How to find local businesses without a website

Searching “plumbers in Fort Myers” shows you the plumbers who are already good at being found — exactly the ones who do not need you. You want the opposite: a complete listing of every business in a trade and an area, including the ones with no web presence to rank.

OpenStreetMap is the practical source. It is a public map database with a real query language, its business listings are contributed by people and businesses locally, and crucially a listing exists whether or not the business has a website. That absence is the signal you are hunting for.

“Has a website” is close to useless as a filter. A one-page site from 2011 with no mobile layout is a better prospect than no site at all in some trades, and a clean recent site means move on. What you are grading is whether the site is currently losing them work:

• No site at all — the strongest signal there is. Nobody has to be persuaded they have a problem.
• No mobile layout, so the site is unusable on the device most of their customers hold.
• No HTTPS, which browsers now label in front of the customer.
• Slow enough that a visitor gives up before it renders.
• Visibly dated — a design that puts the business a decade behind its competitors on the same search page.

This is the step most lists skip, and it is the one that decides whether the day was worth it. A bad website belonging to a business with no customers is not a lead. They cannot fund a rebuild and the site is not costing them anything, because nothing was coming through it anyway.

What you want is the opposite pairing: strong traction, weak site. A business with hundreds of reviews and a phone that rings, on a site that has not been touched in years, already knows the site is embarrassing. You are not creating the need, you are arriving at the moment it is felt. Rank by weakness multiplied by traction and work down from the top.

A contact form on their site is where pitches go to die. A real address is on the page somewhere most of the time — in the footer, on an about page, behind an obfuscation script, or written as info [at] example [dot] com to defeat exactly this. Checking the obvious paths by hand is minutes per business, which is what turns a list of two hundred into a list of twelve.

Lead Scraper is this method, run as one operation. You pick a trade and a location; it queries the listings, grades each business’s site, scores weakness against traction, crawls for a real contact address, and drafts a first email that names the specific thing wrong with that business’s site. The querying and the writing both run on your Mac.

When you have the list, the sending goes through your own mailbox rather than a shared pool, which matters more for whether the email arrives than anything in the copy.

Not with a search engine — a business with no website has nothing to rank, so searching a trade and a town returns the businesses that least need you. Start from a business directory instead. OpenStreetMap works because its listings are contributed locally and a listing exists whether or not the business has a website, so the absence itself becomes the thing you can filter on.

Only some of them, and the filter matters more than the list. A bad or missing website belonging to a business with no customers is not a lead: they cannot fund a rebuild and the site costs them nothing, because nothing was coming through it. What you want is strong traction paired with a weak site — a business with hundreds of reviews and a phone that rings, on a site untouched for years. Rank by weakness multiplied by traction and work down from the top.

Five signals, in rough order of strength: no site at all; no mobile layout, so it is unusable on the device most of their customers hold; no HTTPS, which browsers now label in front of the customer; slow enough that a visitor gives up before it renders; and visibly dated. A clean recent site scores near zero — move on.

It depends on the trade and the density, and anyone quoting a general figure is guessing. One measured example: a run against Collier County, Florida on 28 July 2026 returned 110 named dentists. That is a record of one run in one county in one trade, not a promise about yours — which is why the free tier exists.

Yes — the method above is manual work, not a product, and plenty of people run it with a map and a spreadsheet. It costs an afternoon per area. Lead Scraper automates the same method on a Mac (macOS 26+ · Apple silicon), with 3 scrapes a month free and no card, which is enough to work one county properly.