How to build a B2B prospect list from public data (step by step)
Every sale starts the same way: someone opens a blank spreadsheet and needs names. The good news is that across LATAM and Spain most of the businesses you want to win have already published themselves on public maps and directories precisely to be found. This playbook takes you from your ideal profile to a ready-to-call list, using the exact process the best-converting teams run.
1. Define your ideal customer before searching anything
The most expensive mistake in prospecting is not bad data: it is searching without a profile. Before opening any tool, write down in one line which business you are buying: category (dentists, accounting firms, workshops, logistics), city or metro area, approximate size, and which of your problems it has. With that line, everything that follows becomes measurable.
A quick test: if you cannot say in one sentence “I am looking for X in Y that probably has Z”, stop and fix it. “Dental clinics in Querétaro that still lack online booking” is a search; “more clients” is a complaint. The difference between the two decides whether your list becomes pipeline or obituary.
2. Use the sources where your customer already published itself
Local businesses invest years in their public map listing: phone, address, hours, reviews. It is the richest and cheapest prospecting source that exists, because the business itself published it with the declared goal of being contacted. Add guild directories, chambers of commerce and the “about” pages of their websites.
The golden rule is to collect only public and commercial fields: name, category, address, contact phone, website and reviews. That is exactly the scope RadarLeads automates — and that our Public data policy documents — so your team prospects with a clean conscience and verifiable provenance on every record.
3. Clean and normalize: 80% of the work
A raw file always arrives dirty: phones in three different formats, duplicates from franchises, addresses missing the city. Before calling anyone, normalize phones to international format, deduplicate by name and address, and split records without a phone into an email flow instead of blind calls.
This cleaning is mechanical, which is why it should be automatic. RadarLeads runs it on every search — deduplication, phone normalization and consistent columns — which is why an exported file imports into your CRM on the first try instead of dying in a spreadsheet tab nobody opens again.
4. Prioritize: not every row deserves the same call
With the list clean, order it by opportunity signals, not alphabetically. Recent reviews complaining that “nobody answers” are gold: there is unserved demand. A business with 4.8 stars but no website needs exactly what you sell if your product is a website. A business with 200 reviews has volume; one with 12, maybe a small budget.
Split the list into three blocks A/B/C and attack block A this week. Prospecting is a sequencing game, not a volume game: twenty well-chosen calls to businesses with the right signal outperform two hundred calls in alphabetical order. Your list is a map with treasures marked, not a phone book.
5. Reach out with context and measure everything
The first touch is built from what you already know: “I saw your shop in Polanco has 180 reviews and most of them praise your speed…”. That sentence, which costs ten seconds of reading the listing, multiplies reply rates because it proves you are not calling a random bought list. No full script needed: context, one-sentence offer and one question.
Log the outcome of every touch in the same lead row: contacted, answered, follow-up, closed. Within two weeks you will know which block (A, B or C) converts best and which city replies more, and your next search gets designed with your own data instead of gut feeling.
Turn these playbooks into a real pipeline
Run your first search free: 15 leads, 3 searches, no credit card required.
Start free7-day trial · 3 searches included