The European Commission’s 254-page decision approving Google’s acquisition of Fitbit (the “Decision”) offered the Commission a so-far unique opportunity to apply the 2019 report on “Competition policy for the digital era’s” (the “Digital Era Report’s”) recommendation to assess acquisitions by “gatekeeper platforms” of much smaller companies active in complementary markets using an “ecosystem” approach. Google/Fitbit would seem to be an ideal test case for such an approach, involving a large number of complementary markets for smartphones, smart wrist-worn wearable devices, operating systems (OS) for both types of device, apps, app stores, online advertising and the nascent digital healthcare sector. Unfortunately, the Commission passed up the opportunity.
Indeed, the Digital Era Report argued that traditional theories of harm should be “rethought” where the “acquirer operates a multiproduct platform and/or an ecosystem that benefits from strong positive network effects, which act as a significant barrier to entry” and where “the risk to competition resulting from an acquisition is not limited to the foreclosure of rivals’ access to inputs, but extends to the strengthening of dominance as it fortifies the dominance of the ecosystem.” The report noted that “Even though the incumbent may not be dominant in a complimentary service market when that market is analysed as a separate market, a broader view of the position of the incumbent in a ‘market for the digital ecosystem’ may justify a finding of a significant impediment to effective competition” (pp 121-122).
The Commission did take such an approach in its preliminary report on its sector inquiry into the Consumer Internet-of-Things (the “Preliminary Report”) which noted that “a large number of respondents consider the main obstacle to developing new products and services to be the inability to compete with Google, Amazon and Apple . . .because these vertically integrated companies have built their own ecosystems within and beyond the consumer IoT sector by combining their own and integrating third-party products and services into an offering with a large number of users” (p 7). “Through their ecosystems combining voice assistants with search and/or marketplaces, and/or operating systems and/or app stores,” the Preliminary Report notes, “Google, Amazon and Apple have a unique position in the consumer IoT sector. As respondents point out, with every new smart device or consumer IoT service added these three consumer IoT ecosystems can realise growth through network effects and obtain unprecedented access to user (and sometimes competitor) data” (para 125).
Far from rethinking its approach to online platform acquisitions of complementary businesses in Google/Fitbit, however, the Commission followed its traditional narrow, market-by-market basis. Indeed, the Commission was arguably less exacting than in past cases and accepted remedies in some respects unprecedented and unsupported by its own Remedy Notice. The same remedies were rejected by the Australian ACCC and criticized by the UK’s CMA Chief Executive, Andrea Coscelli.
As might be expected in view of Google’s “multiproduct and/or ecosystem” business, the transaction involved a large number of markets. The Commission took a narrow approach, defining markets in ways that minimized direct overlaps.
Wrist-worn wearable devices
Fitbit is primarily active in fitness trackers and smartwatches, which the Commission groups together as the “wrist-worn wearable” market. The Commission notes that “the Parties do not overlap as regards wrist-worn wearable devices” (para 79), without discussing Google’s rumoured pre-Fitbit plans to launch a “Pixel Watch” and ruling its Verily Study Watch out of the market (footnote 260). The Commission analyses the market positions of Fitbit and its competitors for wrist-worn wearable devices as a whole and separately for fitness trackers and smartwatches (paras 312-327) and briefly discusses other segmentations (para 328).
But the Commission did not consider arguably the most obvious and relevant alternative segmentation of wrist-worn wearable devices, based on smartphone ecosystem. Apple Watches do not work with Android smartphones and thus are not relevant to assessing effects within the Android ecosystem. Arguably, only wrist-worn wearable devices actually paired with Android smartphones should be included, in view of the Commission’s recognition that “[a]ll customers who purchase wrist-worn wearable devices also purchase smartphones” (para 732) and that Android smartphone users “show a significant loyalty to the Android ecosystem” (para 800). The Commission also does not discuss possible segmentation of wrist-worn wearables based on the wrist-worn wearable OS.
With respect to OS, the Commission defines two markets for “licensable OSs for smart mobile devices” and “licensable OSs for wrist-worn wearable devices” (para 100). Google is active in both markets (with Android and Wear OS), while Fitbit is active in neither (because it does not license its wrist-worn wearable OS). Indeed, since Apple does not license iOS or Watch OS, Google’s EEA share in both markets is 90-100% (paras 329-330).
The focus on licensable wrist-worn wearable OSs makes sense when considering the transaction’s effects on Google’s incentive and ability to foreclose access to Wear OS (see below), but obscures other potential anti-competitive effects. These include innovation competition between licensable and proprietary wrist-worn wearable OS or the possibility that the transaction could extend the Google/Apple smartphone OS duopoly into the wrist-worn wearable market(s), as feared by the ACCC.
The Commission’s discussion of licensable wrist-worn wearable OSs also seems curiously outdated. The Commission mentions Samsung’s Tizen as a potentially licensable wrist-worn wearable device OEM (para 547), even though Samsung had not done so in the past. In any case, this possibility is moot in view of Google’s and Samsung’s plans to merge Wear OS and Tizen. The Commission’s statement that Huawei’s future wrist-worn wearable OS (being developed because Huawei was forced off Wear OS by U.S. sanctions), would represent a viable alternative to Wear OS for non-Chinese OEMs (para 547) is open to question.
The Commission distinguishes between separate markets for app stores for a given OS platform of smart mobile devices (in particular Android app stores) and the supply of app stores for a given OS platform of wrist-worn wearable devices (in particular app stores for Wear OS and Fitbit devices) (para 125). On this basis, each of Google and Fitbit have 100% shares in the app stores for their own platforms, with no overlap (paras 334-335).
In the market definition context, the Commission does not discuss the fact that wrist-worn wearable app stores are part of the ecosystem of users’ smartphones since wearable users access companion apps and app stores via their smartphone app store. (This relationship is discussed as a potential vertical foreclosure issue (see below)). As the Commission notes in the Preliminary Report, “Google and Apple are providers of the leading operating systems for smart mobile devices, and as such, they also operate the two main app stores that determine access to consumer IoT services via applications” (para 124).
The Commission stuck with precedent in defining a market for general search services (para 142), a market in which Google’s national shares generally above 90%, but in which there is no overlap (para 337).
Online advertising and ad tech services
Again following precedent, the Commission distinguished between online search advertising services and online display advertising services (para 155), in relation to online advertising, and between search ad network services, display ad SSP services and display ad DSP services (para 168), in the case of ad tech services. Again, while Google has high shares in many such national markets (paras 338-350), there is no overlap.
In addition to app stores, the Commission considers potential markets for a number of individual apps and services provided by apps. These include health and fitness apps (para 186), mobile payment services (para 201), navigation (para 217), virtual assistants (para 231), digital music distribution (para 248) and digital translation (para 260), potentially segmented by platform.
The only overlaps on this basis were in health and fitness apps and mobile payment services. The Commission concluded that the parties’ combined share in “health and fitness apps on smart mobile devices” is only 5-10%, including all active users, regardless of the device on which they were or could be used (the largest being Samsung and Apple) (paras 351-354). The picture was similar for mobile payment services (paras 355-359).
The Commission “examined four potentially relevant data-related activities in the digital healthcare sector: (i) provision of cloud and data analytics services, (ii) patient monitoring services, (iii) provision of data for medical research and real-world evidence [(“RWE”)], and (iv) corporate wellness programmes” (para 266). The Commission found no overlap in any of these markets and low market shares in all (para 378). How these shares were calculated is not explained, but if the Commission’s market size estimates include all revenues, including for products and services for which wrist-worn wearable devices and data are not directly relevant, these low shares would not be surprising.
The Commission acknowledged receiving comments to the effect that these categories may be too narrow, since “the digital healthcare market is dynamic and while the proposed segmentation may be relevant today it is not clear to what extent it will remain the case in the future” (para 268). The Commission could not be expected to examine all possible segments in this rapidly growing sector, but there are some surprising omissions.
In particular, it is surprising that the Commission did not examine potential digital healthcare segments in light of the types of data required and the sources for such data. The Commission lists 91 data metrics collected by Fitbit devices (Table 27), but does not discuss which metrics are used in which digital healthcare applications or the relationship between such data and “health and fitness” data. Similarly, the Commission considers corporate wellness programmes as a whole (where Fitbit has a very low share) as a potential market, but does not consider a potential market for data necessary or appropriate for such programmes (where Fitbit’s position might be more significant).
As with licensable wrist-worn wearable OSs, the Commission’s focus on data provided for medical research and RWE makes sense when analyzing potential foreclosure effects for companies relying on third parties to supply such data. On this basis, Google has a very limited share, while Fitbit is not present at all. But this approach obscures potential horizontal effects in digital-healthcare-related data that are collected and used internally rather than being monetized as a separate product.
It is also surprising that the Commission considers corporate wellness programmes as a relevant market in the digital healthcare sector but not insurance (although the Decision mentions insurance as a possible “use case” (para 490)). The Google group has long been interested in insurance and insure-tech and has made a number of acquisitions in this area. In August 2020, Google’s sister company Verily (which also makes the Verily Study Watch) founded Coefficient Insurance to offer “a new data-driven solution to employer stop-loss insurance.”
Based on the above-market definitions, the Commission identified 28 “affected markets” or segments for further analysis, all but ten of which are described as “vertical” or “conglomerate” in nature (Table 26).
Horizontal affected markets
With one exception, the “horizontal” affected markets involve Google search, online advertising and ad tech. These were considered horizontal affected markets not based on a direct overlap (Fitbit is not active in any of these markets), but from the acquisition of Fitbit as a source of data in combination with Google’s existing data. (The exception is mobile payment services, where the Commission ruled out concerns based on Fitbit’s small share and the presence of numerous competitors (paras 386-387)). The Commission notes that the horizontally affected markets in search, online advertising and ad tech are not affected in the “traditional sense,” but because Google’s access to “data obtained from Fitbit’s wearable devices, would increase Google’s power in some data-based markets by further strengthening Google’s ability to commercially exploit such data” (para 399).
According to the Commission, it is “foreseeable that Google’s availability of a new dataset from Fitbit post-Transaction could add to the current portfolio of data exploitable by Google, in particular for advertising purposes and in digital healthcare. According to paragraph 36 in the Horizontal Merger Guidelines, a merger can significantly impede effective competition if the merged entity gains such a degree of control over an asset that expansion or entry by rival firms may be more difficult. . . In the present case, the Transaction would allow Google to combine its already very prominent datasets with those of Fitbit, thus strengthening the Parties’ ability to supply relatively better services in certain data-based supply markets and foreclose the competitors’ entry and ability to expand in such markets” (paras 400-402).
The Commission concluded that Fitbit data would represent an important asset for Google that would further strengthen Google’s dominance in the markets for the supply of online search advertising services. Such “concerns would exist in view of: (i) the relevance of the data (and data collection capabilities) acquired by Google . . .; (ii) Google’s position in the relevant markets . . . ; (iii) the strengthening of Google’s market position and impairment of rivals’ expansion in the mentioned markets as a result of the data combination; and (iv) the absence of countervailing entry or buyer power” (para 428).
The Commission noted that “the Transaction is likely to have a negative impact on the development of an unfettered competition in the markets for online advertising” given “the large amount of data already used for advertising purposes that Google holds, the increase in Google’s data collection capabilities, which goes beyond the mere number of active users for which Fitbit has been collecting data so far . . . because, given the large analytics capabilities of Google, it cannot be excluded that Google could made inferences . . . for groups of individuals larger than the number of users for which Fitbit today, and Google post-Transaction, collects data” (para 455). While “Fitbit is just one of many sources of health and wellness data, . . . none of Google’s competitors in online advertising has access to a database or data collection capabilities equivalent to those of Fitbit and it is not likely that they would acquire such assets without incurring into significant costs and in timely manner. . . In fact, no competitors of Fitbit seems to make its data available for advertising purposes” (para 457). It would have been useful to hear more about Google’s ability to draw inferences from diverse databases, but this possibility is not further discussed or quantified.
The Commission reached the opposite conclusion in relation to the transaction’s impact on competition in general search services (because Fitbit data will be less relevant) and in digital healthcare. As regards digital healthcare, the Commission concluded that “the Transaction will not likely lead to any significant impediment of effective competition as a consequence of the likely horizontal effects arising from the combination of Google’s and Fitbit’s user databases and data collection capabilities” (para 496).
The Commission rules out such concerns because “the parties are neither actual nor potential competitors in the collection or marketing of user health and fitness data” (para 484), even though the parties were also neither actual nor potential competitors in online advertising. With respect to Fitbit’s data, the Commission noted “that there are alternative data providers available: (i) health data is also collected by smartphones (Apple, Huawei, Samsung) and can be shared with and accessed via health and fitness apps (Strava, MMF, MyFitnessPal, etc.); (ii) there are “aggregation services” or “aggregators”, such as Validic and Human API, which use a platform to connect multiple individuals, mine and collect their health data. . .; (iii) health data can . . . be collected also through corporate wellness programmes (data from health risk assessment, exercise data, lifestyle data, etc.); and (iv) electronic health records, that are aggregated and used to provide analytics services” (para 483).
This list of alternative data sources is curiously superficial. The Commission lumps together data obtained through different sources (e.g., wrist-worn wearable devices, smartphones and medical records) that seem unlikely to be substitutable. For example, smartphones and wrist-worn wearables both generate data on their users over long periods, but wrist-worn wearable devices, in constant contact with users’ skins, likely collect different data. The Commission does not compare the 91 data metrics collected by Fitbit devices to those collected by smartphones or discuss the relevance of those similarities and differences for different use cases. Medical records may be more reliable and precise than smartphone or wrist-worn wearable devices in relation to at least some metrics, but cover much shorter periods (e.g., during visits to a doctor or hospital).
The Commission also lumps primary sources of data such as smartphones and wrist-worn wearable devices with intermediaries like aggregators that are dependent on those primary sources. The Commission similarly lumps together the complete data collected on a real-time basis by smartphones and wrist-worn wearable devices with the individual files users share with health and fitness apps (for example data on individual exercise sessions).
Importantly, contrary to its approach in the online advertising sector, the Commission does not consider the potential for Google to make inferences no competitor could match by combining Fitbit data with its other “very prominent datasets,” including but not limited to health and fitness/medical data. For digital healthcare purposes, Google’s ability to make such inferences would arguably be much greater than in online advertising, since Google could make full use of both parties’ global databases (not just data relating to EEA users) for digital healthcare applications and draw health-related inferences by correlating Fitbit data with many other types of data. As in online advertising, no digital healthcare competitor will have “access to a database or data collection capabilities equivalent to those of” Google and Fitbit.
For vertically affected markets, the Commission considers potential input foreclosure risks, noting that for such concerns to arise the merged entities must have a significant degree of market power in the upstream market to be able to significantly influence prices and supply conditions in the downstream market (para 500). Whether a merger creates an incentive to foreclose depends on the trade-off between lost profits in the upstream market due to any reduction of input sales to rivals and the profit gain, from expanding sales or raising prices downstream (para 501). This Non-Horizontal Merger Guidelines-based summary of input foreclosure analysis is arguably overly narrow in a transaction like Google/Fitbit, where the ability to foreclose is not limited to increasing prices or refusing to sell, and incentives are not based solely on tradeoffs between upstream and downstream profit margins.
In relation to data for digital healthcare, the Commission concluded that Google would have the ability to engage in input foreclosure by restricting access to Fitbit’s Web API (para 520) and that “it cannot be excluded that, after the Transaction, Google will have the incentive to restrict access” (para 525). The Commission recognized that digital healthcare is a nascent and fragmented sector hosting a large number of very active start-up companies and will likely diversify and grow to a significant economic size. A significant number of competitors rely on access to Fitbit’s Web API for user health data. A restriction or interruption of third-party access to the Web API would negatively affect providers of apps and websites that capitalize even on relatively small amounts of Fitbit users’ data to compete and contribute to innovation and diversification of the digital healthcare sector (paras 526-529).
By contrast, the Commission rejected arguments that Google could have the incentive and ability to foreclose access to Wear OS to the detriment of wrist-worn wearable suppliers. The Commission rejected this concern in part on the basis that Samsung’s Tizen and Huawei’s future wrist-worn wearable OS represent alternatives for Wear OS users (para 547) and that Wear OS suffers from technical limitations (para 550). As noted, the Commission seemed unaware of Google’s and Samsung’s plans to merge their wearable OSs, likely addressing Wear OS’ technical limitations and eliminating Tizen as an alternative to Wear OS.
Perhaps more plausibly, the Commission felt that Google would not have an incentive to foreclose access to Wear OS, since it “launched Wear OS precisely with the aim of attracting users to its ecosystem with the aim of preventing their exit” (para 560). Indeed, Google’s strategy of using Wear OS to strengthen its ecosystem begs the question whether Google’s acquisition of Fitbit, eliminating one of the leading proprietary wrist-worn wearable OSs in the Android ecosystem, combined with the merger of Wear OS and Tizen, could enable Google to extend its control over the Android ecosystem into wrist-worn wearables (for example by pre-installing apps on third-party devices that collect data on users).
The Commission further rejected concerns about Google’s post-transaction ability to foreclose access to Google apps and services, access to Google Play or Google Search or access to Fitbit app stores to the detriment of wrist-worn wearable suppliers. The Commission stressed that Google’s strategy is to maximize distribution of its apps in the Google ecosystem (e.g., para 605). Again, the Commission didn’t consider whether this strategy could itself create anti-competitive concerns if Google seeks to leverage the Fitbit acquisition and the Wear OS/Tizen merger to maximize its control of the smartphone/wrist-worn wearable ecosystem.
Under the heading of “conglomerate effects,” the Commission considered whether foreclosure effects could arise through the combination of products in related markets, giving Google the ability and incentive to leverage a strong market position from one market to another closely related market by means of tying, bundling or other exclusionary practices, reducing competitive pressure on Google and allowing it to increase prices (para 711). From the perspective of the transaction’s potential to increase Google’s control over the Android ecosystem, this focus on price increases seems unduly narrow.
The Commission found that Google would not have the ability or incentive to foreclose competition by withholding technical support for competing wrist-worn wearable devices interoperating with Android or by using commercially sensitive information to harm competing third-party apps (paras 772, 830). This contrasts with the Preliminary Report’s findings that, “Some wearable device manufacturers have raised concerns about not receiving full information from smart mobile device operating system providers regarding new devices, operating system updates or privacy policies, which sometimes results in technical problems on the side of the wearable device manufacturers. In this respect, it is reported that unresponsiveness or delayed notifications of updates to smart mobile device operating systems (e.g. iOS and Android) create difficulties for wearable manufacturers. For instance, some features might be disabled following an operating system update, which creates consumer confusion. In other cases, the update might cause bugs and connectivity problems” (para 472).
On the other hand, the Commission considered that Google would have the ability to foreclose competition by degrading interoperability of competing wrist-worn wearables with Android by degrading the relevant application programming interfaces (“APIs”) (para 772). In reaching this conclusion, the Commission noted the large pool of common customers, Google’s control over Android, Google’s dominant position in licensable OSs for smart mobile devices, and the technical possibility for Google to selectively degrade interoperability if it changed its business model.
The Commission further decided that Google would have the incentive to degrade interoperability with Android after weighing Google’s qualitative arguments based on its business model and past practices against a quantitative analysis showing that degrading the interoperability of Android mobile devices with competing for wrist-worn wearable devices could be profitable (paras 792-804). The Commission noted that Google could collect more data by selling more Fitbit devices, but only in passing, cross-referencing the previous, incomplete discussions of Fitbit data in online advertising and digital healthcare (para 805(b)). The Commission did not attempt to estimate the value of Fitbit data for Google’s digital healthcare business (although a 2019 Morgan Stanley study estimated Apple’s market opportunity in healthcare as up to $313 billion by 2027).
To address these concerns, Google offered, and the Commission accepted, three final remedies, referred to as the “Ads Commitment,” the “Web API Access Commitment” and the “Android APIs Commitment.” These remedies included highly unusual, if not unprecedented, provisions.
Under the Ads Commitment (discussed in paras 964-973), Google agreed that Fitbit health and fitness data would be stored in a “silo” subject to strict access rules and not allowed to be used for Google’s online ads. In previous cases, the Commission accepted hold-separate remedies to ring-fence competitively sensitive information (e.g., Wegener/PCM/JV (2005), EdF/Louis Dreyfus (1999) and American Home Products/Monsanto (1998)), rather than to limit the competitive benefits of combining the buyer’s and target’s assets.
Under the Web API Access Commitment (discussed in paras 974-984), Google committed to continuing making available Fitbit’s Web API to allow API users to have access to data that Fitbit releases, subject to compliance with Google’s terms of service and privacy and security requirements. The commitment applies not only to data made available by Fitbit today but also to new data types identified based on benchmarking across other OEMs.
The Android APIs Commitment (described in paras 985-1009), which aims to prevent Google from degrading interoperability with Android via the APIs third-party wrist-worn wearable customers need to interoperate with their Android smartphones, is more complicated. Google committed to continuing making “Core Interoperability APIs” available without discriminating against competing wrist-worn wearable OEMs. Core Interoperability APIs are defined as Android APIs licensed as part of the Android Open-Source Project (“AOSP”) offering at least the functionality of pre-transaction APIs in AOSP, including improvements to allow competing wrist-worn wearable devices (or their companion apps) to perform a number of “core functionalities.”
Since APIs contributed to AOSP are public by definition, and Google decides which APIs to contribute to AOSP and which to keep private (or to make available through other channels, such as Google Mobile Services (“GMS”)), this commitment arguably offers no protection at all. In response to a request for information on this point, Google argued that “any improvements to the Wearable APIs’ functionality in respect of [Fitbit] Devices will be made available, without differentiation, to [Fitbit competitors], and the Wearable APIs will be licensed under the same terms as other AOSP APIs” (para 993). The Decision seems ambivalent on this point, however, since it notes that pre-transaction (with no incentive to foreclose) Google already contributed a less desirable API for a core functionality (location) to AOSP while offering a better API to GMS users. The Decision says this concern is “fully addressed” because Fitbit competitors will have access to GMS APIs to the extent that they are made available to other Android Smartphone App Developers (para 994). In other words, Google can discriminate in favour of Fitbit and against Fitbit competitors even in relation to APIs for core functionalities so long as it also discriminates against Android smartphone app developers.
The Android APIs Commitment contains further obligations that apply to “Android APIs,” defined as “the APIs, including any improvements or bug fixes, that Google licenses to Android OEMs without charge for access, either as part of AOSP or GMS, for use by Android Smartphone App Developers with an Android App.” In this part of the Android APIs Commitment, Google commits not to discriminate against Fitbit competitors in relation to Android APIs via-à-vis Android smartphone app developers (Clause A.3(12) and (13)). Again, Google is allowed to discriminate in favour of Fitbit, so long as it discriminates equally against Fitbit competitors and third-party app developers who do not compete with Fitbit.
This aspect of the Android APIs Commitment is highly unusual. While the Commission has previously accepted non-discrimination remedies in vertical and conglomerate mergers (e.g., Qualcomm/NXP, Broadcom/Brocade, and ARM/Giesecke & Devriendt/Gemalto), Google/Fitbit appears to be the first case in which the Commission accepted a “non-discrimination” remedy that expressly permits self-preferencing. The Decision doesn’t explain why Android smartphone app developers, rather than Fitbit itself, are the appropriate benchmark for a non-discrimination obligation. Indeed, the peculiar wording of the remedy suggests that Google negotiated the language specifically to leave the door open to self-preferencing.
The Commission seems to have missed a golden opportunity to apply the Digital Era Report’s recommendations to take a broader, ecosystem-based based approach to the assessment of acquisitions by multi-product and –service gatekeeper platforms. For example, the Commission did not consider whether the transaction may lead to the current smartphone OS duopoly being extended into the currently diverse wrist-worn wearable space or to Google leveraging its “very prominent datasets” to establish hegemony in the burgeoning digital healthcare sector.
Instead, the Commission took a traditional approach to the competitive assessment of Google/Fitbit, defining markets narrowly in ways that minimized overlaps while failing to examine the transaction’s potential effects in broader ecosystems. Indeed, even on this traditional approach, the Commission refrained from considering plausible market segments that would arguably better match the theories of harm it explored (e.g., a market for wrist-worn wearable devices connecting to Android smartphones and controlled by Google either directly or via their OS).
The Commission’s approach to data-related aspects of the transaction seems especially incomplete. The Commission notes in the online advertising context that Google’s ability to combine Fitbit datasets with diverse datasets could give Google a competitive advantage that others could not duplicate, but it does not explain or quantify this ability. More importantly, the Commission failed to apply a similar analysis in other markets. Dismissing data-related concerns in digital healthcare, the Commission noted the availability of other data sources without discussing their substitutability with the 91 metrics recorded by Fitbit devices. While noting Google’s potential incentive to foreclose competition from Fitbit competitors to increase profits from Fitbit sales, the Commission makes no effort to assess the data-driven incentives to foreclose.
To address the concerns identified by the Commission, Google offered three complex remedies with unusual aspects, including expressly allowing self-preferencing in favour of Fitbit. Time will tell whether this complex and unusual remedy package will suffice.