← All posts

How to Find New Businesses in Your Area: Filing to Lead

GuideAugust 14, 2026 · 12 min read · The LeadMarina team

You can pull 300 new business filings for your county before lunch. Now try calling one. The name on the record isn't the name on the sign, the address belongs to a law firm, and there's no phone number in the file at all — which is why lists of newly registered businesses so reliably produce zero conversations.

This guide ranks the newness signals by how many callable leads each produces, then walks the part other articles skip: turning a filing into a business you can reach, checking you can reach it, and timing the approach so you're not the fourth vendor calling before the doors open.

New business leads split into three different lists

Most confusion here comes from treating one word as one dataset. Three populations exist in any area, and they barely overlap:

  • Registered. Somebody filed paperwork with the state. Largest population, earliest signal, least qualified — a filing is an intention, not an address.
  • Permitted or licensed. Someone is spending real money on a specific location: a build-out permit, a food-service permit, a liquor application. Smaller, later, far likelier to open.
  • Open and trading. There's a listing, hours, a phone, the first reviews. Smallest population, latest signal, the only one you can call today.

What you sell decides which list you want. Signage, POS hardware, commercial insurance, entity formation, bookkeeping — the registration stage is where that money is. Marketing, staffing, linens, food distribution, local SEO — the same list wastes your quarter, because you need businesses that have opened.

For scale: the US Census Bureau's Business Formation Statistics has counted more than five million business applications a year nationally since 2021, and the Bureau reports a separate "high-propensity" series precisely because that headline number overstates how many applications become real employer businesses. Filings-only workflows inherit the gap between those two series.

Where to find newly registered businesses, ranked by callable yield

Ranked by usable conversations per hour of work, not by how early the signal fires. Portal quality varies by county; the ordering is the durable part.

1. Listing tells: a business profile only a few months old

Highest yield by a distance, because the record is already contactable. Google business listings don't publish a founding date, but newness leaves fingerprints: a low review count where every review is recent, hours and category filled in but no photos beyond the storefront, a domain registered this year, and — clearest of all — a business that appears in this month's search of a niche and city but wasn't in last month's.

That last signal is worth building on. Pull the same niche and cities on a schedule and the set difference between runs is your new-openings list, computed from data that already carries a phone number. No filing index gives you that.

The trap: a listing can be new without the business being new. Rebrands, relocations and shops that finally created a profile look identical to a real opening, so check the address against a permit record first.

2. County DBA and fictitious-name filings

The most underrated source here, because it's the one carrying the trading name. In most states an operator working under anything other than the registered entity name must file an assumed-name statement with the county clerk, and in several — California among them — publish it in a local newspaper. That gives you what a state filing rarely does: the name on the sign, and an address that's usually the real one. Clerk sites are inconsistent, but this index bridges entity and storefront.

3. Building and tenant-improvement permits

A commercial build-out permit is the strongest pre-opening signal there is: someone signed a lease and hired a contractor. Most mid-size and large US cities publish permit data through an open-data portal — Socrata and ArcGIS Hub instances are common, plenty of permitting workflows run on Accela — filterable by type, valuation and issue date.

You get the operating address, the valuation as a budget proxy, often the contractor and property owner. What you don't get reliably is the trading name, since permits are usually pulled under the entity or the contractor. Sort by valuation and work the top.

4. Health permits for recently opened food businesses

For anything serving food — restaurants, cafés, bakeries, food trucks, groceries — the county or state health department's inspection portal is the most precise opening calendar available, because a new establishment gets a dated pre-opening inspection that's public in most jurisdictions. High yield, low volume, the right trade. The limits: one vertical, and some counties publish a table while others post a PDF.

5. Liquor and alcohol license applications

State ABC boards publish pending applications, and because most states require public notice with an objection window, those applications carry dates, addresses and the applicants' names — one of the few public sources handing you a human being attached to an address. The catch is lead time: an application can sit months ahead of the opening, so it builds pipeline, not this week's call list.

6. Secretary of State business registrations

Every US state runs a free entity search, and most let you browse or export recent filings; OpenCorporates aggregates across jurisdictions if you need several states at once. Volume is enormous and the data authoritative — officer and registered-agent names live here. It sits sixth on purpose, for reasons that get their own section below: a registration confirms a business and names its people, but it doesn't produce contacts.

7. Chamber rosters, local press, and job ads

Small volumes, excellent quality, near-zero cost: chamber new-member announcements, the "now open" column in a business journal, and job ads for a location that isn't trading yet all identify real, staffed businesses.

Why a new business filing is not a lead

Here's the wall every article on this topic drives you into and then stops at. You export 300 registrations and nothing in the file is actionable. Three reasons.

The registered agent isn't the business

Every entity must name an agent for service of process, and the address on the filing is usually the agent's — a commercial registered-agent service, the formation lawyer, an accountant. Dial the number attached and you reach a call centre handling legal documents for tens of thousands of companies. Even when an owner is their own agent, that address is often a home or a mailbox. It's a legal service address; treat it as one.

The entity name isn't the name on the sign

"JMR Holdings III LLC" is the filing. "Bruno's Pizza" is the business. Search a listings database for the entity and nothing comes back, because no customer has ever typed it. This is the biggest reason new-business exports feel unusable: the join key you were given isn't the one the rest of the world uses. Two fixes, in order of reliability — the county fictitious-name index, which exists to record that mapping, and the street address, which survives a rebrand.

Plenty of applications never open anything

Holding companies, single-property real estate LLCs, defensive name registrations, entities formed for a deal that collapsed, and side projects abandoned in month two all file the same paperwork as the café opening on Main Street. That the Census Bureau reports high-propensity applications separately is an admission of exactly this. Filings are candidates, not prospects; qualification is a step you own.

Matching a filing to a business you can actually call

The chain the guides skip, in five steps. Do twenty by hand and you'll know what to automate.

  • Get the trading name. Check the county fictitious-name index, look for a "doing business as" field on any permit or licence record, or search the officer's name and see what they're attached to.
  • Anchor on the street address, not the name. Permits, health records and listings all carry the operating address, and it survives rebrands and ownership changes. Where a name match fails, an address match usually holds.
  • Find the trading business in listings data. Search the trading name plus the city, then confirm category and address line up with the permit — a name match alone across a metro is a coin flip.
  • Take contact details from the business, never the filing. The website, the listing's phone, the Facebook page. The filing contributes names; the business contributes contacts.
  • Use the filing for what it's best at: the human. Officers and organisers are named on state filings, which is a decision-maker you can ask for by name — our guide to finding a business owner's name covers the other sources and the honest hit rate.

Nobody matches every record: realistically a minority of a county's filings resolve to a trading business with contacts, and that minority beats the whole raw export.

Verifying contact details for newly opened businesses

New businesses are the worst case for unchecked contact data, for a reason easy to miss: everything about them is weeks old. The domain was registered last month, the mailbox set up by whoever built the site, the number bought online in an afternoon.

  • Email. A brand-new business skews hard towards a catch-all on a cheap host, or a personal Gmail printed on the door. Both accept mail; only one is read by anybody, which is why the check belongs at the address level — the labels are broken down in email verification statuses explained.
  • Phone. Expect mobile and VoIP, since a virtual number is among the first things a new owner buys. A mobile at a two-month-old business is very often the founder, but only the line type tells you which is which — the reasoning is in verifying phone numbers before cold calling.
  • Names. Match the officer name from the filing against whoever signs the review replies. When they agree, you have a decision-maker and an opener in one record.

Timing: when to contact a newly opened business

The same business contacted at three points gives three different conversations.

  • Pre-opening, permit issued, doors closed. They're buying build-out, signage, POS, insurance, licensing help and staff. Nobody answers a business line because there isn't one — email and the owner's mobile are the only channels.
  • Opening week to month two. The worst time to sell most things: the owner is working eighteen-hour days and every vendor in the county has already called. If your offer isn't urgent right now, wait.
  • Month two to month six. The best window for most agency and service offers — the rush has flattened, the owner has learned which assumptions were wrong, and there's budget with anxiety behind it.
  • Month six to year one. Efficiency buying — bookkeeping, payroll, cleaning contracts, insurance renewal. Also when vendors hired in a panic during week one get replaced. Whichever stage you work, the list wants a date column: permit issue date, licence approval date or first-review date all serve as a rough clock.

Run your new business list as a standing weekly pull

New businesses are a moving target, and a list of new openings goes stale in weeks. So repeat mechanics matter more here than the initial build: if your process can't be re-run cleanly, you'll rebuild it by hand monthly or quietly stop.

Our bias, up front: LeadMarina is ours, so weigh the next two paragraphs accordingly — the workflow logic holds in anything pairing listings data with scheduled re-runs. It finds local businesses by niche and city from Google business listings data, as many as 30 cities in one bulk search, and what arrives is a finished record rather than a name and a maybe: as many as three emails, each probed at the mail server and graded safe, risky or invalid; as many as three phones, each stamped with line type and carrier; an owner name where one was identifiable and a blank where it wasn't; socials, ratings, reviews and the profile. We don't publish accuracy percentages, and we'd be sceptical of anyone who does on this kind of data; field definitions are the honest version.

Two properties fit this job. A search can be put on a schedule — once, daily, weekly, monthly or yearly — pointed at any destination, which turns a static pull into a feed. And exports match on a stable Google business identifier and update in place, so re-running the same niche and cities edits rows you already have rather than minting a second copy of your territory. That's what makes the diff work: first-time rows are the new entries.

Destinations are Close, GoHighLevel, Google Sheets, and emailed CSV, Excel or JSON. None overwrite what you created by hand: notes, stages and custom fields survive a refresh, and GoHighLevel tags are added rather than stripped. To compute the diff yourself, the REST API is on every plan including the free one, and the MCP server's seven tools work from Claude, ChatGPT developer mode, Cursor or any MCP client.

What listings data can't tell you is a filing date, because that field doesn't exist in it. If your offer depends on catching businesses before they open, the permit and licence sources above are irreplaceable — use them for discovery, listings data for contactability.

Before you contact anyone: the legal part

None of this is legal advice, and no dataset — certainly not ours — makes an outreach campaign lawful. Sourcing a number tells you whether you can dial it, not whether you may. Calls and texts answer to the TCPA and state telemarketing rules, commercial mail to CAN-SPAM, and anyone you reach in the EU or UK pulls GDPR into the picture. Which of those bite depends on your campaign and jurisdiction, which is a conversation for counsel.

One wrinkle is specific to new businesses: a brand-new sole proprietor's business number is very often the founder's personal cell, exactly where the line between B2B and consumer calling gets uncomfortable — DNC scrubbing for B2B cold calling covers that hygiene. And a filing is public information about a person as much as a company, so handle an officer name from a state register with the care you'd want applied to your own.

How to find new businesses in your area: quick reference

  • Decide which "new" you need — registered, permitted, or open — because your offer only sells to one of them.
  • Work sources by callable yield, never dial a registered-agent address, and bridge the entity name to the trading name through the county fictitious-name index.
  • Check email at the address level, resolve phone line type, date every record, and re-run on a schedule matched to a stable business identifier.

Newness and contactability sit at opposite ends of the public record: filings are early and uncallable, listings callable and undated. Everything useful happens in the join between them, and the teams who own a territory built that join once, then ran it weekly. To test the contactability half on your own niche, the free plan includes 100 fully verified leads — pull one city and count how many of the newest listings come back with a mobile. Start free.

SharePostLinkedInShare

Keep reading

Try it on your own market.

100 free leads — every email and phone verified, API included, no credit card.

Get started free