827 seconds.
That's how long my program took on a single search for sports. Then 708 seconds on the second, then 702 on the third.
No error, no message. Just a program reading ordinary pages and never getting out of them.
Maybe you've been through something like this too: a job that looks simple from the outside, then you find a maze inside.
1. The email that doesn't want to be read
From a distance, pulling an email address out of a web page is one line of code: look for something with an @ in it. That's how I started.
Then target customers' websites started teaching me.
- A page with a long block of JavaScript with no @ in it at all, so the search crawls through it character by character and starts over from every character. That's where the 827 seconds came from.
- A site that hides its email behind Cloudflare protection, so only symbols show up in the page.
- A riad that writes info [at] riad [dot] ma so programs can't catch it.
- A phone number stuck to the email: +212 5 24 38 99 00contact@riad.ma, which turns the email into 00contact, an address that doesn't exist.
- A French word with an accented letter, réservation@, which the program read as servation@.
- Example addresses in contact forms: votre@gmail.com and jean.dupont@gmail.com. The program found them in a real test and counted them as customers.
Every line in this list came from a real experience: a discovery, then a fix, then a test to make sure the mistake doesn't come back.
And nobody sees this part. The user sees a table of clean emails, and doesn't see how many wrong addresses were removed before it reached them.
2. Why I'm giving it away
The full Lead Miner program looks for target customers itself: you type the field and the city, and it finds the sites, extracts their emails and learns from your ratings.
But many of you already have a list of sites: from an old spreadsheet, from a Google Maps search, or from a trade directory. What's missing is the next step: opening each site, looking for the "contact us" page, and copying the email.
So today I took the extraction engine itself out of Lead Miner, with all its fixes, and put it in a small free tool.
3. What Lead Miner Free does
- You paste the links: one site per line, up to 50 sites at a time.
- It reads each site and then its important pages: "contact us", "legal notices" and "about us", inside the same site only.
- It respects robots.txt: if the site forbids reading a page, it doesn't read it.
- It drops what's useless: noreply@, example addresses, and image names that look like emails.
- It exports the result: a CSV file that opens in Excel in Arabic and French without broken symbols.
Before writing this post I tried it on four real sites. Two of them gave 14 emails, and the tool read three pages on each. The third site refused the visit, and the fourth doesn't exist, and each showed up in the table with its reason instead of stopping the run.
4. Its limits, honestly
- You're the one who brings the list of sites. The tool doesn't search for them.
- 50 sites at a time.
- It runs on Windows and needs Python 3.11 or newer.
- Some sites refuse programs, and show up in the table as "unreachable".
⬇️ Free download
Unzip, then double-click 1-install.bat once, then 2-start.bat, and the browser opens on the tool.
💳 Price and purchase
Want the program to find the sites itself, by field and city?
Lead Miner: $99 / a full year license
Question before buying? Message me on WhatsApp.
The part nobody sees is the part that makes the difference.
And you, how do you collect your target customers' emails today? By hand, or with a tool? Tell me in the comments.
A new free tool reaches subscribers of the newsletter first every month.
Comments
Comments
Post a Comment