$20 setup → $26,957/month
The play is two moves: find a local shop with no website, then put a finished one in front of the owner before you pitch. AI handles both. A 26-year-old line cook in Tampa ran it on his days off, cleared $26,957 in 30 days, and started on about $20. He has never written a line of code.
He tried it on foot first. Between shifts he walked into shops and asked if they wanted a website. Owners waved him off before the laptop opened, and the polite ones said they would think about it. An owner cannot picture a site he has not seen. So he built it first, walked back in, and turned the screen around. The maybe became a yes. Finding the shops and building the sites both run inside Claude now. The close is the only part he still does by hand.
1. Open Cowork on the $20 plan
Even the cheapest Claude plan opens Cowork, the automation surface inside the desktop app. Build one project, name it Google Maps Lead Generation, and write the project instructions first: your name, the trade you target, the city, what you sell, the tone you want on outreach. Skip this and you pay for it later in misfires that take an hour to unwind. Three tools connect from here. Apify scrapes Maps, Firecrawl reads the websites, Lovable builds the new ones.
2. Scrape the shops that have nothing
Connect Apify and aim it at one city and one trade. It returns the name, phone, rating, reviews and website field for every listing in range. Keep the rows where the website field sits empty and the rating clears 4.5. You now hold a list of shops that already make money and earn trust, with nothing online to find them by. Apify gives you 1,100 listings free, so your first 300 prospects cost nothing. Run a 10-row test, check the columns, then scale to hundreds.
3. Pull the proof off their listing
Apify hands you the listing, not the words that sell it. Connect Firecrawl and let it read each shop's reviews straight from Google. "Excellent customer service." "A great job at reasonable prices." A 5.0 from 18 neighbors. The owner's own customers wrote your headlines, and lifting them costs nothing. Firecrawl reads 500 sites a month on the free tier.
4. Build the site before you speak
Drop the business name and its top reviews into Lovable. 9 minutes later you have a live one-page site: the best review as the headline, the job count, one call button, the trade laid out like a studio built it. You wrote none of the copy and met none of the owners. The page is live before the shop knows your name.
5. Sell the thing they can already see
No deck, no script. You send one line and the link to the finished page. A shop owner looking at his own name on a real homepage, his real reviews as headlines, does not haggle over $500. Route 30 Smoke Shop said yes in a day. Four of those at $899 is $3,596, the same daily figure the phone gave him, now built while he slept.
6. Run the search on a schedule
Set the scrape as a task that fires at 8 a.m. Every morning it pulls fresh no-website shops in your city and writes them to leads.xlsx, then stops. You open the sheet to a list your automation built overnight, then ask Cowork for a one-screen dashboard tracking shops found, sites built, and replies in. You pick from prospects instead of chasing strangers.
7. Run the cost math once
Under 300 leads a month, both tools stay free. Past that, Apify bills per use and Firecrawl's paid tier runs near $125 for thousands of pages. At full volume the rig of Apify, Firecrawl and a Claude plan stays under $200 a month. Sell three sites at $500 and you cover the month ten times over.
8. The progression is slow on purpose
Month 1: he built six sites by hand and made $3,200. Month 2: the 8 a.m. task surfaced 280 no-website shops, he built 19 and closed 11. Month 3: $26,957 on the same rig at the same under-$200 cost. No viral spike moved this. The prospect sees the product before the price, and the close rate holds.
Scrape a city first and you own it. Every shop on that map without a website is unbilled inventory.
Thanks for reading to the end. If the breakdown earned its place, repost the first move so the right person reads it before their competitor does, and follow for the next build.
