Sold in the DM: What the Evidence Actually Shows About Latin America's Conversational Commerce Edge

The number that doesn't hold up

You have probably seen the claim, sometimes with a specific-sounding range attached: conversational commerce in Latin America converts at 45 to 60 percent, roughly twelve times what a normal online store manages. It shows up in pitch decks, in blog posts about selling into Mexico or Brazil, in more than one place with almost identical wording. That last detail is the tell. When the exact same figure, phrased almost the exact same way, appears across multiple unrelated marketing sites with no footnote, no linked study, and no named researcher behind it, the read is not that the number is well established. It is that the number has been repeated enough times to feel established, which is a different thing entirely.

We went looking for the study underneath it. We could not find one. Not from Meta, not from AMVO, not from any of the regional ecommerce bodies that would plausibly have run this kind of comparison. The figure fails the most basic test a claim has to pass before it ships under our name: can you follow it back to a source you can actually read. This one cannot be followed anywhere. So the version of the question this study set out to answer is narrower than the popular one. Not 'how much better does chat convert,' but 'what does the verified evidence actually support about conversation as a commerce surface in Latin America, and where does real measurement simply not exist yet.'

The exact same figure, phrased almost the exact same way, showing up across unrelated sites with no source behind it, is not evidence of consensus. It is evidence of repetition.

What we chose not to publish

Showing the work that got cut matters as much as the work that survived. Beyond the headline 45-to-60-percent figure, three more specific claims came into this research looking credible and did not survive a direct check of their own cited sources. A widely repeated statistic, that 72 percent of Latin American consumers have purchased through a messaging app compared with 45 percent in Europe and 38 percent in North America, traced back to a single blog page that stated the number with no attribution at all.

Two more claims were attributed specifically to AMVO's 2026 Estudio de Venta Online, the kind of named, credible trade-body source that should be trustworthy on its face: that 64 percent of Mexican ecommerce transactions are initiated or closed through WhatsApp, and that 73 percent of buyers value pre-purchase questions while 81 percent feel more confident after a conversation. When we pulled the actual trade-press article these figures were supposedly drawn from, the 64 percent number did not appear anywhere in it. The 73 and 81 percent figures did not appear either. What did appear nearby, attributed to a different research group called Concepto Movil rather than to AMVO, was a related but distinct statistic. Publishing any of those three numbers under AMVO's name would have misrepresented what AMVO actually found. That is a real error, not an unverifiable gray area, and it is exactly the kind of thing worth catching before a client makes a market-entry decision based on it.

What Meta's own numbers actually say

Strip away the uncited claims and what remains from Meta's own disclosures is still a meaningful story. In its fourth-quarter 2025 earnings call, Meta reported that WhatsApp business messaging, the API and business-messaging revenue folded into its broader Family of Apps other revenue line, crossed a two billion dollar annual run rate. That wider revenue category grew 54 percent year over year, and in the United States specifically, click-to-message ad revenue, ads designed to start exactly the kind of chat thread this study is about, grew more than 50 percent year over year.

More specific to Latin America, Meta disclosed more than one million weekly consumer-to-business AI conversations across its messaging platforms, and named Mexico, alongside the Philippines, as an early-traction market for that business-AI feature. None of this proves a 45 percent conversion rate, or any conversion rate at all; Meta did not disclose one. But it proves something arguably more useful to a business owner deciding where to invest: a real, growing, multibillion-dollar platform business, with Mexico specifically called out by Meta's own leadership as a market where businesses are already using AI-assisted conversation at meaningful scale.

A region genuinely outgrowing the world

Separate from the messaging-specific data, the broader regional trend is well established and, on its own, striking. eMarketer's 2025 forecast puts Latin America's retail ecommerce sales at 191.25 billion dollars, up 12.2 percent year over year, roughly 1.5 times the global average growth rate. That marks the first time since 2021 that Latin America has reclaimed its position as the world's fastest-growing ecommerce region, ahead of Asia-Pacific and North America.

Mexico is the sharpest edge of that trend. eMarketer projects Mexican ecommerce penetration will reach 17.7 percent of total retail sales in 2026, edging past the United States at 17.0 percent for the first time. That is not a conversational-commerce statistic specifically, it is a whole-of-market growth statistic, but it sets the frame for why the conversation-versus-cart question matters commercially in the first place. A region growing this fast, with this much retail spend still shifting online for the first time, is exactly the kind of market where the channel a business bets on early can compound for years.

Brazil and Mexico are not the same market

It is tempting to talk about 'Latin America' as one commerce environment, but the DataReportal and Kepios figures make clear that the two largest markets look meaningfully different on the surface most relevant to DM-based selling. In Brazil, Instagram reached 147 million users as of October 2025, equal to 69.0 percent of the country's total population, a near-universal surface among Brazil's connected population. In Mexico, Instagram reached 53.6 million users in the same period, 40.5 percent of total population, markedly lower penetration than Brazil.

That gap matters for anyone building a channel strategy off a single regional assumption. A DM-first strategy built around Instagram reach in Brazil is starting from a very different baseline than the same strategy in Mexico. Nothing in the verified evidence here breaks WhatsApp usage down by country with the same rigor, which is itself a real gap worth naming: the platform doing the most measurable commerce work, WhatsApp, is also the one where clean country-level penetration data proved hardest to pin to a single verified primary source in this research.

The behavioral case for conversation, with a caveat

There is real behavioral evidence that talking to a business before buying changes outcomes, and it comes from a source worth taking seriously on its own terms: a Meta-commissioned, Kantar-executed survey of 11,056 online adults across 22 markets, fielded in two waves in April and September 2025. It found that respondents were 72.4 percent more likely to purchase from a brand they could message, 73.3 percent preferred messaging as a way to contact a business, and 74.6 percent said they trusted a business more when they could message it.

The caveat matters as much as the finding. This is a global 22-market study, not a Latin-America-specific one, and it was commissioned by Meta, the company whose messaging product it validates. That does not make it worthless, the methodology is disclosed and the survey design is legitimate, which is why it is tiered as emerging rather than dismissed outright. But it should never be presented as evidence about Latin American buyers specifically. It is evidence that conversation, generally, changes purchase intent. Whether that effect is larger, smaller, or the same size in Mexico and Brazil specifically is not something this survey, or any other source found in this research, actually measures.

It is evidence that conversation changes purchase intent, generally. Whether that effect is larger in Mexico and Brazil specifically is not something any source found in this research actually measures.

The digitization paradox

Here is the counterweight to that picture. AMVO data, reported through Mexican trade press and independently confirmed word for word against the source text, found that fewer than 15 percent of Mexican small and midsize businesses have a fully digitized, end-to-end sales process. That is a striking gap next to how visible chat-based selling appears across the region. A business can look conversation-first from the outside, active WhatsApp threads, DM replies, order confirmations sent by hand, while having no CRM, no backend order system, and no way to actually track what share of those conversations become sales.

That paradox is worth sitting with, because it also explains why a clean conversion-rate comparison is so hard to find. If most of the businesses doing this selling are not capturing structured data on it, no measurement of a 'chat conversion rate' comparable to a website's session-to-purchase rate can exist yet at scale. Related infrastructure has been attempted: MercadoLibre's payments arm, Mercado Pago, was reported in testing as a payment processor for an in-WhatsApp checkout tool being built with Meta in Brazil, but that report is from December 2022, three years stale relative to this study, and no more recent confirmation of live status was found. The building blocks for measurable, trackable conversational commerce exist. Whether they have actually been assembled at scale is a separate, unanswered question.

The rising rival and the unmeasured frontier

Two forces sit at the edge of this picture, both real, both still small relative to WhatsApp's scale, and both worth watching rather than ignoring. TikTok Shop has been testing Mexico as a regional market since late 2023, and its Brazil launch in May 2025 is estimated, in trade and analytics press rather than TikTok's own disclosure, to have grown from roughly one million to 46.1 million dollars in monthly gross merchandise value by August 2025. That figure should be read as directional, not precise, but the direction itself is unmistakable: livestream and video-native commerce is establishing a real foothold in the same markets where chat commerce has been building for years.

The second, less visible force is the AI-answer layer. Spanish and Portuguese rank among ChatGPT's leading non-English languages, according to secondary reporting on OpenAI's usage data, which suggests meaningful reach into Latin America's two largest language markets already exists. But no independent source found in this research measures what share of Latin American product discovery or purchase consideration actually happens inside the answer an AI engine gives, as distinct from a human WhatsApp chat or a general chatbot interaction. That is not a small gap. It is the edge of what anyone, anywhere, currently knows about how buyers in this region are starting their purchase journeys.

Where this leaves your attention map

Put together, the verified record supports a real story, just a more careful one than the viral version. WhatsApp business messaging is a genuine, fast-growing, multibillion-dollar channel with specific early traction named in Mexico. Latin America is genuinely the world's fastest-growing ecommerce region. Conversation, in general, measurably raises purchase intent and trust. And underneath all of it sits a digitization gap large enough that most of the businesses selling through chat in Mexico cannot yet prove, with their own data, what that channel is actually worth to them.

That combination, real momentum plus real measurement gaps, is exactly the situation the Visibility Corpus exists to map. Not a borrowed industry statistic standing in for your specific market, but an actual read of where your buyers' attention sits today across WhatsApp, Instagram, search, and the AI-answer layer, and how that mix is shifting under the region's fastest ecommerce growth in five years. The businesses that win the next phase of Latin American commerce will not be the ones repeating the twelve-times number. They will be the ones who found out, for their own category and their own two markets, what is actually true.