Behind the Paper

The use of e-commerce applications in Latin America: individual and structural influences during COVID-19

How did individual behavior and market structures shape e-commerce adoption in Latin America during COVID-19? This study reveals how trust, social media, and economic infrastructure drove digital shopping across 18 countries.
Caroline Xavier Valdivino Nov 23, 2025

E-commerce Adoption in Latin America During COVID-19: Individual Behavior, Structural Conditions, and What Happens Behind the Research

Introduction

The COVID-19 pandemic triggered one of the fastest and most profound digital transformations ever recorded, particularly in emerging economies such as those of Latin America. While e-commerce adoption surged almost overnight, the mechanisms driving this shift were far from uniform. My research explored a central question: Why did some people adopt e-commerce apps much more readily than others, and how did national market conditions shape this behavior?

This blog post expands on the scientific article I recently published, offering not only a summary of the findings but also the behind-the-scenes elements of the project — how the study was conceived, why these questions matter to me personally, and how they can benefit members of this Community.

Why This Study? A Personal and Regional Motivation

Growing up and working in Latin America, I have always witnessed the stark inequalities that shape access to technology. When COVID-19 hit, these gaps became painfully visible — people who relied entirely on face-to-face transactions suddenly faced lockdowns and mobility restrictions. Yet others shifted seamlessly into digital purchasing.

I wanted to understand:

  • What explains these differences?

  • Are they purely individual — trust, digital habits, social media engagement?

  • Or are they structural — infrastructure, purchasing power, logistics, connectivity?

  • And how much does each level actually matter?

This research became, in many ways, a way to make sense of what I was seeing around me: families struggling with digital divides, small businesses forced online, and entire markets accelerating into digitalization unevenly.

Behind the Research: How the Study Was Built

To answer these questions, I used two large datasets:

1. Latinobarómetro 2020 (individual-level data)

A long-standing public opinion database with responses from 18 Latin American countries. It provided:

  • interpersonal trust

  • social network usage (Facebook, Instagram, WhatsApp)

  • self-reported social class

  • age and gender

  • and most importantly, which e-commerce apps people used (Mercado Libre, Amazon, food delivery apps)

After cleaning missing values, the final dataset included 12,309 respondents.

2. Market Potential Index (country-level data)

This database captures national-level market characteristics, such as:

  • market size

  • market intensity

  • consumer capacity

  • commercial infrastructure (logistics, internet speed, broadband, paved roads, mobile subscribers, etc.)

This allowed us to pair each individual with the structural profile of their country.

Why Use a Multilevel Model? The Hidden Challenge

One of the biggest methodological challenges was the fact that individuals are nested within countries. This means that:

  • People from the same country share economic conditions, cultural patterns, and digital infrastructure.

  • This violates the independence assumption required in traditional regression models.

To solve this, we used a multilevel regression model with random intercepts. This approach lets us:

  • examine individual differences and structural differences simultaneously

  • quantify how much variation in behavior is due to personal factors versus country factors

  • avoid biased conclusions that might blame individuals for behaviors shaped by structural constraints

An interesting behind-the-scenes moment:

When we ran the Null Model, we discovered that 12% of all variation in e-commerce use comes strictly from country-level differences — a remarkably high number for behavioral research.

Key Findings: What Really Drives E-commerce Use?

1. Interpersonal trust matters — a lot

Trust reduces perceived risk. People who trust others are more willing to buy online, share information, and try digital payments.
This aligns with Latin America’s context, where institutional trust is often low, making interpersonal trust even more valuable.

2. Social media usage is the strongest predictor

Individuals who used more social networks (especially WhatsApp, Instagram, and Facebook) were far more likely to adopt e-commerce apps.
Social networks serve as:

  • marketplaces

  • channels for recommendations

  • spaces where risk is reduced through social proof

During lockdowns, this became even more evident as small sellers and large platforms converged on social media.

3. Structural factors matter — but less than expected

Three structural variables had positive, but modest effects:

  • market size

  • consumer capacity (purchasing power)

  • commercial infrastructure

These factors help determine access to digital platforms (internet quality, logistics, payments).

4. Market intensity had no significant effect

Contrary to expectations, per capita income and private consumption did not explain differences in e-commerce use.
This suggests that adoption is not simply a matter of national wealth — digital readiness, trust, and social habits play a larger role.

Interpreting These Results: What They Mean for the Community

For researchers, practitioners, and policymakers in this Community, the study provides several insights:

1. Digital adoption is both social and technological

E-commerce adoption is not merely a product of “better technology” but rather:

  • how people perceive trust

  • how they use social networks

  • how countries structure their commercial ecosystems

This supports the UTAUT model’s emphasis on social influence and facilitating conditions.

2. Infrastructure still sets the boundaries

Even if individuals are willing, adoption can only flourish where logistics, connectivity, and payment systems are available.
This is especially relevant for members working on digital transformation strategies.

3. Social media is no longer auxiliary — it is central

For businesses and digital innovators, social media is not just marketing.
It is a gateway to e-commerce adoption, especially in emerging markets where formal digital ecosystems are uneven.

Why This Study Matters to Me — and to Our Community

This research sits at the intersection of economics, technology, and social behavior, a combination I find deeply meaningful.

For me personally, the study reflects a desire to understand how digitalization can either empower or exclude people in regions marked by structural inequality.

For the Community, I hope this research contributes by:

  • encouraging discussions on inclusive digital transformation

  • offering a methodological reference for multilevel analysis

  • and fostering debate on how trust, infrastructure, and social behavior shape technology adoption

Conclusion

The adoption of e-commerce applications in Latin America during COVID-19 was not random. It resulted from a complex interplay of personal trust, social media engagement, and national economic and technological conditions.

By combining individual and structural data, this study demonstrates that digital adoption in emerging markets requires both human and infrastructural readiness.

I look forward to hearing your thoughts, questions, and perspectives, and engaging further with this Community to deepen these discussions.