9,188 More New Business Records Captured Across 4 Metros

The four municipal new-registration feeds we capture now hold 22,785 records, up from 13,597 at the start of the window, a gain of 9,188. That gain counts records newly captured by the clock. It is not a count of new businesses opened.

Most of the gain is the clock collecting for longer, not a rise in business formation. The rolling 90-day view tells a different story: Chicago and Los Angeles both fell, and New York City's rolling count only started to register as its feed was captured. Read the cumulative and rolling measures separately.

Like-for-like comparison of records captured from four municipal new-registration feeds (Chicago, Los Angeles, New York City, San Francisco) between August 7 and October 10, 2026. Each count covers one city's own municipal dataset, so nothing here is a national total or a complete city total.

Captured records rose from 13,597 to 22,785 in 64 days.

Key Findings

• Captured records rose from 13,597 to 22,785. Gain of 9,188, according to our sealed business-formation snapshots.

• Los Angeles holds the largest captured gain, at +4,017. Its captured count went from 6,557 to 10,574.

• New York City's captured count went from 303 to 2,137. Its rolling count went from 0 to 876 as capture started.

• The +782 rolling change is not growth. New York City's +876 is larger than the total.

• Construction share fell from 16.1% to 14.8% among coded records. Retail share rose from 14.3% to 17.6%.

The Comparison at a Glance

The headline and rolling totals are the four feeds added together, as the sections below show by feed. For the monthly census across these same four feeds, read the September index post.

Captured Records by Feed

Every feed rose on the captured measure, which is what a clock reading open feeds should produce. The spread matters more. Los Angeles added the most, and it is covered in depth in the Los Angeles formation report. New York City began the window with the lowest captured count, 303, so its +1,834 should be read against that low start.

Captured totals only accumulate while the clock runs. The rolling figure can fall, as Chicago's -17 and Los Angeles's -122 show.

The Rolling View Reads Differently

The rolling count is a net balance. The 90-day window is measured on each record's publication date: a record counts while that date falls within the 90 days before the snapshot, and it ages out once it passes the 90-day mark, so a negative figure means the balance fell over the window. It does not mean businesses closed. New York City's rolling count went from 0 to 876 because records with publication dates inside the window only began to be captured for that feed, so its +876 reflects a start-up effect.

Rolling net change in Chicago was -17; Los Angeles, -122.

The +782 total for all four feeds is not evidence of growth. New York City's +876 is larger than the total because the other three feeds' rolling counts net down.

Sector Shares Among Coded Records

Construction's share fell from 16.1% to 14.8%, a change of -1.3 points, while retail's share rose from 14.3% to 17.6%, a change of +3.3 points. Only the Los Angeles and San Francisco feeds publish a NAICS code, so these shares describe coded records from those two feeds alone. A 3.3-point move in one sector over 64 days can reflect the timing of captures rather than a change in what businesses are starting. Treat it as something to watch, not a finding.

What the Sources Can and Cannot Carry

A feed reports a record when the city publishes it, which can lag the day a business registers. The clock cannot see that lag, and a feed can pause or change format. A captured record is therefore a publication event, not a verified opening date. For how the activity-label feeds read in practice, see the Chicago formation report and the New York City formation report.

Captured records: 64 days, net listings, not outcomes.

Methodology: How the Comparison Was Built

Source attribution: US Tech Automations Research derived comparison, computed from two sealed business-formation snapshots, the earlier one taken at the start of the window and the later one sealed in October. Each change figure is an exact difference between two verbatim counts from those snapshots.

Honesty statement

Every figure is a verbatim count or an exact difference of two verbatim counts from the same sources. A net change in listed or captured records is not a count of new businesses, projects approved, or projects built. This is a short window, not a long-run trend. Nothing is estimated or modeled.

Each count is a census of one city's municipal dataset only. The four feeds are not a national total, and no city figure here is that city's complete new-business count. Los Angeles and San Francisco publish a NAICS code; Chicago and New York City publish a free-text licensed activity, which this comparison does not treat as a sector. Sector shares cover only the records that carry a code. Nothing is extrapolated, and nothing is estimated or modeled.

Known limits: this is a 64-day window; counts are net listings in captured records, not outcomes; and rolling figures are net balances, not gross registrations.

1. Collect. The clock reads each of the four municipal new-registration feeds on its own schedule.

2. Normalize. Each record keeps its feed's classification: the NAICS code where one is published, or the free-text licensed activity where one is not.

3. Dedupe to first-seen. Each record is counted once, on the day the clock first captured it.

4. Seal daily. The day's records are content-hashed into an append-only snapshot.

5. Compare like for like. Two sealed snapshots covering the same four feeds are compared count by count, and every change is that exact difference.

Frequently Asked Questions

Q: Why does the captured count rise when no new businesses have opened?

A: The captured count is cumulative. Each day the clock adds records it has not yet seen, so the total moved from 13,597 to 22,785 as the window ran. The +9,188 measures records newly captured, not businesses opened.

Q: Why did New York City's rolling count go from 0 to 876?

A: The rolling count was 0 at the start because none of the 303 records captured for that feed at that point had a publication date inside the 90-day window. By the end the rolling count reached 876. The +876 is a start-up effect, so it should not be read as growth or as a sign of new activity.

Q: Why are sector shares limited to coded records?

A: Only the Los Angeles and San Francisco feeds publish a NAICS code, so only those records can be grouped by sector. Chicago and New York City publish free-text licensed activities, which this comparison does not treat as sectors. The shares therefore describe coded records only.

Q: What does a negative rolling change such as -17 mean?

A: The rolling 90-day figure is a net balance: records count while their publication date is inside the window and age out after 90 days. A -17 change means the balance fell over the window. It does not mean businesses closed, because the count does not track outcomes.

Q: Can the 9,188 figure be quoted as new businesses across four cities?

A: No. It is a net change in captured records across four municipal feeds, each a census of one city's dataset. It is not a count of new businesses, and it covers a 64-day window rather than a long-run trend.

Put Formation Data to Work

Three roles fit this comparison. Each should work from the sealed tables and from the feed that matters to them, not from the headline total.

A B2B sales lead at a bookkeeping or payroll firm. Checks the Los Angeles and San Francisco coded feeds each week for records in the sectors the firm serves. The captured totals, 6,557 to 10,574 in Los Angeles, show where the feed is active, but the sales action runs on coded records in the target sector, routed to a rep the day they appear.

A business-insurance broker planning first-touch marketing. Reads the rolling table at the start of each month before committing spend. Chicago's -17 and Los Angeles's -122 show a falling recent balance, so the broker treats the +4,017 cumulative gain as a sign of longer capture, not demand, and moves budget only when a rolling net turns up.

A customer-service operations manager onboarding small-business accounts. Confirms each feed is still being captured before changing staffing. New York City's rolling count went from 0 to 876 as its capture started, so the manager checks that feed's status first, then reviews coded sector shares before deciding which new-account queues to staff.

US Tech Automations can be configured to watch each feed on a schedule, flag a sudden change in capture status, route new coded records to the right rep, and draft first-touch replies for review.

See the sales and customer-service agents built for this kind of monitoring.

Source: US Tech Automations Research — computed from two sealed business-formation snapshots, August 7, 2026 – October 10, 2026.