Raw counts can't compare two markets. Our rates can.
Net formation and closure per 1,000 active businesses — at county × NAICS × month grain, normalized across state SOS filings. So Polk-County Construction and Travis-County Professional Services sit on the same axis, not two different scales. Dashboard and API.
Coverage
IL & CA in the expansion sequence
“Travis looks ~6× hotter.”
Anyone can count filings. The rate is the hard part.
Scraping a Secretary of State portal gives you counts. The signal that actually informs a decision is the rate— and the rate is only as good as the math underneath it. Two things we do that raw-count feeds don't:
Every signal is a rate, not a raw count
A county with ten times the businesses always posts more formations — raw counts just rank by size. We normalize every signal to a rate per 1,000 active businesses, at county × NAICS × month. That's the grain where a small county can out-rank a big one, and you can see it.
The rate is only as honest as what it divides by
We divide by the businesses actually active when the period started — a consistent base, so the same rate means the same thing across every market and month. Too few businesses to measure? You get no number, not a fake one. No three-business neighborhood posting a 400% formation spike.
Formations + Closures + Survival. One Feed.
Small-business filing data is scattered across 50 Secretary of State portals — each with its own schema, search interface, and update cadence, and all of it in raw counts. We pull the corpus nightly, normalize it, and compute the rates so you can score local-economy health, prospect tenants, or underwrite policies in a single query.
Market Signals — county × NAICS × month, as a rate
Formations, closures, and net formation rolled up by state, county, 2-digit NAICS, and month — normalized to per-1,000-active-business rates so any two markets compare on one axis.
Drill any market down to a 2-digit sector
Split any state or county rate by NAICS sector to find where the formation pulse concentrates and where the closure rate bites.
The raw event stream behind every rate
The time-series of formation_date and dissolution_date across the full registry — the signal every aggregate is built from.
Every entity, normalized across states
Per-entity records from state SOS portals, one normalized schema across every source.
Early-closure detection, maturing with the data
A boolean flag on entities that dissolve within ~18 months of formation — the leading indicator for thin-market segments where churn outpaces organic growth.
Built for insurers
Get the event the moment it lands
Subscribe a county, NAICS sector, or business name and get a webhook the instant a formation or dissolution posts.
Built for insurers
Two Markets, One Axis.
Pick two markets and watch raw counts and normalized rates tell different stories. Sample figures below are illustrative while the production crawl completes its baseline pass — the methodology is live; the numbers populate per source.
Market A
Market B
Leads on raw count
Travis County
Leads on rate / 1K active
Polk County
Ranking reversal: Travis County wins on raw filings, but Polk County wins on the comparable rate. Raw counts would point you at the wrong market.
Formations
312
Closures
168
Net Formation
+144
Formation Rate
12.4/ 1K active
Closure Rate
6.7/ 1K active
Early Closure %
16.2%
Formations
1,840
Closures
1,015
Net Formation
+825
Formation Rate
9.1/ 1K active
Closure Rate
5/ 1K active
Early Closure %
14.8%
Read: Polk posted 312 new construction entities — a smaller raw count than Travis, but against a smaller active base it's a 12.4 / 1K-active formation rate. On the axis that matters, Polk is the hotter market.
Illustrative only. Live numbers populate as the nightly SOS scrapers complete a baseline pass against each source.
What Subscribers Do With the Data
Real workflows from the four personas we serve.
Score Local-Economy Health for Loan Originations
CDFIs and small-business lenders watch net formation rate by county and NAICS sector to gauge where new-business activity is rising or contracting. Pair with neighborhood heat maps to size loan books against a rate you can compare across submarkets — not raw counts and not last decade's census tract data.
Surface Tenant Pipeline From Formations
Commercial landlords and brokers track LLC formations in target ZIP codes and industry codes to identify newly-formed entities — the signal often arrives weeks before a tenant is shopping space. Rank target ZIPs by formation rate, not just raw volume, so a dense urban core doesn't drown out a faster-growing suburb.
Underwrite Commercial Policies With Survival Data
Insurance underwriters pull early-closure rates by industry and geography to price premiums against actual survival rates by industry and geography — normalized, so a high-volume sector isn't mistaken for a high-risk one. Cross-reference with compliance signals to identify high-churn segments.
Monitor Formation/Closure Trends for Policy + Research
Economic development offices and policy researchers consume the county × NAICS × month aggregate feed — formations, closures, net formation rate — to track the small-business pulse of a region. Compare your region against peers on the same per-1,000-business axis — the comparison most published data can't make at county × NAICS grain.
Always As Fresh As the SOS Allows
Nightly crawls feed the entity registry; the nightly aggregator recomputes the county × NAICS × month signals. Insurer subscribers get webhooks the moment a formation or dissolution event lands.
| Source | Update Frequency | Origin |
|---|---|---|
| Iowa SOS + DOR | Nightly | Socrata feeds for SOS active entities + DOR retail-sales permits |
| Texas Comptroller | Nightly | data.texas.gov franchise-tax holder feed (proxies paid TX SOSDirect) |
| Florida DOS (Sunbiz) | Daily incremental + quarterly full bulk | FL DOS official SFTP bulk feed (~4.95M entities) |
| Market Signal Rates | Nightly | Computed per 1,000 active businesses from biz_entities (county × NAICS × month grain) |
| Webhook Alerts | Real time (next nightly pass) | Triggered on diff vs. last snapshot |
| Illinois SOS | Next in sequence | ilsos.gov corporate / LLC search |
| California SOS | Next in sequence | bizfileonline.sos.ca.gov |
Plans for Every Biz-Signals Workflow
Start with a 7-day free trial. Annual billing saves 2 months on every plan.
Insurer
For commercial insurance underwriters pricing business-survival risk.
- Everything in Lender
- Business survival rate by industry + geography
- 90-day early-closure signals
- Cross-vertical compliance correlation
- API access (10,000 req/mo)
- Unlimited export
- Webhook alerts on new formations + closures
Three States Live. Two Next.
Iowa SOS and the Texas Comptroller franchise-tax feed ingest nightly. Florida arrives via the official FL DOS SFTP bulk feed — daily incremental plus quarterly full bulk, ~4.95M entities loaded. We go live where the data is openly licensable and sequence the rest deliberately: Illinois and California are seeded in the registry and next in the expansion path. Live row counts surface here as each source matures.
- Iowaia-sos
- Texastx-sos
- Floridafl-dos
- Illinoisil-sosnext
- Californiaca-sosnext
Frequently Asked Questions
Which states do you cover?
Three states are live: Iowa (SOS active entities + DOR retail-sales permits, nightly), Texas (Comptroller franchise-tax holders, nightly — proxies the paid TX SOSDirect portal), and Florida (FL DOS official SFTP bulk feed, daily incremental + quarterly full bulk, ~4.95M entities loaded). Illinois SOS and California SOS are seeded in the source registry and next in the expansion sequence — gated on negotiated portal access rather than open feeds. Additional states are added quarterly based on subscriber demand.
What does “rate-normalized” mean, and why does it matter?
Raw formation counts can't be compared across markets — a county with ten times the business base will always post more new filings. Every market-signal row we publish is normalized to a rate per 1,000 businesses active at the start of the period, at county × 2-digit NAICS × month grain. That's what lets you rank submarkets or sectors on one axis instead of re-measuring population. Where a cell has no active base to divide by, the rate returns NULL rather than a fabricated spike — so a three-business cohort never shows up as a runaway formation rate. Rate-normalized signals are on the Lender and Insurer plans; the Analyst plan shows aggregate monthly trends.
How current is the data?
We crawl every Secretary of State portal nightly where the source supports it, and write a `last_scraped_at` timestamp against each source. Some state portals only refresh weekly or monthly upstream — the dashboard's data-freshness panel surfaces the actual lag per source. New filings typically surface within 24–72 hours of being posted to the SOS.
What is the early-closure flag and how is it computed?
Early-closure is a boolean on every entity record — TRUE when dissolution_date is within 18 months of formation_date, NULL until enough lifecycle data exists. It's the leading indicator behind the survival-rate aggregates exposed to Insurer-tier subscribers. Surface it as a screening signal for underwriting; it's not a substitute for full borrower or policyholder review.
Can I query by NAICS industry sector?
Yes. Lender and Insurer plans include NAICS 2-digit breakdowns on every market-signal endpoint — drill a state or county aggregate down to NAICS 23 (Construction), 54 (Professional Services), 72 (Accommodation + Food), or any other 2-digit sector. The Analyst plan exposes monthly trends in aggregate; per-NAICS slicing starts on the Lender plan.
How is the active-business denominator computed?
The denominator is the count of entities active at the start of each period for the given county × NAICS cell — a consistent cohort base, not a total-ever-filed figure and not an end-of-period count that drifts with the formations you're measuring. Holding the base consistent is what makes a rate comparable month to month and place to place. Cells with an empty cohort return NULL, never an invented denominator.
Is API access included on every plan?
API access starts on the Lender plan ($699/mo, 1,000 req/mo) and expands on the Insurer plan ($1,299/mo, 10,000 req/mo). The Analyst plan ($299/mo) is dashboard-only — most economic researchers run ad-hoc queries through the dashboard's export rather than programmatic pulls.
Is there a free trial?
Every plan includes a 7-day free trial with full access to that tier's features. No refunds are issued after a paid subscription begins — see the Terms of Service for details.