Google, Amazon, Facebook, Apple and Microsoft (GAFAM) make huge investments in research and development, with a cumulated investment of over USD 71 billion in 2017. In addition to these important investments, GAFAM have engaged in extensive mergers and acquisitions (M&A) activity. Between 2015–2017, GAFAM acquired 175 companies, most of which seem to be young and innovative start-ups.
Despite their intense merger activities and the vivid debates they generate, little is known about the the GAFAM’s merger strategies. With the exception of a report reviewing the CMA’s decision-making, there is no systematic analysis of the merger activity of the main digital platforms. This paper, available here, provides detailed information and statistics on the merger activity of GAFAM, and on the characteristics of the firms they acquire.
Section 2 present the digital platforms’ business model.
The authors identify the segments in which each GAFAM firm operates, i.e. the main categories of users they serve and the main revenue sources of each firm, using the information contained in their 10-K reports. Amazon derives revenues from the four segments where it is active: sales of goods (merchants), media (editors), devices (platform), and sales of digital services, mainly cloud services for business (businesses). Apple obtains most of its revenues from its devices (platform) and connected services such as the Apple Store (content), Apple Pay (merchants), iWork (businesses) and Apple apps (end-customers). Facebook is active in three segments: advertising, content (e.g. gaming) and consumers (Facebook, WhattsApp, Instagram). Google is active in four segments: editors (Google Apps, YouTube), consumers (search, maps, chrome, etc), advertising (AdWords, AdSense), and platforms (Android, Nexus phones). Microsoft is active in business (Office, Azzure), platforms (Windows, Surface, Xbox, Lumia phones), content (development tools), consumer (Bing) and advertising.
For every firms except Microsoft, a single segment generates more than 80% of all revenue (Amazon with merchants; Apple with its devices; Facebook and Google with advertising. Business services are the source of almost 60% of all of Microsoft’s revenues).
Section 3 provides detailed information on GAFAM merger activities between 2015-2017.
The authors review 175 mergers involving GAFAM between 2015 and 2017. Microsoft and Google were by far the most active in terms of acquisitions, with 52 and 40 respectively, while Facebook was the least active with 20 acquisitions. In this short sample period, 2015 was the busiest year for these companies, with 65 acquisitions. GAFAM mostly bought rather small and young technology companies, with some outliers. The median acquired firm was aged four, had completed two funding rounds and collected USD 7 million.
In addition to these statistics, the authors collect information on the target companies’ business and products. The two most important target segments are business (61 acquisitions) and editors (43 acquisitions). By contrast, the segments with fewer mergers were advertising (one acquisition) and merchants (five acquisitions).
Section 4 reviews the individual M&A activity of each GAFAM company.
This section takes a closer look at the number of acquisitions by each GAFAM firm, and the segments where such acquisitions took place.
Amazon acquired 30 firms in this period, focusing on companies operating in the business, merchants and editor segments. These figures reflect a fairly strong focus on the company’s already successful product lines. Amazon used the acquisitions to strengthen its core business and to develop the services it provides to the business segment.
Apple made 33 acquisitions. Twelve occurred in the platform segment, mainly concerning technologies for its devices; eight were in the business segment, mainly concerning data analytics; seven acquisitions concerned editors, particularly companies focusing on the development and management tools for music, podcasts and videos; and six acquisitions took place in the consumer segments, e.g. navigation and mapping services, health software and a picture editing application. As with Amazon, the numbers seem to indicate that Apple is using its M&A activity to reinforce its current business model.
Facebook undertook 20 acquisitions in those three years. Eight, i.e. the majority, took place in the editor segment, mainly concerning video streaming technology and content management services for publishers. The consumer and platform segments saw five acquisitions each. The analysis reveals that Facebook used its M&A activity for two purposes. First, by acquiring consumer services’ companies, mostly social media apps, it strengthened the non-money side of its core business. By providing more content on its social networks, Facebook keeps the consumers within its ecosystem and it can monetise the time spent by them on the platform. A second motivation for its acquisitions lies in Facebook’s move into VR hardware and software. This constitutes an entry into a new segment for the firm.
Google was the most active firm, with 55 acquisitions in this period, covering all product segments under analysis. Acquisitions in services for advertisers, as well as 11 acquisitions in the consumers segment (e.g. mapping and navigation services, photo editing and storage software as well as web browser add-ons), can be seen as investments in Google’s most important activities. The 14 acquisitions in the business segment are purchases of cloud services, productivity software and professional communication products. This broad M&A strategy might suggest vast ambitions. The acquisitions in consumer and advertising services are clearly meant to bolster Google’s main source of revenue, while the 17 mergers in the editor segment reflect the importance of content and applications made available through Google’s online services and Android platform. However, the 14 acquisitions in the business segment represent an attempt of entry into a new segment.
Finally, Microsoft made 40 acquisitions, 26 of which in the business segment (cloud services, productivity software, management software like CRM or sales software, analytics technology and professional communication products). There were also eight acquisitions in the editor segment (mainly development tools for applications and games), in addition to some acquisitions concerning the operation of platforms (e.g. keyboard functionalities). In other words, the vast majority of Microsoft’s acquisitions concerned its core focus on business clients.
In short, it appears that GAFAM use acquisitions to reinforce their business model, by acquiring firms active in their core segments or in complementary segments. This does not suggest intensive competition between the five firms, but rather that they use mergers to reinforce their most successful products where they already enjoy a strong market position.
Section 5 analyses how acquired firms and products fare after the merger.
105 branded products, amounting to 60% of all mergers, were discontinued within a year of the transaction. In 47 cases, the targets’ products remained active and continued to be offered just as before the acquisition. In 23 cases, or 13% of all products, there was not enough or clear information about the target’s product. This can be seen as indication that GAFAM firms are often not interested in the market performance of the firms and products they acquired, but rather in the knowledge they possess / incorporate.
Looking at each firm in turn, Apple shuts down most products. In 26 out of 33 cases, the target’s product disappeared in its initial form after the transaction. This reflects Apple’s choice of a closed system of products. A similar pattern emerges for Facebook. Amazon, Google and Microsoft, however, keep between 27% and 37% of their targets’ products up and running. For Amazon and Microsoft. this could mean that they want to diversify their product offering. For Google it might as well be a consequence of its rather expansive M&A strategy. When entering new segments, an acquirer could have higher incentives to keep acquired products on the market. To examine this question further, the authors run a Probit regression. The estimations show that younger firms are less likely to be continued. Indeed, younger startups are more likely to be bought for their knowledge rather than their products, making shutdown more likely. In addition, GAFAM are more likely to discontinue a product when it is part of their core segment.
Section 6 looks at potential killer mergers.
Finally, the authors use the data to search for a possible killer acquisition, i.e. the acquisition of a potential competitor. In the digital economy, a firm that managed to attract a large user can extend its product bundle and turn to a sizeable competitor of an existing platform, even if products are a priori different. Identifying products segments by user groups, as this paper does, rather than by functionality is an interesting tool to identify potential competitors of incumbent platforms.
A first condition to identify a potential killer merger is to have an acquisition in the firm’s core segment where the acquirer enjoys a strong market position. To be a potential competitor, the target should also have a sufficiently large user base. Last, the product of the target should be continued: when a firm has a large user base, changing its brand name may move consumers away, especially if there are network effects, switching costs and brand loyalty. For these reasons, a sizeable competitor is likely to continue to operate under its pre-acquisition name. Few (12) acquisitions appear to meet these criteria, and most of them complement the acquirer’s existing products, i.e. these are mainly vertical mergers improving the acquirers’ products. Of these twelve acquisitions, there are only three cases where the target could represent a competitive threat to the buying firm because of its large user base: the Amazon/Souq, Microsoft/LinkedIn and Facebook/Masquerade (MSQRD) deals.
This is an insightful analysis of the merger strategies of digital giants. According to the authors, two main conclusions flow from the analysis. First, most acquisitions are undertaken in core segments or other segments in which these firms are already active. Second, the majority of acquired products is discontinued post-acquisition. This suggests that many GAFAM acquisitions are driven by the desire to purchase valuable R&D inputs, such as the technology, IP rights and/or people of the target firms. Second, GAFAMs’ focus on already known and important segments raises the question whether these acquisitions are undertaken to increase market power or to realize synergies. The answer to this question is far from obvious and requires a case by case analysis.
The authors consider that, given the small size of target products, classical synergies seem rather implausible. Except for beneficial effects on innovation, the likely motives in these cases are the desire to improve market positions and to increase market power by adding new functionalities to GAFAM’s already successful products. From a competition perspective, the fact that these firms are acquiring mostly in their core segments suggests that they are seeking to reinforce their market positions. However, almost all GAFAM acquisitions fly under the radar of competition law. The low revenues of many targets prevent these cases from falling under the jurisdiction of antitrust authorities. As such, there is a case for amending merger control thresholds to allow such transactions to be subject to scrutiny by competition agencies.
This is fair enough as far as it goes, but it also seems to run against the authors’ conclusion that most transactions seek to improve the dominant product and are not driven by attempts to exclude potential competitors. If this is the case, it is unclear how each individual merger can detrimentally affect competition and consumer welfare. It follows that it is also unclear why the authors think that more stringent merger control would be needed – unless one assumes that the cumulative acquisition of target companies by GAFAM’s detrimentally affects competition in the latter’s core markets, or the development of competition in complementary markets. I understand this is not the focus of the paper, but the same way I would have liked to see a clearer explanation of how the analysis of GAFAM firm’s merger practices links to the discussion on product discontinuation / killer acquisitions (interesting as both discussions are), I would also like better to understand what is the competition harm that the authors infer from the results of their analysis.