody on the team has driven through.
Which metro or region should we enter?
Network leaders deciding which cities to enter can roll tracts up to the metro and rank metros by how many Desert-tier tracts they hold and how many people live in them. The Gospel Desert Index runs 0–100, higher meaning less access, and only 793 of 83,344 scored tracts reach the Desert tier (75 and above). Another 15,846 are Arid (60 to under 75). Public tools on GospelGraph usually treat 60 and above as a Gospel desert in the broad sense.
Do not assume the answer is the urban core. The desert moved to the growth edge found the steepest Gospel-access gaps in new subdivisions and exurbs rather than in urban cores, and showed why counting the unreached one way inverts the answer. The Frontier the PCA Is Not On maps where one denomination plants against the frontier it is not on.
How do we shortlist candidate neighborhoods?
Inside a metro, filter the map explorer to Arid and Desert tracts and look at each candidate’s domain breakdown. Then put the access gap beside need pressure, the weight the neighborhood carries now. The methodology splits the two at a GDI of 60 and a need pressure of 50, which gives four kinds of place that call for four different postures.
- Desert (GDI 60+, need 50+): high need, thin access. Prioritize for planting assessment, partnership and field validation.
- Empty (GDI 60+, need under 50): low need, thin access. Little to draw on when pressure rises; build reach before it is urgent.
- Healthy (GDI under 60, need 50+): high need, adequate access. Reach is not what is missing; answer the weight of the place, often alongside churches already there.
- Quiet (GDI under 60, need under 50): low need, adequate access. Build in the quiet season, before pressure arrives.
What should a planting team check on the ground?
Run a free field report on an address in each candidate tract and read four things. First, the five Shalom domains ranked by pressure: two tracts with the same composite can carry their weight in very different places. Second, the Gospel Impact Metric, the change in Shalom from 2020 to 2024, which tells a hard place that is getting better from a comfortable one coming apart. Third, built form: whether the streets have third places, intersections and public rooms where people run into each other, or whether a new church will have to create every meeting from scratch. Fourth, the neighborhood personalities present, because the households on one block can want very different things from a church than the households on the next.
Then go. Visit at different hours. Talk to the pastors already serving nearby, the school secretary, the shop owner. The data can tell you a tract has a high poverty rate and no nearby evangelical church. It cannot tell you about the widow in Building C who leads a Bible study for twelve of her neighbors.
What can site-selection data get wrong?
Church-location data undercounts small, storefront and immigrant congregations, which are under-registered in every source, and it may still count churches that have closed. It also carries a survivorship problem: failed plants and dissolved congregations are absent from every dataset, so a tract’s history of hard ground is invisible. Ask locally who has tried before.
A high GDI is not a statement that a tract has no church. The count of congregations inside a tract is almost uncorrelated with the score (r = +0.02); distance to the nearest church is what drives it (r = +0.44). Check the church layer on the map and the churches just over the boundary.
Finally, a plant’s survival depends on more than its neighborhood. Which Churches Stand On Their Own finds that in the Assemblies of God, the hardest ground is held by churches a district is carrying, and that a church’s class is predicted better by its district than by its neighborhood. Plan the support, not only the site.
Common questions
What data should church planters use before selecting a site?
Can the Gospel Desert Index pick a planting site for us?
Does a high GDI mean there is no church in the neighborhood?
Can we compare several candidate neighborhoods?
Read your candidate neighborhoods
Run a free field report for an address in each one, then open the map to compare the tracts around them.