Thirty thousand scored businesses across 286 ZIP codes in Kansas and Missouri. Where SBA access is weakest, businesses are no smaller — median revenue is flat across the entire access gradient — but they carry more critical gaps, score fifteen points lower, and are funded at two-thirds the rate. The constraint is not scale. It is the distance between a viable business and a bankable one.
Which ZIP codes should this capital go to, and why those?
Everything below is the evidence for those three sentences. You can read only this box and still follow the argument. Underlined terms open a plain definition — the full glossary has them all.
ZIP codes sorted into five bands by the SBA access index carried in the platform data. Every readiness measure moves monotonically across the bands. Business size does not move at all — which is what makes the gradient interpretable.
Three lines track readiness, funding rate and revenue across access quintiles, each indexed to 100 at the lowest-access band.
A rising line means that measure improves as access improves; the flat revenue line shows business size does not track access at all.
Readiness, funding rate and median revenue across access quintiles, each indexed to 100 at Q1.
Every access quintile with its ZIP count, mean readiness score, share ready-or-above, critical gaps and funded rate.
Read down each column — every one rises from Q1 to Q5, showing readiness and funding both improve steadily with access.
286 ZIP codes, 30,000 businesses.
| Band | ZIPs | Score | Ready+ | Crit gaps | Funded |
|---|---|---|---|---|---|
| Q1 lowest | 67 | 163.6 | 32.4% | 1.99 | 5.17% |
| Q2 | 53 | 165.9 | 32.6% | 1.92 | 5.23% |
| Q3 | 53 | 167.9 | 34.1% | 1.85 | 5.43% |
| Q4 | 69 | 171.3 | 36.6% | 1.74 | 6.19% |
| Q5 highest | 44 | 174.9 | 39.1% | 1.7 | 6.58% |
$304,318 · $285,010 · $310,084 · $304,024 · $300,808
No trend. A business in the least-served quintile is the same size as one in the best-served quintile. What differs is whether it can document itself to a lender.
This is the gradient the research design predicted, and it holds. But it holds on readiness and funding, not on business scale — which narrows the claim and strengthens it at the same time.
A capital-availability story would predict smaller businesses in low-access areas, because undercapitalised firms stay small. A readiness story predicts equivalent businesses that cannot clear a documentation and credit threshold. The data shows the second. That is a solvable problem, and it is the one this program is built to solve.
Priority is the intersection of the bottom third on SBA access and the bottom third on mean readiness — a transparent two-axis rule, fixed before the run, auditable, and replaceable by the Economic Mobility Index decile the moment Kauffman delivers the tract file.
Nine in Wyandotte County, twelve in the Missouri-side urban core. 2,328 businesses, $277M of unmet supportable capital.
The rule also selects 18 ZIPs in non-metro Kansas, 15 in non-metro Missouri, and 9 in the St. Louis urban core. If the target frame stays both states rather than the KC metro, the rural share of priority geography is larger than the KC share — which is a decision for the Research Lead, not an analytical result.
Nine measures comparing the 63 priority ZIPs against the remaining 223 ZIP codes statewide.
Priority ZIPs trail on access, readiness and funding despite carrying near-identical revenue to the rest of the state.
| Measure | Priority | Rest |
|---|---|---|
| ZIP codes | 63 | 223 |
| Businesses | 6,332 | 23,668 |
| Access index | 0.72 | 1.08 |
| Mean readiness | 160.0 | 170.8 |
| Ready or above | 29.8% | 36.2% |
| Pre-readiness | 12.2% | 9.3% |
| Critical gaps | 2.11 | 1.77 |
| Median revenue | $295,050 | $304,024 |
| Funded rate | 4.85% | 5.95% |
Ten and a half readiness points and 0.34 critical gaps separate the two groups. Median revenue separates them by 3%.
Wyandotte and Johnson counties, one state line apart, from the platform rather than the loan register.
| Measure | Wyandotte | Johnson | Gap |
|---|---|---|---|
| Businesses scored | 1,027 | 2,742 | — |
| Mean readiness score | 159.4 | 175.1 | −15.7 |
| SBA access index | 0.58 | 1.42 | 0.41× |
| Ready or above | 29.7% | 39.2% | −9.5 pts |
| Critical gaps per business | 2.10 | 1.69 | +24% |
| Thin credit file | 34.2% | 28.8% | +5.4 pts |
| Days to capital-ready | 158 | 135 | +23 |
| Funded rate | 4.58% | 6.71% | 0.68× |
| Median annual revenue | $290,020 | $304,356 | 0.95× |
| Employees per business | 1.76 | 1.95 | 0.90× |
| DSCR proxy | 1.107 | 1.162 | 0.95× |
The three rows at the bottom are the point. On the measures a lender would call fundamentals — revenue, employment, debt-service capacity — Wyandotte businesses are within five to ten percent of Johnson businesses. On the measures that determine whether a file gets approved, they are twenty-four percent worse and twenty-three days further away.
Thirty thousand businesses across stage, sector, size and ownership. Two of these cuts carry large variation and two carry almost none — and which is which is a finding in itself.
All 30,000 businesses split into five readiness tiers, each with its firm count, share of population, DSCR and routed lending product.
Reading down, DSCR falls as tier worsens — the bottom two tiers cannot service debt at the product they would otherwise be routed to.
Current tier, with the product the platform routes to.
| Tier | Firms | Share | DSCR | Routed to |
|---|---|---|---|---|
| Strong Readiness | 3,142 | 10.5% | 1.44 | 504 / bank term |
| Ready | 7,347 | 24.5% | 1.31 | 7(a) / bank term |
| Moderate Readiness | 8,785 | 29.3% | 1.16 | Express / LOC |
| Developing | 7,764 | 25.9% | 0.97 | Microloan / CDFI |
| Pre-Readiness | 2,962 | 9.9% | 0.80 | No product |
Median firm age is 3.7 to 3.9 years across every cluster; mean employment 1.8 to 2.2. This is a population of very small, established-but-young businesses — not startups, and not employers at scale.
A horizontal bar for each of twelve sectors, ranked by mean readiness score.
Longer, green bars are higher-readiness sectors; the amber bars at the bottom show where readiness is weakest.
Mean readiness score and share reaching Ready or above.
24.3 points separate Construction from Gig & Platform — and 15.4 percentage points separate their share reaching Ready. Sector is the largest single discriminator in the dataset.
By ownership, there is almost nothing to report — and that is the report. Mean readiness ranges from 166.2 to 169.9 across every race and ethnicity category, a spread of 3.7 points on a 300-point scale. Share reaching Ready or above ranges from 34.0% to 36.0%. Gender shows the same flatness; veteran status shows none. Thin-file rates sit between 28.8% and 33.5% for every group.
Two readings are available and Phase 2 should carry both. Either readiness genuinely does not vary by owner demographics once a business is on the platform — or the platform's intake population is already selected in a way that removes the variation, in which case the businesses that never reach intake are where the disparity lives. The second is testable against the intake-channel field and should be tested before this result is published.
129,641 open and closed gaps across ten dimensions. A third of them resolve in under two weeks and change nothing about creditworthiness. The ones that do change it take four to eight months.
All ten readiness dimensions with their gap count, mean days to close, mean readiness points earned, and share of gaps rated critical.
The dimensions at the top close fastest but pay the least; the ones lower in the list take longer and matter more to creditworthiness.
Sorted by mean days to close. Critical share is the proportion of gaps in that dimension rated critical.
| Dimension | Gaps | Days | Points | Critical |
|---|---|---|---|---|
| Packet Documentation | 42,165 | 13 | 0.0 | 16% |
| Licensing & Compliance | 9,746 | 39 | 11.1 | 65% |
| Behavioral Signals | 10,638 | 63 | 13.3 | 65% |
| Cash Flow Health | 10,090 | 103 | 14.6 | 83% |
| Workforce Stability | 11,117 | 110 | 9.0 | 35% |
| Debt Service Capacity | 7,688 | 124 | 15.4 | 73% |
| Revenue Consistency | 9,971 | 131 | 13.3 | 63% |
| Business Maturity | 7,873 | 140 | 12.1 | 54% |
| Industry Benchmarks | 7,516 | 190 | 11.5 | 55% |
| Owner Credit | 12,837 | 240 | 7.1 | 19% |
The five most common gaps in the entire dataset are all documentation: no profit and loss statement (7,005), no entity formation documents (6,914), no two-year tax returns (6,909), incomplete debt schedule (6,882), no balance sheet (6,853). Together they are a third of everything recorded and they close in a fortnight.
They are also worth zero readiness points, because they are packet items rather than credit facts. Closing them makes a business submittable. It does not make it approvable.
The dimension that most needs work is the one that responds slowest and pays least. Owner credit carries 12,837 gaps, takes 240 days on average, and returns 7.1 points — the worst ratio in the table. In Wyandotte, 34.2% of owners have a file too thin to score at all.
A technical-assistance program that measures itself on gaps closed will report excellent numbers while moving almost no one into a credit box. The measure has to be tier migration, and the honest timeline for that is two to three quarters, not one.
$4.62B of supportable capital identified, $518M deployed. The failure is not evenly distributed across the tiers — it has a precise location, and it is not where microlending sits.
| Tier · routed product | Firms | Avg supportable | Total supportable | Funded | Funded rate | Unmet | DSCR |
|---|---|---|---|---|---|---|---|
| Strong Readiness · SBA 504 / competitive bank term | 3,142 | $615,417 | $1.93B | 818 | 26.0% | $1.57B | 1.44 |
| Ready · SBA 7(a) / community bank term | 7,347 | $230,774 | $1.70B | 785 | 10.7% | $1.55B | 1.31 |
| Moderate Readiness · SBA Express / community LOC | 8,785 | $89,047 | $782M | 112 | 1.3% | $773M | 1.16 |
| Developing · SBA Microloan / CDFI | 7,764 | $26,305 | $204M | 0 | 0.0% | $204M | 0.97 |
| Pre-Readiness · development pathway, no product | 2,962 | — | — | — | — | — | 0.80 |
| Total | 30,000 | $4.62B | 1,715 | 5.7% | $4.10B |
The cliff is at Moderate Readiness. It is the largest tier in the dataset — 8,785 businesses, 29.3% of the population — it carries $782 million of supportable capital, its debt-service proxy is 1.16, and its funding rate is 1.3%.
The tier above it funds at 10.7% on a DSCR of 1.31. The tier below it funds at zero. Moderate Readiness businesses can service debt, need tickets averaging $89,000, and are not being served by anything. In the KC metro alone this tier is 3,266 businesses and $291 million.
The Developing tier is where microloan and CDFI product routes, and it is the natural home for a microlending vehicle. It contains 7,764 businesses with $204M of supportable capital at an average ticket of $26,305.
Its mean debt-service proxy is 0.97 — below one. On average, these businesses do not currently generate enough cash to service the debt they would take on. That is the same population, and the same credit profile, that produced Community Advantage's 28.3% charge-off rate on a book only 36% resolved.
Microlending would reach real businesses that no one else is reaching. It would also lend into a sub-1.0 coverage ratio. Both are true, and an instrument built here needs its loss assumptions set against Community Advantage experience rather than programme-wide 7(a) averages.
The binding constraint is at $50–150K, not under $50K. Moderate Readiness has coverage above 1.15, an average need of $89,000, and effectively no product reaching it. That is a gap in the market, not a gap in borrower quality.
The Developing tier needs readiness capital, not credit. 7,764 businesses at DSCR 0.97 plus 2,962 at 0.80 is 10,726 firms — 35.8% of the population — for whom the correct intervention is preparation. A vehicle that treats them as borrowers will book the losses that follow.
Two products, as Erika framed it, maps cleanly onto this. A facility at the $50–150K band with TA-conditioned underwriting, and a preparation pathway with no credit attached. The second manufactures borrowers for the first, and tier migration is the measurable link between them.
Engagement works and is measurable. What follows engagement is where the volume is lost.
Six measures comparing businesses that engaged an action plan against those that did not.
Every measure — score change, gaps closed, lender matching, funding — is far better for engaged businesses.
| Measure | Engaged | Not engaged |
|---|---|---|
| Businesses | 14,374 | 15,626 |
| Mean score change | +5.40 | 0.00 |
| Gaps closed | 1.58 | 0.00 |
| Current score | 184.7 | 154.0 |
| Lender-matched | 33.8% | 16.4% |
| Funded | 8.4% | 3.2% |
Engaged businesses fund at 2.6 times the rate. Note the direction is not established — more capable businesses may be likelier to engage. The score change, however, is only observed among engaged firms, and that is a genuine treatment signal.
Score change and gaps closed for engaged businesses, grouped by how long they have been on the platform.
Score gains keep rising through month twelve and then flatten, while gaps closed per period barely moves — later gaps are harder, not fewer.
| Tenure | Firms | Score change | Gaps closed |
|---|---|---|---|
| 1–3 months | 8,206 | +0.97 | 0.75 |
| 4–6 months | 6,540 | +2.48 | 0.76 |
| 7–12 months | 7,910 | +4.30 | 0.77 |
| 13–18 months | 3,961 | +4.47 | 0.75 |
| 19–24 months | 354 | +4.63 | 0.79 |
Movement is real through month twelve and then flattens at roughly four and a half points. Gaps closed stays constant at 0.75 to 0.79 throughout, meaning later gaps are harder, not fewer. Crossing a tier boundary takes more than the platform alone currently delivers — which is the argument for paid, accountable technical assistance rather than self-serve remediation.
The leak is at the lender, not the borrower. Half of engaged businesses reach a lender match; fewer than a quarter of matches convert to funding. Whether that is credit-box mismatch, pricing, documentation friction at underwriting, or lender capacity is not answerable from this data — it needs the originator assessment across the lender types, run as a fixed instrument. That is the highest-value open question in Phase 2 and it sits on the supply side, which is exactly where Erika asked whether anything was missing.
Covered. Target ZIP profile with distress and access indicators; business stratification across stage, sector, age, size and ownership; needs and barriers via the gap taxonomy; current funding routes and capital-needs findings by tier.
Not covered — EMI anchoring. The priority ranking uses the SBA access index in place of the Economic Mobility Index, because the tract file has not arrived. The two-axis rule is built to accept the EMI decile as a drop-in replacement; the ranking will change and the method will not.
Not covered — comparative analysis. The 3–5 NBA member communities require the same data outside Kansas and Missouri, which does not exist in this export. Opportunity Atlas remains the portable measure for that arm.
This is platform population, not business population. 30,000 scored businesses are not a census of Kansas and Missouri firms. Every rate here is conditional on reaching intake, and the intake-channel field should be used to characterise that selection before any figure is published externally.
Pricing is not in this export. Observed cost of capital — the decisive test for whether a facility improves on what these businesses already pay — is not present. It is available from transaction enrichment and should be added.
The access index construction is not documented here. It is used as an ordinal throughout and nothing in this analysis depends on its scale.
Direction of causation is untested. Readiness and access move together at 0.57. Whether weak access depresses readiness, or low readiness suppresses access, or both respond to something else, is not resolvable from cross-sectional data.
This is document 5 of 5 in the reading path. Back to the Research index.