B2B expansion into Latin America: how to pick the first country and find the buyers

Step 1

Which country should you enter first?

Enter first the country where you can name the most reachable buyers at a size you sell to, before you spend on outreach. Public registries let you count them. Market size, tariffs and headlines tell you little about who will answer.

Most expansion plans start from a macro story: nearshoring, a trade agreement, a fast-growing economy. Those stories are real inputs, but they describe a country, and you do not sell to a country. You sell to a purchasing manager at a distributor in an industrial corridor. The first question is therefore not "which market is biggest" but "in which market can I list the companies that could buy from me, with a name, a location and an activity code?"

That list exists in most Latin American countries because companies must register to operate. What differs is how much of the register the state publishes, in what format, and how current it is. We score each candidate country on six criteria before anyone writes a single email.

Two of the six criteria, reachable universe and concentration, come straight from public data and can be computed in a day. The other four need judgment, and that is where your sales leadership earns its keep. The scorecard exists to stop the decision from being made by whoever spoke last in the meeting.

Step 2

Where can you find buyers in each market?

In each country's public business registry or establishment directory. Colombia (RUES), Mexico (DENUE), Chile (SII), Peru (SUNAT) and Brazil (Receita Federal CNPJ) each publish a different slice, so the first job is learning what each one contains and what it leaves out.

No registry is a customer list. Each is a legal or statistical record, built for tax or census purposes, that happens to be public. Treat it as the denominator: it tells you how many companies could exist in your segment and where. We checked each official page on October 13, 2026; where a page could not be read, the table says so.

The pattern matters more than any single row. Registries differ in what they let you filter by, such as company size, location, legal form or activity, and those fields change between editions. Before you build a country list, open the registry and check which fields its current file actually carries. Your first-country scorecard should reflect those differences, because a filter you cannot apply is a list you cannot trust.

Worked example

What does a registry count look like in practice?

For Colombian food companies, an October 8, 2026 query of a RUES extract loaded July 23, 2026 returned 12,713 commercial registrations with main activity code CIIU 10. Of those, 4,265 are registered in Bogotá and 8,448 elsewhere in the country. They are registrations, not plants.

The distinction is the whole lesson. A registration is a legal record. This count already excludes natural persons, inactive registrations, those not renewed since 2024 and those in liquidation, and a registration can still be a company with no real operation or a large processor. Counting registrations gives you the outer edge of the market, never the number of buyers. Every filter you add after this point, legal form, status, location, size where available, shrinks the number toward the companies worth a conversation.

Source: DealTunnel query of RUES records (rues.org.co), run October 8, 2026, grouped by main CIIU code 10 and by city scope. The shares are our arithmetic on those counts. Registrations only; no claim about operating plants or active companies.

Notice what this tells a US distributor of food-processing equipment. About one in three registrations sits in a single metro area, which is a concentration you can visit in a week. It does not tell you how many have more than ten employees, because the RUES activity code does not carry size. That gap is why you pair registries with a second source or a screening call before spending on outreach.

Allies

Where do trade shows fit in a Latin American entry?

Trade shows are allies. They put you in front of buyers who walked past your booth. A registry map adds the ones who did not attend, so the show becomes one input to a market map instead of the whole plan.

A booth in Bogotá or Mexico City gives you a few hundred conversations, many of them real. It does not give you the purchasing manager at a plant three states away who never bought a ticket.

“The show introduces the buyers who walked by your booth. We map the ones who didn’t.”

Nicolás Calderón, Founder, DealTunnel

The practical use is sequencing. Before the show, map the registered companies in your segment and mark which exhibitors and attendee lists overlap with your map. At the show, spend your time on the overlap and on the ones you know matter. After the show, contact the rest of the map with a reference to the event you both know exists. A calendar of one show per quarter in the target country is enough to pace the first year.

Language

Tuteo or usted, Spanish or Portuguese?

Write in the buyer's language and match the register their own company uses in public. Brazil is Portuguese, not Spanish. In Spanish markets, tuteo and usted both occur in B2B; norms differ by country, sector and seniority, so test with a native reviewer.

This section is our editorial practice, not a researched survey, and you should treat it as such. We have not found a primary source that ranks countries by formality, so we do not claim one. What we do is read the buyer's own website, LinkedIn posts and press mentions: if the company speaks to customers in tuteo, a first message in tuteo reads as native; if it writes formally, usted costs you nothing. DealTunnel writes in tuteo in its own Spanish-language work.

Two rules hold regardless. First, never send machine-translated English; buyers read it as a form letter. Second, keep the first message short and specific to the company, with one concrete observation drawn from public data, such as its registered activity or location. A native-speaking reviewer for every new country is the cheapest control you can add.

Compliance basics

What data protection questions come first?

Before contacting anyone, ask four questions per country: what personal data you collect, from which public source, how the person can opt out, and who stores it. This is not legal advice. A local lawyer must confirm the rules for each country.

Every country in this guide has its own personal data law, and several have changed recently. We have not verified the current text or enforcement of each law against primary sources, so we do not summarize them here. Doing so from memory would be exactly the kind of unsourced claim this guide avoids.

1. List which fields you hold on each contact: name, role, email, phone.
2. Record the public source and date for each field.
3. Give every message a clear sender identity and a working way to opt out.
4. Decide where the data is stored and who can access it.
5. Ask a local data protection lawyer to review the process before the first send.

Company-level registry data (a business name, an activity code, a tax address) is a different category from a named person's email address. Keeping the two apart in your files makes the legal review faster.

Method

How do you go from market map to first meetings?

Run seven steps in order: scorecard, registry pull, filters, buyer roles, language review, a small first wave, and a review of every reply. Skipping from the map straight to volume is the usual way to waste a first quarter.

1. Score the candidate countries with Table 1 and pick one. Write down why, so the decision can be checked later.
2. Pull the registry for your activity codes. Record the date and the exact filters, so the count can be reproduced.
3. Filter down by legal form, status, location and, where the registry allows, size. Keep a note of how many each filter removed.
4. Define the buying roles you want in each company: who signs, who specifies, who uses.
5. Review language and register with a native speaker, using real examples from the target companies' own communications.
6. Send a small first wave to a few dozen companies, spread across at least two cities, so one local factor does not decide the result.
7. Read every reply yourself in the first weeks, including the refusals. They tell you which filter was wrong.

The Revenue Audit follows this logic: we build the market map from public records, name the segments and the buyers, and show you what you could not see in your own lists. It arrives within 72 hours in a private hub, and you can read how the larger system works in our Revenue Engineering overview, which describes the 16 steps from the public registry to the signed contract.

First 90 days

How do you tell signal from noise in the first 90 days?

Signal is a pattern repeated by several different buyers. Noise is a single loud reply, a vanity metric, or a result from one city. Wait for the same observation from at least three separate conversations before you change anything.

Our working rule: one conversation is an anecdote, three are a signal worth testing, and five are a rule worth building on. This is a rule of thumb from our own practice, not a statistical threshold. With small samples, say so openly and avoid presenting a percentage from twenty companies as a market truth.

Report every figure with its denominator: replies out of how many contacted, meetings out of how many replies. A number without its denominator is the first thing to distrust. We do not promise outcomes or timelines; the useful output of ninety days is a clearer map, not a guaranteed pipeline.

Topical map

What else does this guide cover?

This page is the hub for B2B expansion into Latin America. Deeper articles are planned on each subtopic below. None is published yet, so none is linked; this list will turn into links as each article goes live.

• How to choose your first Latin American country with public data. Planned.
• A guide to Latin American business registries. Planned.
• Mapping buyers in Colombia with RUES. Planned.
• Mapping buyers in Mexico with DENUE. Planned.
• Trade shows in Latin America: following up the buyers who never visited. Planned.
• B2B outreach in Spanish and Portuguese: register and tone. Planned.
• Your first 90 days in a new Latin American market: signal versus noise. Planned.

Related reading: B2B market intelligence from public records covers how we build the underlying market map.

FAQ

Frequently asked questions

Sources

Sources

All pages checked October 13, 2026 unless noted.

1. INEGI, DENUE. inegi.org.mx/app/mapa/denue. Quote: "se ofrecen los datos de más de 6 millones de establecimientos que hay en nuestro país", edition based on the Censos Económicos 2024. No update date on the page.
2. SII Chile, Personas Jurídicas y Empresas. sii.cl. Quote: "Listado de todos los contribuyentes personas jurídicas catalogados como empresas por el SII"; "Cifras preliminares"; companies nomina marked "Actualización: noviembre 2025".
3. SUNAT, Padrón reducido del RUC, para descarga. orientacion.sunat.gob.pe. Quote: "Es el Padrón del Registro Único de Contribuyentes, pero con información reducida." Update date not on the page.
4. RUES, Registro Único Empresarial y Social. rues.org.co. Portal text not readable on our check; role described by the Superintendencia de Sociedades (Decreto Ley 019 de 2012) in a draft circular.
5. Receita Federal, CNPJ open data. dadosabertos.rfb.gov.br/CNPJ. Server did not respond on October 13, 2026; contents not verified by us.
6. Colombia RUES counts by CIIU 10: DealTunnel query, run October 8, 2026, of a RUES extract loaded July 23, 2026 with registrations through July 3, 2026 (12,713 registrations; 4,265 Bogotá; 8,448 rest of country).

Author: Nicolás Calderón, founder of DealTunnel. Questions: nicolas@dealtunnel.io.