Public data · Florida · ZIP-code grain
A statewide market screen built entirely on public data — and an honest account of what public data can and cannot see. Most market maps quietly hide their gaps. This one features them — because the gaps are where the real work begins.
Two buyers, two questions
Two very different buyers look at a business like this. A search fund building a platform needs to know which markets to hunt across — a portfolio question. An individual buyer staking an SBA loan on one location needs to know about one corner — an existential one.
This statewide screen narrows a whole state to the handful of markets that earn a closer look; the per-market drill-down answers the next question — whether one specific corner does too. Most of what follows is written for the second buyer — the one for whom this is the only deal that matters.
The screen
Every Florida ZIP code (ZCTA) is scored 0–100 for laundromat demand fit, from U.S. Census data on a transparent five-part rubric — the industry customer profile, weighted and disclosed, not a black box. 983 ZIPs scored.
The limits are the point
Public place-data undercounts small businesses badly and unevenly. We hand-reconciled three markets against the ground truth. In one, a public floor of 12 laundromats was really 25 — a 52% miss.
So a low floor is a candidate to verify on the ground — never "underserved." A high floor is reliable enough to screen out. The screen says where to look; the corner beats the county.
Screen, then drill
Keep only the ZIPs where demand is high and the supply floor is low — high demand-fit (≥ 70) and a floor at or below 1. That is the honest definition of "look closer": strong signal, thin visible competition (which the floor may be undercounting). 48 candidates surface statewide.
The filter narrows attention; it does not conclude. Each surviving ZIP is a question to answer at street level, not an answer.
The shortlist's #1
The screen ranks ZCTA 33136 first in the state on demand fit (). The mechanism is structural, not a matter of place: laundromat demand rises where households rent and lack in-unit laundry — that points to dense, high-renter, multifamily housing. On the measured inputs, 33136 is renter-occupied and multifamily, and all five demand components score high.
The supply read is the honest part. Public place-data shows zero laundromats inside this ZIP. Given the calibrated 50–52% undercount, a floor of zero is most likely an undercount, not an empty market. This is a candidate to verify on the ground — at trade-area radius, with a real supply walk — before any conclusion. The screen has done its job: it says look here, and exactly what to check.
From screen to diligence
The credential isn't a verdict on 33136; it's the discipline of separating what the public method can show from the question you carry into diligence — ranked by what actually decides a laundromat.
The public method shows — commercial water/sewer rate schedules are public and vary widely by municipality: a real, knowable margin input for any laundromat. (This screen does not yet display a per-ZIP rate — that's a build item, not a number to invent.)
You take to diligence — the store's actual water bill, private and obtained in diligence, is the revenue truth-serum: metered gallons back out wash-cycle volume and real revenue, cross-checked against the seller's P&L. Market rate is context; the bill is proof.
The public method shows — a floor, not a count. The statewide screen is ZCTA-grain, and public place-data undercounts supply by roughly half (the calibration above: ~50–52%). Statewide, that is honestly all it can claim.
You take to diligence — the real corner count is a radius drill-down: the Overture floor, reconciled against records, then a drive-by to verify on the ground. The same method as the calibration — run at one address.
The public method shows — nothing. The engine has no line of sight into a specific store's machine vintage — and pretending otherwise would break the whole premise.
You take to diligence — a capex item the public data does not cover: machine age, water efficiency, and remaining life, inspected on site. Named here because the gaps get labeled too.
The public method shows — context only: local density and labor-cost signals that shape how hands-on a location tends to be.
You take to diligence — operator diligence answers it: staffing, hours, and the attendant model for this specific store.
The same honest engine, a different market
Point the same engine at self-storage and the rubric changes with the customer: storage demand skews to higher-income, fast-growing single-family exurbs — the opposite income direction from laundromats. The cleanest proof is a single ZIP.
ZCTA 33136 — the #1 laundromat market — is near the bottom for self-storage. Same place, opposite ends. That is not a bug; it is the method being honest about who each business actually serves.
Every figure above traces to a public dataset and a disclosed method. The limits — the floor, the ZIP grain, the provisional calibration — are stated in the open, because a screen that hides its gaps is worth less than one that names them.
Sources & method
Source: ACS acs/acs5 2023 (ZCTA); TIGER/Gazetteer 2020/2023. Self-contained · opens offline · Leaflet 1.9.4 + topojson-client 3.1.0 (vendored, tile-free).