Team Analysis: Mercado Libre
estudiante
MercadoLibre.pdf
CASE: IB-105
DATE: 04/04/13
Gary Mekikian prepared this case under the supervision of Professor William Barnett as the basis for class
discussion rather than to illustrate either effective or ineffective handling of an administrative situation. Parts of this
case were excerpted from Stanford GSB case IB-19A, MercadoLibre.com, written by Andrea Higuera and Lauren
Pressman under the supervision of Professor William Barnett.
Copyright © 2013 by the Board of Trustees of the Leland Stanford Junior University. Publically available cases are
distributed through Harvard Business Publishing at hbsp.harvard.edu and European Case Clearing House at
ecch.com, please contact them to order copies and request permission to reproduce materials. No part of this
publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by
any means –– electronic, mechanical, photocopying, recording, or otherwise –– without the permission of the
Stanford Graduate School of Business. Every effort has been made to respect copyright and to contact copyright
holders as appropriate. If you are a copyright holder and have concerns, please contact the Case Writing Office at
[email protected] or write to Case Writing Office, Stanford Graduate School of Business, Knight Management
Center, 655 Knight Way, Stanford University, Stanford, CA 94305-5015.
MERCADOLIBRE
The company was doing great, but with growth and success came tremendous challenges. I sat
together with my manager of customer service, and answered calls in our call center to learn first
hand what our challenges are.
—Marcos Galperin, CEO and Chairman, MercadoLibre
As 2013 approached, Marcos Galperin and his team of MercadoLibre top executives were
meeting to celebrate the breathtaking growth of their company, and contemplate the challenges
ahead. Since starting the company with his Stanford Business School classmates in 1999,
Marcos had transformed the company from an internet auction site akin to eBay, to Latin
America’s leading online marketplace unique in its own right, and on par with Amazon. The
MercadoLibre team had not just cloned eBay and Amazon in Latin America; it had made a
number of key innovations in order to compensate for, and in some cases, take advantage of the
lack of commerce infrastructure in its markets. On the technology side, MercadoLibre had
started with a then state-of-the-art internet application based on an industrial-grade stack of
Oracle technologies. The application handled millions of transactions, provided a good user
experience, and facilitated the company’s growth and IPO in 2007. In 2008, Galperin had made
what he described as a “bet the company” decision to completely replace the company’s
technology by building and deploying a parallel system; not just another application, but a
sophisticated e-commerce platform based on Web 2.0 and the latest mobile web standards. The
gamble was paying off. The company’s transition was being hailed as another success for
Galperin and his team. But there was a larger challenge looming on the horizon.
In 2013, Galperin and his team watched as the biggest e-commerce company in the world,
Amazon, entered the markets that had been hitherto dominated by his company. MercadoLibre
had built a powerful brand and a significant footprint in Latin America through localization,
customization, and adaptation of best ecommerce practices. But competing with Amazon, with
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 2
its huge stockpile of cash and sterling reputation worldwide, presented significant strategic
challenges.
FOUNDING MERCADO LIBRE, 1999
In the spring of 1999, Marcos Galperin, then an MBA student at the Stanford Graduate School of
Business, had a vision: to build an e-commerce company focused on serving the nascent but fast-
growing Spanish and Portuguese-speaking markets in Latin America. While many U.S.-based e-
commerce companies were beginning to think about international expansion, Marcos felt
strongly that a home-grown company built from the ground up to serve Latin American markets
could compete against the foreign transplants and win. Galperin’s goal was to identify a
business model that had proven successful in the United States and could be easily transferred
and adapted to Latin America (see Appendix A). During the spring of 1999, he had been
thinking about selling cars or real estate over the Internet when he had an epiphany. Why not
sell everything? That day, the idea of an online auction site for Latin America was born.
Before his graduation in June 1999, Galperin had secured angel funding and recruited his cousin,
Marcelo Galperin, to be the company’s chief technology officer. After graduation, Marcos
Galperin returned to Argentina, and in the spirit of a Silicon Valley start-up, he and Marcelo
began working out of a small garage in Buenos Aires. They quickly recruited Hernan Kazah, one
of Marcos’ classmates from Stanford, and on August 2, 1999, Marcos’ vision became a reality.
MercadoLibre launched its first country-specific C2C and B2C online auction website in
Argentina.
MercadoLibre was well received by consumers in Latin America. Aggressive marketing and
promotions campaigns drove users to the site and transactions grew quickly. By October of
1999, less than three months after launch, MercadoLibre had over 15,000 users in Argentina who
had completed more than 6,000 transactions worth over $2 million in value; but MercadoLibre
was not the only company with a flag rushing to claim Latin American e-commerce markets.
Galperin and his team knew they needed to move quickly to play in the largest of Latin
American markets, Brazil, and move quickly they did. Shortly after entering the Argentine
market, Galperin launched a site in Brazil, and within a few weeks, more than 4,000 Brazilians
had registered as users on the site. With Brazil and Argentina showing great promise for the
upstart auction site, Galperin knew he had a winner, and pushed himself and his team to fulfill
their mission―“to build a Latin American company where users get together to socialize,
browse through items, and buy and sell almost anything in an efficient, easy, trusting and fun
environment.”
Regional Organization and Marketing with Local Feel
From the company’s early days, Galperin and his team thought about the organizational structure
that would best support their strategy of becoming a dominant player throughout the countries in
which they wanted to do business. It was obvious that a uniform technology platform could
handle the needs of all markets and scale across borders as the company grew. Not so obvious
were the localization and customization of its marketing efforts that would be required to project
an image of a local e-commerce company doing business according to local customs and needs.
It did not take long for Galperin to decide that he needed to hire local country managers, and to
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 3
build around each a local sales and marketing organization. This alone, he thought, would
ensure that every country’s needs would be well understood and met.
He also believed that developing a strong, well-recognized and unique brand perceived as cool,
irreverent and fun, would serve three key objectives: 1) to bring sellers and buyers to the site by
attracting current web users; 2) to position the brand in the minds of millions of people who
would become web users in the near future; and 3) to increase the loyalty of current users.
Galperin recognized that buyer and seller acquisition could be very time consuming and
expensive, if done in traditional ways. He spent significant effort evangelizing about his
company in the press, and when MercadoLibre launched in Argentina and Brazil, he worked
with a PR firm to make sure that the company received significant press coverage in local
newspapers and magazines. Even the U.S. press, including publications such as the Wall Street
Journal, the New York Times, Industry Standard and Red Herring, took notice of Galperin and
his company, which provided much-valued credibility in the eyes of internet-savvy Latin
Americans who follow the American press for the latest advances on the web.
Partnerships
Galperin developed partnerships with major Latin American portals to help acquire users. He
believed that portals would drive traffic to the site and help capture customers, and at the same
time, burnish the image of MercadoLibre as a major player on the Latin American internet scene,
both in the minds of the public and international investors.
Argentina’s Ciudad internet portal was MercadoLibre’s’s first partner, followed by StarMedia
and (an impressive feat for a start-up) AOL Brazil, an internationally recognized and prized
brand that dominated the Internet in the late 1990s. An important component of the company’s
partnership agreements was non-exclusivity. Anyone that partnered with MercadoLibre could
and did post MercadoLibre’s auction site as an option in their sites’ shopping channels. This
arrangement drove user acquisition while keeping costs to a minimum.
Technology
MercadoLibre believed that in order to succeed in the auction space, it was crucial to have
control of its technology. Initially, MercadoLibre used an “off-the-shelf” technology platform in
order get to market as quickly as possible. From the outset, however, the company hired an
internal IT team, led by Marcelo Galperin, to develop a robust, scalable, proprietary technology.
MercadoLibre believed that having a proprietary technology would be a key competitive
advantage for several reasons: 1) it would give flexibility to launch in any market without the
restrictions that most license agreements impose; 2) it would allow for continuous upgrades in
response to users’ feedback; 3) it would allow the company to offer customized solutions to its
business partners. To ensure reliability and scalability, the technology architecture was being
built on best-of-breed systems: Oracle database, Sun Microsystem servers, the Unix operating
system, and Exodus hosting. The decision to rely on big technology names and proprietary
systems and software to deploy its application seemed logical in the early days of the Internet.
Reliability and scalability were of paramount importance, and large technology companies such
as Oracle were safe bets.
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 4
Ecommerce Infrastructure
There were many reasons why Latin America was considered virgin territory for internet
companies at the turn of the century, not the least of which was lack of infrastructure on all
fronts. Low internet user penetration rates, high access costs, and inferior technology were
among the infrastructure challenges that threatened MercadoLibre’s ability to grow rapidly and
gain the necessary scale to exploit the benefits of the online auction model. Online commerce
relies heavily on physical infrastructure for delivery of goods once transactions are completed.
Historically, the official postal systems in Latin America had been unreliable, even though they
had improved dramatically in several countries, such as Brazil and Argentina. At the same time,
companies like Federal Express and DHL shipped goods internationally, but did not offer
affordable, domestic delivery services. To grow his company in spite of these seemingly
insurmountable obstacles, Galperin decided that he could innovate to address these issues. One
such innovation would be MercadoPago, an online payment system, which would become a
MercadoLibre crown jewel.
KeyCompetitor
Another start-up with Ivy League pedigree, DeRemate.com, opened its site for commerce in
Argentina only a few weeks after MercadoLibre’s launch. DeRamate.com was founded in July
of 1999 by 11 cofounders, some of whom were classmates from Harvard Business School, Yale
School of Management, and Kellogg School at Northwestern. The formidable cofounders, under
the leadership of CEO Alec Oxenford, invested $50,000 each in a seed round of financing, and
committed another $50,000 to be invested in the next round.1 In what would become
DeRamate.com’s hallmark for focusing on speed, Oxenford and his team decided that they
would purchase the auction site software, rather than build it, and rushed to market on August
31, 1999―a mere six weeks after their founding. They gave Aucland, a French firm running an
auction site in Europe, 10 percent equity in exchange for the right to use their software in Latin
America. Oxenford thought that he could quickly adapt the Aucland software platform to the
Latin American market because it was built to handle multi-currency transactions, and had been
tested in live sites for a number of years. He was half right.
The software was quickly translated to Spanish and deployed, and DeRemate.com went live on
schedule, with Argentinian television channels providing extensive coverage for the launch of
the site on shows devoted to teenagers. The launch was such a tremendous marketing success
and generated so much traffic that the DeRemate.com site promptly crashed.
While the team worked to stabilize the Argentinian site, it was hard at work identifying a suitable
strategy to enter the Brazilian market quickly. Since the Brazilian market already had a number
of e-commerce players, Oxenford and his team decided to trade more equity for speed to market
and growth. They convinced the Brazilian founder of E-Bazar, Mr. Andrade, to sell the site to
DeRemate.com for $50,000 plus 2.4 percent equity. Within two weeks, E-Bazar was rebranded
and re-launched as Arremate.com, and this was only the beginning.
1 Harvard Business School, “DeRemate.com: Building a Latin American Internet Auction Site,” January 24, 2002.
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 5
Meanwhile, cash was running low, so Oxenford embarked on a fundraising trip to New York and
Miami, hoping to raise enough cash to carry the company to IPO within a short time frame.
After being shunned by traditional VCs (most had no interest in Latin American markets),
Oxenford negotiated a $12 million Series-A financing at $15 million pre-money valuation from
Citigroup, Merrill Lynch, eQuest, TPG, and SLI.
With $12 million in the bank, Oxenford pushed forward with his expansion plans. Since the
founders of the company were from different Latin American countries, each founder opened a
local office, and coordinated with Oxenford to deploy the software and marketing in order to
start generating traffic and demand. Oxenford also tagged Sergio Grinbaum and Alejandra
Herrera, a former Boston Consulting Group administrator, to develop a formulaic approach to
opening offices. As rapid growth accelerated their cash burn rate, Oxenford convinced
Terra―one of the leading European internet portals owned by Spain’s Telephonica phone
company―to invest $45 million at post-money valuation of $150 million. For a brief moment, it
seemed Oxenford’s eight-month-old start-up was speeding to exit in the summer of 2000, with
underwriters JPMorgan and Merrill Lynch predicting a multi-billion dollar IPO. In April of
2000, a month after receiving the checks from Terra, the internet bubble burst, and the market
crashed, effectively foreclosing DeRemate.com’s IPO hopes in the near term.
Oxenford quickly regrouped and reduced its cash burn rate by 70 percent, but did not roll back
expansion plans. By August of 2000, DeRemate.com had around 1 million users, transacting
auctions in eight Latin American countries, plus the United States, where the company was
targeting the vast Spanish-speaking population. Oxenford’s goal was to reach profitability
without additional capital by 2003. But this was easier said than done. DeRemate’s founding
was based on the assumption that a quick IPO would provide liquidity to the 11 or so co-
founders who had left very lucrative positions to help build an internet start-up with a multi-
billion dollar exit opportunity.
When the tech bubble burst and the U.S. stock market crashed in 2000, the dream of a quick exit
through public offering or outright sale evaporated, and along with it the incentives of the
original team that was sprinting to an exit. Oxenford updated his business plan to target cash
flow breakeven by the middle of 2002. But reinventing the company around the new market
realities would be too big a challenge, leaving a huge opening for MercadoLibre to leap ahead,
and eventually acquire DeRemate.
GROWTH AND IPO OF MERCADOLIBRE, 2007
Meanwhile, by 2006, MercadoLibre had achieved what even Galperin could not have dreamed a
short five years prior. The company hosted the largest online trading platform in Latin America,
and had established itself as the market leader in e-commerce in Argentina, Brazil, Chile,
Colombia, Ecuador, Mexico, Peru, Uruguay and Venezuela, respectively, based on unique
visitors and page views measured in 2006. Additionally, Galperin had launched online trading
platforms in Costa Rica, the Dominican Republic, and Panama. With a market of over 550
million people and a region with one of the world’s lowest but fastest-growing internet
penetration rates, the company provided buyers and sellers a state-of-the-art online trading
environment that fostered the development of a large and growing e-commerce community.
Most importantly, the company offered a technological and commercial solution designed to
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 6
meet the distinctive cultural and geographic challenges of operating an online trading platform in
Latin America.
As of December of 2012, the company offered users two principal services:
The MercadoLibre marketplace: The MercadoLibre marketplace was a fully automated,
topically arranged and user-friendly online trading service. This service enabled both businesses
and individuals to list items and conduct their sales and purchases online in either a fixed-price
or auction-based format. Additionally, through online classified advertisements, the company’s
registered users could list and purchase motor vehicles, watercraft, aircraft, real estate and
services. Any internet user could browse through the various products and services that are
listed on the website and register with MercadoLibre to list, bid for, and purchase items and
services.
The MercadoPago online payments solution: To complement the MercadoLibre marketplace,
the company had developed MercadoPago, an integrated online payments solution akin to
PayPal in the United States. MercadoPago was designed to facilitate transactions on the
MercadoLibre marketplace by providing a mechanism that allowed the users to securely, easily
and quickly send and receive payments online.
MercadoPago went beyond PayPal’s service template by introducing key innovations. Credit
cards and bank accounts had only recently begun to penetrate the vast populations of Latin
American countries where the company operated. This created a unique challenge for
Galperin―inventing a process whereby buyers and sellers could exchange goods for cash or
check. MercadoPago solved this problem by creating an ingenious escrow service, interfacing
with a network of collection agents, who acted as payment goods intermediaries. When a cash
buyer completed a transaction online, she was instructed to make the payment to a local
collection agent. Once the payment was made and registered by the local agent in the system,
the seller would be notified to send goods to the buyer.
By 2006, MercadoLibre users paid approximately $87.9 million for items bought through
MercadoPago, which represented 8.3 percent of the gross merchandise volume for that year.
Buyers were charged an additional commission when paying via MercadoPago. The
commission rates averaged 8.0 percent of the sales price of a listed item, and varied depending
on whether a buyer made payments through MercadoPago by credit card, debit card, check, cash
or bank transfer. Argentina was one exception, where as of May 2006 the company charged a
flat 6.49 percent.
For cash, bank deposits, checks and electronic transfers, MercadoPago’s commission rates
ranged from 1.99 percent to 3.99 percent depending on the funding method and country, with a
minimum of approximately $1.00. For credit card payments, the commission rates ranged from
6.49 percent of the payment amount for a single lump-sum payment, up to approximately 43.99
percent of the payment amount (which included any applicable interest charged) for an 18-month
installment plan.
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 7
Galperin designed MercadoPago to make online commerce in Latin American markets more
efficient and more secure as compared to traditional offline payment methods such as checks,
money orders and credit cards via merchant accounts. Traditional offline payment methods
could place several obstacles in the online commerce process, including lengthy processing
times, general inconvenience, undesired disclosure of personal information, and risk of non-
delivery. With MercadoPago, individual sellers could offer a much broader array of payment
options to buyers without the need to have a credit card merchant account or collection accounts
with different banks.
The combination of MercadoLibre and MercadoPago contributed to the company’s dizzying
growth and market uptake, as evidenced by the impressive numbers it published in 2007. During
2006, visitors browsed an average of 2.5 million total listings per month, organized by country,
in 2,000-plus product categories. According to the company’s public offering prospectus (S-1),
as of December 31, 2006, MercadoLibre had 18.2 million registered users, 1.7 million unique
sellers, and 4.4 million unique buyers, with 13.8 million items sold.
Just prior to the IPO in 2007, MercadoLibre had achieved what eluded most internet start-
ups―profit. It had recorded $52 million in revenues and $1.1 million net income for the full
year in 2006. Galperin had achieved all this, while not only vanquishing its main competitor,
DeRemate, but also acquiring its assets for $40 million. Galperin’s singular focus on building a
world-class e-commerce company for the long term clearly worked to his advantage, as
compared to Oxenford’s goal of building a company for quick cash out. “From day zero, my co-
founders and I wanted to build a company that would survive us and thrive for generations.
Some of our competitors wanted to become rich and famous in the short term,” said Galperin,
then adding thoughtfully, “This is why we remain successful, and many of our competitors are
gone.” The company was on a roll.
2007-13: COMPANY IN TRANSITION
After 2007, Galperin and his team had precious little time to celebrate their IPO, as the new
challenges of operating a public company seemed to multiply with each passing month.
Maintaining a breakneck growth pace in both revenues and profits was critical to the viability
and success of MercadoLibre, and growing revenues meant growing users, growing transaction
volume and per transaction amounts, and finally growing services by edging into new markets.
At the founding of the company, Chief Technical Officer Marcelo Galperin had decided to build
an internet auction application using Oracle Corporation’s technologies, which proved to be
scalable and reliable, if not flexible and cost effective (see Appendix B). MercadoLibre had
grown to be a $50-million-a-year company using the same technology stack for eight years. But
as the company grew and expanded, so did the needs of its customers and the speed with which
changes to MercadoLibre’s systems needed to keep up with demand for myriad new features and
functions. “After the IPO, when I met with Stanford students and professors, or Wall Street
bankers, I was treated like a hero,” Galperin remembered during an interview in January of 2013.
“Then I would get in a cab and ask the driver to take me to MercadoLibre’s office. On the way
over, I would find out that the driver was one of my customers with a long list of complaints.
That’s when I decided to move my office from the executive suite to a cubicle next to the CTO’s
office, to learn the issues our technical team was dealing with first hand.”
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 8
Besides MercadoLibre’s growing pains, Galperin was also dealing with tectonic changes in e-
commerce brought on by introduction of Web 2.0 standards-based technologies. In the United
States, companies like eBay, Amazon, and Facebook had successfully transitioned from being
web applications to providing robust commerce and communication technology platforms. For
example, Amazon was deploying its strategy of enabling customers to buy from its competitors
if it did not carry the inventory for a particular item. Amazon did this by allowing small
businesses to build e-commerce sites using its flexible software and hardware platforms. A
seller of athletic footwear that otherwise competed with Amazon could develop a storefront on
Amazon’s sprawling e-commerce site, using Amazon’s computing resources, e-commerce
applications, and payment processing capabilities. Meanwhile Amazon took care of buying and
maintaining computer servers, providing transaction security, processing payments, and offering
storage resources to accommodate peaks and valleys of demand on its servers.
Amazon had become one of the largest retailers in the world without so much as a single brick-
and-mortar store. But it was much more than a retailer; it was the provider of infrastructure to its
competitors on and off line. This strategy meant that Amazon would rarely turn away a
customer searching for an item because it did not have that item in its own inventory. By
allowing third parties to plug their “inventory” into its own systems, Amazon increased customer
satisfaction. At the same time, it created yet another fast-growing revenue stream by monetizing
its excess computing infrastructure and charging competitors for selling their wares on its sites.
Amazon could do this because it had one of the most sophisticated computing and
communications infrastructures on the web, allowing it to transition from an application to a
platform company. It was not alone in this regard. Facebook was implementing technologies to
perform a similar transition, transforming itself from a social network to a worldwide
communication platform, allowing third parties to build games and services using its
communications infrastructure and member database.
Another trend still in its infancy but clearly gathering force was the transition of commerce and
communication from desktop computers to mobile devices. After the introduction of Apple’s
iPhone in 2008, Galperin watched as pure-play mobile commerce apps and companies grew like
mushrooms the morning after a wet storm. More than 300,000 individuals and companies signed
on with Apple to develop apps for the iPhone, and Apple’s app store soon hosted hundreds of
thousands of apps, most of them free for downloading by anyone in the world with an iPhone.
Google was not far behind Apple with its Android mobile operating system, which enabled
smartphone makers such as Samsung and Motorola to rush to market with iPhone competitors.
Galperin watched the transformation of U.S. companies with some trepidation. He too was
working furiously to transform his company from an application to a platform that would allow
third parties to connect to its network, and be mobile-app friendly. But the robust and reliable
proprietary system that had taken his upstart company to a dominant force in the market was now
acting as an anchor on innovation and speed. System enhancements and new feature
implementations were being delayed because the technologies Galperin used to launch his
company eight years prior were not designed for ongoing change. Plus, his system employed
Web 1.0 technologies that could not accommodate the transformation MercadoLibre had to make
from a stand-alone e-commerce application to a capable ecommerce platform.
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 9
FROM PRODUCT TO PLATFORM
After moving his office to be near his technologists, Galperin promoted a young but tough-as-
nails engineer, Dani Rabinovich, to take the CTO’s position, which had recently been vacated by
Eduardo Sokolowicz. Rabinovich’s task was as clear as it was almost insurmountable: to rebuild
the company’s systems from the ground up to support speed of innovation, third-party
integration, and mobile apps―and all this while the company continued to deliver on the
promises it made to its investors when it went public. Rabinovich described the challenge:
A platform was necessary for us to deliver a product that would survive in the
following decade. Even if we never shared our platform with any external
developer, a platform was (still) absolutely key for (our) survival. Main reason:
SPEED OF EXECUTION AND INNOVATION. A platform allowed us to
decouple a large monolithic system (and team) into many small cells. Each cell is
like a small company: (with its) own team, own infrastructure, own code and own
data. Only one aspect in common: a strict [adherence to] interface APIs and
strong authentication standards. Now the user-cell talks to the listings-cell exactly
the same way it talks to Facebook or Twitter.
Galperin and Rabinovich wrote a technology transition plan in 2009, and called it New World.
Galperin said during an interview in January of 2013.
“This was a bet-your-company type of a decision,” said Galperin when he was
asked if he was confident that the transition would succeed. “We ended up
calling the project New World, because the new system was going to enable us to
take advantage of new market opportunities.”
Rabinovich continued in an e-mail:
We had been discussing this since the early 2000s, but with our old, monolithic
technology, it would [have been] impossible or at least unsustainable. We tried
some shortcuts around 2006, but a true platform can't go far if you don’t use it
yourself. It was in 2009 that we decided to rewrite our entire product from
scratch.
Galperin and Rabinovich navigated organizational and technological minefields to reinvent the
company’s technology infrastructure and give life to one of its founding goals―to be the
dominant e-commerce platform provider in Latin America. After officially opening its platform
and publishing its APIs in Sao Paulo, in October of 2012, MercadoLibre was seeing significant
uptake in third-party developers adopting its platform for all types of sell-side and buy-side
applications. Its platform could be accessed by applications on mobile, WebTV, and tablet
devices as well as web and mobile browsers based on the new HTML5 standard―all developed
and deployed by third parties. To further encourage adoption, the company set in place
incentives for developers. For example, the company would share revenues with developers of
sell-side applications that bring traffic to its many sites.
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 10
By July of 2012, the company’s revenues were running at a $350 million annual rate, aiming for
net income around $100 million. The company’s market capitalization was at $3.5 billion―an
all-time high―and headed to $4 billion, and the analysts were looking at MercadoLibre as a
premier internet company with significant upside.
While these results might make the founder of any company pause to enjoy the accolades and
recognition for his achievements, Galperin was contemplating the entry of Amazon into the
markets he had created and dominated. Amazon was coming to Latin America with close to $10
billion in cash and marketable securities, $50 billion in annual revenues, and profits that had
averaged over a $1 billion in the past three years. Valued at $110 billion as measured by market
capitalization, Amazon was worth 35 times more than MercadoLibre. Moreover, Amazon was
not the only U.S. company preparing to enter the fast-growing Latin American markets.
American and European retailers, large and small, were eyeing Latin America to fuel their
growth. But Galperin was building a company tailor-made for the unique opportunities and risks
of the Latin American markets, designed to grow its markets and at the same time defend itself
from competitors. Galperin declared:
MercadoPago was in the business plan I wrote while at Stanford. To do
commerce we needed a capable payment system. We just needed to have the
resources and to build and deploy the capability, which we have now done. But,
as a world-class e-commerce company, we are now focused on two other critical
areas―logistics and fulfillment. 2
While the parcel delivery services had improved significantly since 1999, in 2013 Latin America
still lagged far behind developed markets in offering cost-effective and timely delivery services,
especially in the “last mile” delivery segment terminating at the buyer’s location. Galperin
identified this as a major obstacle that must be overcome in creative ways to continue growing
the business. In congested large cities, often the only viable form of transport was the
motorcycle; so motorcycle-based delivery companies were sprouting up to address the “last
mile” problem. But the service areas were patchy, costs were high, and late deliveries affected
customer satisfaction and customer service costs. As a down payment on trying to solve this
problem, Galperin contracted with reputable delivery companies, allowing them to integrate their
scheduling systems with MercadoLibre’s systems, giving them and their customers full visibility
of the shipping process. But this was only the beginning.
“In the next few years,” said Galperin, “we have big plans to address the logistics and fulfillment
problem with solutions specific to each of our markets.”
The path ahead was far from clear for Galperin and his team in 2013, even as they enjoyed the
tremendous successes garnered by their hard work.
2 The case draws on interviews by Professor Barnett at the MercadoLibre offices in 2011 and 2012, as well as the
author’s interviews with Marcos Galperin and Dani Rabonivich in 2012 and 2013.
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 11
Appendix A
THE SPANISH AND PORTUGUESE LANGUAGE INTERNET MARKET
Latin America
The region, which consists of South America, Central America, the Caribbean and Mexico, is
home to over 550 million people, or approximately 8.5 percent of the world’s population. The
International Monetary Fund estimates that Latin America’s combined annual gross domestic
product in 2006 was greater than $2.9 trillion. Based on information released by
InternetWorldStats.com, internet penetration as of June 30, 2012 was 66.4 percent for Argentina,
45.6 percent for Brazil, and 36.5 percent for Mexico. Between the end of 2000 and 2011,
InternetWorldStats.com estimated that Latin America’s internet user base grew 1,111 percent.
Such growth presented a significant opportunity for an internet-based marketplace provider. For
developing markets, e-commerce platforms offered advantages of scale, information availability
and accessibility to markets, while addressing many of the inefficiencies associated with
traditional offline trading. These included limited access to product and price information,
excessive number of parties in the value chain, limited inventory, and obstacles to efficient
communication and interaction between market participants. The Latin American markets also
had daunting challenges, above and beyond the challenges facing e-commerce companies, such
as liability arising out of sale of prohibited, dangerous or defective, and counterfeit or stolen
items.
Specifically, MercadoLibre faced risks associated with political and economic crises, terrorism,
civil unrest, and expropriation. These are not risks typically faced by western companies, but
MercadoLibre and other new economy companies were especially vulnerable to local politics
and social instability. For example, shortly after the founding, in 2001 and 2002, one of the
company’s key markets, Argentina, was mired in social and political upheaval, which had a
significant impact on its economy. Further, profitable companies were often targets of socialist
governments such as Venezuela, which had nationalized large and powerful western oil
companies, showing that they could move on any company they deemed too rich and powerful.
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 12
Appendix B
INTERNET APPLICATIONS AND ECOMMERCE PLATFORMS
In 1994, a small company by the name of Mosaic Communications Corporation released a
software product named Mosaic Netscape 0.9, which allowed Windows and Unix computer users
to connect to rudimentary network sites known as the World Wide Web through an intuitive,
easy to use, graphical user interface known as the browser. In 1996, when Mosaic, by then
renamed Netscape, went public, the Internet revolution was in full swing. Just about every
Global 2,500 company and government agency was developing an internet strategy almost
overnight, so as not to cede any competitive turf. Companies spent billions of dollars
collectively to project their internal systems over the Internet, by building and deploying
browser-based user interfaces to their existing customer service, human resource, enterprise
resource planning, and other corporate applications. These systems were mostly mainframe
computer-based or PC-minicomputer-based, client-server software systems. While the new
browser-based user interface allowed access to the system through the Internet, the actual
functionality and database of the systems were built on antiquated integrated technologies. It
was during this period that enterprise software companies such as Oracle, IBM, and SAP,
introduced internet versions of their applications and systems, allowing companies to build and
deploy their own custom internet applications using the vendor’s proprietary programming
languages and protocols. The key benefit of these applications was that they could be scaled to
accommodate large numbers of users as demand for access grew. The disadvantages, however,
were numerous and in many cases crippling. An application built using a single vendor’s
technology stack meant that innovation depended completely on that vendor’s ability to keep
pace with the furious speed at which internet technologies were evolving. Also, adding new
features and functions to these all-in-one integrated systems was time consuming, costly, and
small bugs in new functions could cause system-wide application outages.
By 2005, internet technologies and standards had matured to a point that scalable and reliable
web systems could be built using best-of-breed technologies from tens of vendors, enabling
companies such as eBay, Amazon, and Facebook to deploy application platforms with standard
APIs (application programming interface) and well documented interface programming guides.
Internet platforms are applications with separately implemented database, functionality, and user
interface layers, where each layer can be accessed by standards-based APIs, without disturbing
other layers. This meant that a new feature could be built by developing it on a separate
networked server and using published APIs only to connect with the platform database or other
needed functions, eliminating the need to change the code in the user interface and functionality
of the existing system. Because of this, it was not uncommon for the majority of a platform’s
capabilities to be developed by third parties. Decentralized API-based development had another
significant advantage. Developers in different parts of the same company could quickly build
and deploy new capabilities cost effectively, without the danger of introducing system level
errors that could bring down the entire system.
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 13
Exhibit 1
MercadoLibre’s Timeline
1999 2000 2004 2006 2007 2008 2009 2010 2011 2012
Founded Argen na Brazil
Mexico Uruguay
Columbia Venezuela
Chile
Peru Costa Rica Venezuela Panama
Users 6.5 18.2 24.9 33.7 42.6 52.9 65.8
Revenues 12.7 52.1 85.1 137 172.8 216.7 298.9 ~350
Net Income (2.2) 1.1 9.7 18.8 33.2 56 76.8 ~100
IPO Acquires DeRemate for $40M
March 10, 2000 Internet Bubble
Bursts
Figures in MM
Sources: SEC Form S-1 Registra on Statement, March, 2007 MercadoLibre Investor Presenta ons, 2008, 2011
SEC Form 10K, 2011 Annual Report, Feb. 28, 2012
Sept. 15, 2008 Lehman Bro’s Bankrupt
Mar. 6, 2009 Dow drops to
6,626 – down 50%
From high in 2008
Portugal
New Pla orm
Start Project New World
$3.5B Mkt. Cap
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
MercadoLibre IB105
p. 14
Exhibit 2
Financials – 2008 - 2012
For the exclusive use of C. ALFARO, 2024.
This document is authorized for use only by CLAUDIA ALFARO in International Entrepreneurship Sp 24 taught by DILEEP RAO, Florida International University from Dec 2023 to Jun 2024.
