Skip to Main Navigation
Download the report

Leveling the playing field

Data-driven business models have emerged as a way of converting data into economic value. Platform businesses, in particular, are solving development challenges through data-driven matching of suppliers and consumers, including in lower-income countries. Yet such businesses tend to concentrate. Newer and smaller players must overcome many obstacles in order to compete against larger established companies. Market concentration can be detrimental to the welfare of individuals who connect to those platforms as consumers, suppliers, or service providers. Policy measures are needed to ensure a level playing field and healthy competition in the market.

Data-driven business models

Data are increasingly being used by firms as a key input to create value in the production process. At the same time, data are continuously being produced as a byproduct of economic activity, and feeding back into the production process. With growing capacity to collect, store, and process that data, the ability of businesses to extract value from this data has been rising exponentially in recent years. At the same time, data-driven business models that create better goods and services are helping to address many development challenges.

Businesses use data from various sources as an input. Using data allows companies to make better data-driven decisions. This leads to quality improvements, lower costs, and new and innovative products. In India, for example, farmers can access a data-driven platform that uses satellite imagery, artificial intelligence (AI), and machine learning to detect crop health remotely and estimate yield ahead of the harvest. In Chile and Cameroon, telehealth clinics and their specialists can remotely monitor and diagnose patients in difficult-to-reach regions using sensors that collect patient health data and artificial intelligence to analyze such data.

Platform businesses, one of the most ubiquitous and transformative data-driven models today, reduce transaction costs and alleviate market failures by enabling better matching of supply and demand. In West Africa, for instance, platforms are matching farmers with idle tractors with farmers in need of a tractor.

Input dataEconomic activitiesUserprovideddataPublicdataBig dataanalyticsand AIIoT andother smartdevicesDigital ledgertechnologyincludingblockchainInferred datageneratedthroughanalyticsDataobserved onusers, objects,and processes,through economicactivityBig datacleaning andstorageBetterdata-drivendecisionmakingInnovativeproductsData-drivenmatching ofdemand andsupplyImprovedinter-mediationQualityimprove-mentsLowercosts

Note: AI = artificial intelligence; IoT = internet of things.

Crucially, data-driven businesses produce data as a byproduct, which in turn can be used as an important input for the data-driven business cycle. For example, e-commerce platforms use data created as a byproduct of browsing behavior and purchases to improve their product offerings. Credit card companies sometimes sell their transaction data for a specific location to tourism providers in that location.

Platforms in developing countries

Platform firms are one of the most transformative data-driven models. They are expanding across low-income and middle-income countries. They range from start-ups to businesses operating at scale and are a mix of both locally grown and foreign firms.

The diversity of new platforms is evident in recent research examining both start-ups and established platforms. At least 959 platform firms have established a physical presence in a set of 17 low-income and middle-income countries from all regions across four sectors that are important for jobs and/or economic productivity: e-commerce; transport and logistics (including both freight and passenger transport); agriculture; and tourism.

Platform firms by country and by sector

East Asia and Pacific
Europe and Central Asia
Latin America and the Caribbean
Middle East and North Africa
South Asia
Sub-Saharan Africa
45272210371110641993413612292132236141318913436109451718176113207455739121013624421117077E-CommerceTransport/LogisticsTourismAgricultureArmeniaTunisiaMoroccoSri LankaPeruUkrainePhilippinesColombiaBangladeshEgypt, ArabRep.KenyaSouthAfricaMalaysiaNigeriaRussiaIndonesiaBrazil

Source: World Bank Group Developing Economies Digital Platform Database 2020, using data from Crunchbase, Factiva, Thomsonreuters, and Alexa (downloaded in 2020:Q2).

E-commerce is the leading sector for platform firms, representing 82 percent of firms in the sample. The highest shares of e-commerce firms are found in South Asia and North Africa and the lowest in Europe and Central Asia. The agriculture sector tends to have the smallest share of firms across regions, with the exception of Sub-Saharan Africa.

Nigeria, Bangladesh, and Indonesia are platform hotspots, as measured by a relatively large number of platform firms in proportion to GDP, especially in e-commerce. In agriculture, Kenya and Nigeria stand out.

Across the focus countries, most platform firms are recent entrants; 55 percent were established in the past five years. Only 11 percent of firms were established more than 10 years ago. Firms also tend to be small. More than 80 percent have 50 or fewer employees. Almost half (47 percent) have 10 or fewer employees.

Ownership and traffic appear to be concentrated in a few large global platform firms

Although local data-driven firms are on the rise in low-income and middle-income countries, foreign-headquartered firms have a significant presence, underscoring the global nature of the data-driven economy.

Two-thirds of the top 25 sites in sampled low-income and middle-income countries are owned by just 6 US based companies.

A measure for the balance between domestically owned versus foreign-headquartered firms is their share among the top 25 most visited websites in countries.

Domestic and foreign-owned websites

Social media
Video communication
Software/computing resources
Real estate
ArmeniaBangladeshBrazilColombiaEgypt, Arab Rep.IndonesiaKenyaMalaysiaMoroccoNigeriaPeruPhilippinesRussiaSouth AfricaSri LankaTunisiaUkrainePlace in the ranking of the top 25 most visited websitesArmeniaBangladeshBrazilColombiaEgypt,Arab Rep.IndonesiaKenyaMalaysiaMoroccoNigeriaPeruPhilippinesRussiaSouthAfricaSri LankaTunisiaUkraine
Number undefined in undefined

Source: Alexa (downloaded 2020:Q2).

These are the top 25 most visited websites in the country sample, ordered from left to right according to the most visited site. The sites are colored by category. The “other” category includes news pages and government websites, as well as sites linked to academia, financial institutions, and telecom operators.

These are the foreign-owned websites. They represent 59 percent of the top websites, ranging from 84 percent in the Philippines to 20 percent in Indonesia. The presence of these foreign-owned websites underscores the global nature of the data economy and of the presence of large global firms in the data-driven ecosystem.

In fact, most of the top websites are owned by the same handful of companies headquartered in the United States. Two-thirds of the top 25 sites in the country sample are owned by Google, Microsoft, Facebook, Verizon, Amazon, and Zoom. Their presence is a reminder that the data economy is still nascent in lower-income countries relative to high-income economies.

Google, Facebook and Microsoft alone own half of the top websites. They are among the top 10 most visited websites in all low-income countries for which data are available.

When the market power of platform firms becomes entrenched, it can hinder new entry and harm users

While the rise of data-driven businesses can drive pro-development market opportunities, the possession of data can also provide firms with a competitive advantage that may tip markets into situations of entrenched concentration and market power, increasing the risk of exclusion of smaller firms and entrepreneurs, and exploitation of individual users.

Concentration in data-driven markets and entrenched market power brings risks to both consumers and suppliers, including service providers.

The risks of concentration in data-driven markets

Increased risk of anticompetitive practices For consumers, this may mean abuse of dominance that raises prices and reduces consumer choices, algorithmic collusion. For entrepreneurs, this may mean exclusion of smaller firms from the market.

Unequal bargaining power For entrepreneurs, this may mean unfavorable and untransparent contract terms and conditions for suppliers transacting on platforms.

Risk of excessive data collection For consumers, this may mean excessive collection and processing of their personal data. For entrepreneurs, this may mean platforms collecting and using data on their suppliers’ products that can then be used to enhance the platform’s own products.

Discrimination and bias For consumers, this may mean use of data to discriminate against them based on their consumer profiles or their willingness to pay. For entrepreneurs, this could mean platforms’ search and ranking algorithms being biased against independent entrepreneurs products and toward the platform’s own products.

Lower welfare and innovation A lack of market contestability reduces incentives for firms to innovate and improve productivity. In the longer term this results in higher prices and lower quality for consumers and businesses.

So, how do digital markets become entrenched? Let's take a mature e-commerce platform as an example. It holds a lot of (historical) consumer data and can create a more customized shopping experience, with more accurate product recommendations, preordered shopping baskets, and more consumer reviews. With its large number of consumers, this platform will also attract more suppliers through indirect network effects, raising users’ costs of switching to competing platforms.

Newcomers to the market, on the other hand, lack the big historical data of the established player, and will not benefit from the network effects and economies of scale. As a result, new firms will generate less data to help develop and improve their products and services and therefore attract fewer consumers. This, in turn, makes it even more difficult for them to innovate and compete with large firms.

Data can also ease a platform’s entry into adjacent markets. Well-known examples include M-Pesa’s move from money transfer into savings and loan products; Uber’s entry into food delivery and freight delivery; and Google’s evolution from search to shopping, maps, and other markets.

By combining multiple types of data, platforms can benefit from the broader scope of their data, which has spurred an increasing number of mergers aimed at accumulating data. A prime example is Facebook’s acquisition of Whatsapp.

What can regulators do?

Governments have two complementary competition policy tools to safeguard against the risks of entrenched market power. The first tool is enforcement of antitrust laws, with potential adaptions to the context of data-driven businesses. This involves detecting and punishing anticompetitive practices (where a firm abuses its dominant position or a group of firms enters into an anticompetitive agreement) or preventing anticompetitive mergers. The second tool—just as important as the first—is the design of regulations to allow data-driven firms to enter markets and compete on a level playing field while also protecting users.

Enforce antitrust rules

To tackle entrenched data markets, competition authorities can enforce antitrust laws. When a firm abuses its dominant position or a group of firms enters into an anticompetitive agreement, these anticompetitive practices are illegal and can be punished.

For example, with the heightened risk of dominant firms emerging in the data economy, several competition authorities around the world have detected and tackled cases in which firms abuse their dominance. The Global Digital Antitrust Database, a World Bank database of all finalized antitrust cases involving digital platforms, contains 30 cases of abuse of dominance from around the world that can be categorized into 13 different types of practices. Among these are practices that are directly related to the use of data by firms, including restricting competitors’ access to data, manipulation of search algorithms, and exploitative data collection and processing policies. The latter highlights the interface between data protection and competition policy.

Explore the different types of abuse and the associated cases finalized by competition authorities worldwide by using the filters below.

Restricting access to data
Filter by country
Restricting access to data
Dominant platforms restricting users’ access to data that could be used on competing platforms, or limiting access to data to only certain users.
Restricting access to data in online advertising
Google’s Terms of Service of the AdWords’ API (application programming interface) preventing advertisers from transferring data from Google’s platform to competitors' sponsored search platforms.
Restricting access to data in online advertising
Google’s Terms of Service of the AdWords’ API (application programming interface) preventing advertisers from transferring data from Google’s platform to competitors' sponsored search platforms.
Restricting access to data in online advertising
TREB restricting real estate brokers’ and consumers’ access to historical home sales data.

Source: Global Digital Antitrust Database, World Bank, Finance Competitiveness & Innovation Global Practice.

Another way competition authorities can enforce antitrust regulation is by blocking or enforcing conditions on potentially anticompetitive mergers and acquisitions involving digital platforms. For example, in 2019 the Federal Economic Competition Commission of Mexico (COFECE) decided to block Walmart’s proposed acquisition of Cornershop because the new company would have access to data on the sales of competing retailers through the platform, which was believed to put smaller rivals at a disadvantage.

Ensure a level playing field through regulation

Apart from taking antitrust measures after anticompetitive behavior has been identified, market regulation can be designed to allow data-driven firms to enter markets and compete on a level playing field while protecting users. This ex ante regulation includes data governance regulations, regulations focused on large data-driven platforms, and traditional sector regulation. Such regulation is just as important as antitrust enforcement—if not more so—especially where there is no functional competition authority.

Mandatory and voluntary schemes to improve access to data are an example of ex ante regulations. There are several options available to regulators here including facilitating voluntary data sharing between firms or providing the right to portability where consumer are able to port their data between different data controllers. Mandatory data interoperability—which goes further than portability by easing technical barriers to sharing—is useful when continuous sharing is required.

A successful example of this are so-called “open banking” regimes—whereby financial service providers are mandated to share data on user accounts to third parties through open application programming interfaces (APIs)—are a good illustration of this kind of regulation. Banking data have specific characteristics that make them well suited to data sharing initiatives because they are relatively homogenous and standardized—and the concept of “open banking” is now becoming well established in Europe. The United Kingdom’s Open Banking initiative is generally considered to have been particularly successful in spurring market entry and innovation, with 134 third-party providers currently registered and supplying services. At least nine other jurisdictions also have emerging open banking regimes in place.

Finally, governments can optimize offline regulation. For some data-driven businesses, the key to being able to enter and compete lies in improving traditional “offline” regulation. In some cases, these regulations protect or favor traditional or incumbent players at the expense of data-driven firms. In Morocco and Tunisia, for example, state-owned enterprises are not subject to the same data protection obligations that are binding for the private sector.