What are the takeaways?
Online platforms are an increasingly important part of the economy. They have specific economic features which require adjustments to conventional merger analysis and can have radically different implications for the assessment of how mergers between them will affect competition.
These features were seen in the recent merger between Just Eat and Hungryhouse, two online takeaway aggregation platforms, which was cleared unconditionally by the UK Competition and Markets Authority (CMA) in November 2017.
Online platforms are a type of two-sided market. They are intermediaries that bring together two types of participants – sellers of a particular product or service and their consumers. Sometimes the platform provides a direct forum in which these two groups can participate. For instance, online property portals such as Rightmove put estate agents in touch with house-hunters. Other platforms attract end-customers indirectly by providing a desirable independent product. A prominent example is Facebook, which draws people to its social network and sells their “eyeballs” to advertisers.
There are several economic features of online platforms that are useful to draw out.
First, because online platforms are two-sided markets, the prices that are charged to each side need to be set collectively to maximise the attractiveness of the platform overall. A typical pricing model is one where prices are free to end-consumers, while sellers pay. Sellers can be charged through one of several mechanisms, including a commission rate related to sales, a one-off fixed fee or a monthly payment for access.
Second, online platforms often exhibit network effects. These occur when the appeal of a product depends on the number of people using it. A distinction can be made between two types of effect:
- Direct network effects. These arise when the attractiveness of the platform to one side of the market depends on the number of participants on that side of the market. Social networks are good examples. The attractiveness of Twitter or Snapchat depends on how many other people use them. Direct network effects can also exist offline: telephones serve more purpose if there are lots of users.
- Indirect network effects. These occur when the attractiveness of the platform to one side of the market depends on the number of participants on the other side. Auction sites and marketplaces are good examples – in both directions. The attractiveness of eBay for buyers depends on the number of sellers, and vice versa. Content-based products selling advertising are good examples of markets exhibiting indirect network effects in only one direction. The more readers the Guardian newspaper has, the more attractive it is to advertisers.
Third, online platforms with network effects often exhibit “vertical differentiation”. This happens when, at the same price, one firm is preferred to another by all customers. For instance, if one marketplace has 50 sellers, and a second has the same 50 plus 50 others, all end-consumers are likely to prefer the second marketplace (other things equal). Vertical differentiation effects can be offset through changes in prices, so if the first marketplace is cheaper to access, not all consumers will have an obvious preference for one platform over the other. However, with online platforms, vertical differentiation is often present from the perspective of the end-consumer, for whom accessing the platform is typically free.
Fourth, online platforms may or may not exhibit “horizontal differentiation”. This occurs when some customers prefer one firm’s offer and others plump for its rival, even if both firms have the same prices. For instance, one platform might provide a pure retailing service with a particular emphasis on fast delivery, while another might provide retailing of the same products with more by way of editorial and market commentary. These platforms would attract different types of consumers.
What it means for merger analysis
Online platform markets with indirect network effects in both directions will have a tendency towards consolidation, possibly leading to a situation of “winner takes all”. This is because once a platform is seen to have a substantially better offer by one side of the market, the other side will follow, and vice versa. So an online platform with more sellers will attract more buyers, which in turn will lure more sellers, and so on.
The firm with the less appealing offer will find it increasingly hard to attract either side of the market. To begin with, it can overcome these disadvantages by subsidising one side or the other. However, once the largest platform has taken a sufficient lead, a smaller rival is unlikely to find it profitable to keep providing subsidies, as its anticipated earnings will not cover the costs involved. In markets with no horizontal differentiation, only one firm is likely to be left standing in the long run.
The extent to which this effect obtains depends on the strength of the network effects and the degree to which the platforms are horizontally differentiated. If network effects are weak or localised and platforms have highly differentiated offers, then multiple firms could survive in the long term. But they are unlikely to do so if network effects are strong and there is no horizontal differentiation.
Merger analysis of online platforms needs to take these features into account to understand whether a market is inevitably going to consolidate as smaller firms become unable to compete, or whether it can sustain multiple competitors.
An example: the Just Eat/Hungryhouse merger
Just Eat and Hungryhouse are online takeaway aggregators in the UK. They announced their intention to merge in late 2016. Just Eat and Delivery Hero (Hungryhouse’s parent company) operate in many countries. The merger related to the UK only.
Online takeaway aggregators bring restaurants and hungry consumers together. Just Eat and Hungryhouse operate a “pure aggregation” business model that targets traditional takeaway restaurants: they take the orders and pass them to the restaurant, but the restaurant is responsible for the delivery of the food.
Both Just Eat and Hungryhouse have been operating in the UK for over a decade. During this time, almost all other online takeaway aggregators have exited the market. However, a related business model, namely “aggregation and delivery”, has emerged. Deliveroo was the first to enter this market in 2014, followed by UberEATS (a subsidiary of Uber, the ride-hailing firm) in 2016. The aggregation and delivery model is most attractive to restaurants without their own delivery service.
At first sight, the Just Eat/Hungryhouse merger appears to be between the two closest competitors in this market segment, as both firms operate a largely identical business model and appeal to the same types of restaurants. At Phase 1 of its review of the proposed tie-up, the CMA had concerns that the pair had 100 per cent of the online takeaway aggregation market. Even if the definition of the market was broadened to include aggregation plus delivery operators, the combined share of the merging parties was well over 80 per cent.
How then did the CMA become comfortable at Phase 2 of its review that the merger would not result in a substantial lessening of competition? The answer lies in the existence of indirect network effects. Just Eat was a much larger business than Hungryhouse – eight times larger in terms of UK revenue, with a substantially larger number of restaurants and orders.
It seems reasonable that consumers would prefer a platform that provides a greater choice of restaurants. It is also reasonable to assume that restaurants would prefer a platform which generates a larger number of orders. As a result, the two parties were strongly vertically differentiated. The fact that the parties had similar business models – and that almost all of Hungryhouse’s restaurants were on Just Eat’s platform (with only a few listing solely on its own site) – meant there was limited horizontal differentiation. From the introductory discussion, it follows that there would be a tendency to consolidation in the market, and it would be difficult for a much smaller firm to compete with a much larger rival.
Chewing over the analysis
The CMA did not simply accept these observations from a theoretical perspective. At Phase 2, it went through an extensive evidential process to assess whether indirect network effects were present and whether Hungryhouse acted as a competitive constraint on Just Eat. In relation to the latter, it carried out an econometric analysis, looking across different areas to determine whether the presence of more restaurants on Hungryhouse in a particular area reduced the number of orders placed on Just Eat. The CMA’s study broadly followed an analysis that Frontier Economics carried out on behalf of Just Eat, which was submitted during the Phase 1 process.
The regulator found that Hungryhouse had had a competitive effect on Just Eat’s orders in the past, but that this constraint had diminished over time as Just Eat had pulled further and further away from Hungryhouse; by 2016 the constraint had disappeared entirely. The CMA also found that there was no prospect of Hungryhouse being able to reverse this trend, even if it did not find that Hungryhouse was inevitably a failing firm.
Ultimately, therefore, the CMA was reassured that the merger would not lead to a loss of competition, because competition between the two had essentially already disappeared.
The CMA did recognise that competition had emerged from rivals with differentiated business models, such as Deliveroo and UberEATS, and its econometric analysis found that these rivals were having a competitive impact on the parties. However, the focus of its analysis was very much on the competitive interaction between the parties.
Digesting the lessons
The economic framework used by the CMA for assessing the Just Eat/Hungryhouse merger was substantially different from that used in conventional mergers. This is a consequence of the unusual economic factors at play in the world of online platforms. These require different arguments to be made – for instance, that firms with horizontally identical business models may not be close competitors in a vertical differentiation sense; and that the existence of strong indirect network effects inevitably means that there are high barriers to entry for firms with identical business models to preexisting players, but with smaller numbers of buyers and sellers.
The CMA observed that the existence of indirect network effects and of the winner-takes-all phenomenon in online platform markets could not be taken for granted; it would need to be demonstrated in each case. Therefore, merely observing the presence of indirect network effects does not absolve parties from the need to demonstrate that they do not act as a competitive constraint on each other.
Having said this, the CMA’s recognition of the unique features of online platforms and the analytical framework set out in the Just Eat/Hungryhouse merger decision provide useful guidance as to how merging online platform companies should approach their submissions to competition regulators.
Frontier Economics advised Just Eat throughout the merger clearance process.