Platform lobbying: Policy influence strategies and the EU's Digital Services Act

Introduction

Amidst growing public, policy, and scholarly concern about the increasing societal influence of privately-operated platforms, an interdisciplinary body of work has begun mapping how technology firms with a range of business models can exert social, economic, infrastructural, and political power (Cohen, 2019; Cioffi et al., 2022; Van Dijck et al., 2019). Recent years have seen an increase in efforts to broaden, historicise, and contextualise the varying dimensions of platform power (Culpepper & Thelen, 2020; Laurer & Seidl, 2021), and to situate it within broader global and political economic structures (Rikap, 2021; Pfeiffer, 2022; Staab, 2019).

There is an understandable tendency in research on platform power, as this special issue demonstrates, to focus on the new and distinctive characteristics of today’s digital markets and infrastructures and the role that major platform firms play within them. Nevertheless, this focus has arguably led extant scholarship to under-emphasise how platform firms wield classic forms of corporate influence that have long been the subject of research in political science, public policy, and transnational regulation scholarship. Platform companies may be some of the most profitable and wealthiest corporate actors in history (Vaidhyanathan, 2018), but they should still be understood as corporations active within the framework of the global economic and political system (Gorwa, 2019; Dolata & Schrape, 2022).

A core component of the political power of global corporations is their ability to engage in “lobbying”: techniques that exploit their access to policymakers and regulators to minimise the impact and costs of new regulatory frameworks, and shape them towards their preferences and perhaps, away from outcomes that best adhere to the public interest (Klüver et al., 2015). A small body of work has established how technology firms engage in these activities in the US context (Popiel, 2018; Li, 2023). It has examined corporate lobbying at certain contested transnational tech policy junctures, such as the negotiation of the General Data Protection Regulation, or GDPR, in the EU (Atikcan & Chalmers, 2019; Kalyanpur & Newman, 2019; Christou & Rashid, 2021). There has also been some notable work addressing how platform firms are increasingly deploying creative strategies of technologically-facilitated consumer mobilisation that go beyond classic understandings of lobbying (Culpepper & Thelen, 2020). The question is: how do these different strategies fit, conceptually and empirically, within our understanding of the political power of platform companies?

The goal of this article is to contribute to the debate on the power of platform firms through an exploratory analysis of platform policy influence during the Digital Services Act (DSA) negotiations. The DSA, initially proposed in the European Union (EU) in 2019 and signed into law in 2022, is the EU’s major reform of its legacy intermediary liability regime, and is a regulatory framework intended to set new transparency, due process, and accountability rules shaping how large platforms moderate the content shared to those platforms by their users (Barthelemy & Penfrat, 2020; van Hoboken et al., 2023). Through a synthesis of existing interdisciplinary work on business power, corporate lobbying, and lobbying by the tech industry, we theorise policy influence as an important component of political platform power. In the first part of the paper, we develop a five-part typology of corporate policy influence strategies that can be applied to the platform domain — access lobbying, coalition building, stakeholder mobilisation, public relations, and funding. We then draw upon freedom of information requests, the EU Transparency Register, and research from corporate influence watchdog organisations to explore how the diverse mechanisms and strategies of policy influence discussed in the public policy literature functioned within the specific context of the EU’s political institutions and platform regulation landscape during the DSA negotiations. In the concluding discussion, we use the case study insights to develop some forward-looking hypotheses regarding platform power and policy influence in the unique regulatory setting of the European Union.

Conceptualising strategies of policy influence by platform companies

Direct

In political science, the classic response from companies facing potentially costly new rules is “voice” — understood as efforts to influence decision-makers and prevent new regulations from going into effect (Fuchs, 2007; Drezner, 2007). Existing research indicates that, in recent years, platform companies have indeed undertaken a range of similar efforts at policy influence when faced with disadvantageous regulatory proposals (Popiel, 2018, 2022). In the agenda-setting markets of the United States and the European Union, the technology sector has recently become the biggest lobbyist overall — spending more money than classic corporate influence sectors like the fossil fuel, tobacco, pharmaceutical, and defence industries (Bank et al., 2021; Public Citizen, 2021; Corporate Europe Observatory, 2023a).

The key lobbying strategy highlighted in the literature on business power is “inside” or “access”-oriented lobbying. This kind of direct policy influence seeks to obtain direct contact with decision-makers to shape policy outcomes (Beyers, 2004; Weiler & Brändli, 2015). Here, a firm’s employees (or specially hired lobbying consultants, or industry association representatives) meet directly with policy staff, discussing either macro level policy strategy or at the micro level, the details of voting on certain legislative efforts (Dür & Mateo, 2013). The goal of lobbyists is to shape the preferences of key individuals involved in the policy process, ideally trying to turn their perception of the public interest so that it is aligned with industry positions (Li, 2023). Firms can offer “carrots” to policymakers, such as increased investment and jobs in the constituencies that they represent; they can also make threats, such as partial or total exit from a jurisdiction if costly new rules might prompt a firm to relocate or to cut off access to its products (De Bruycker & Beyers, 2019; Bradford, 2019).

Seemingly crucial to the strategy of inside lobbying are the personal connections and networks that facilitate the kinds of close access that could change policymaker preferences and behaviour. Traditional corporate “inside” strategies include “wining and dining” policymakers and their staff, offering campaign contributions to elected officials, engaging in face-to-face policy meetings, writing draft legislation, and engaging staff with direct communication and policy feedback (Klüver, 2013; Coen et al., 2021). To help facilitate these efforts, the world’s largest corporations have long sought to employ former politicians and regulators with both knowledge of, and close contacts to, the people and systems that might be engaging in policy (Blanes i Vidal et al., 2012). The tech industry is no exception, and major firms providing platforms for user-generated content have, from India (Perrigo, 2020) to Ireland (Fitzgerald & McDermott, 2022), actively acquired key former policy staffers and politicians for the jurisdictions they are active in. These examples go beyond just the classic notion of the “revolving door” between certain industries and governments: leading platform firms are pulling in key regulators and policymakers into their orbit not just in the US, but also in the global context.

At the same time, political science literature has emphasised that in many cases — including in particular policy issues with high degrees of regulatory or technical complexity — policymakers may delegate policy development to industry experts, allowing them to directly influence regulations that bind their own commercial activity (Culpepper, 2010). However, when policy issues become more politicised and publicly visible (that is, politically salient), the strategy of ‘quiet politics’ and back-room dealings between industry and government may be difficult to uphold. In such cases, the mere involvement of industry actors may render a given policy initiative illegitimate in the eyes of observers (Kalyanpur & Newman, 2019). In the technology space, some industry actors, knowing that their influence efforts might be perceived poorly in the public eye, have sought to channel their preferences through third-party industry associations, and other intermediaries, allowing greater anonymity in political influence. In the United States, for instance, groups like NetChoice have received large amounts of tech funding to engage in litigation against state-level platform regulation (Poritz et al., 2024).

Another crucial aspect of private business policy influence involves the formation of interest coalitions by firms and associations. One benefit derived from this approach is the pooling of resources and expertise among firms (Hojnacki, 1998; Weiler & Reißmann, 2019). Furthermore, this strategy helps firms to indicate that a wide range of the business and societal actors are united in their issue preferences, which may increase legitimacy of their positions in the eyes of policymakers. Sometimes, coalitions can be secret, instead seeking to pose as non-corporate affiliated “grassroots” civil society actors that can bolster industry positions under a veneer of independence (Walker, 2012). One classic corporate tactic (particularly noteworthy for its core role in the playbook of the natural resource extraction and tobacco industries) involves the creation of “front groups”, which are sponsored by business actors but nevertheless appear as independent, that can be used to provide additional channel through which industry can access policymakers (Apollonio & Bero, 2007).

Indirect

In the literature on business power, all these “inside” forms of influence, directly targeting policymakers, are often complemented by “outside” strategies (Beyers, 2004), which aim to influence policy processes in a more indirect manner. Outside strategies seek to engage a broader range of societal stakeholders and their interests (e.g. the public at large, but also more specifically academics, professional groups, or journalists). These strategies of funding, agenda-setting, issue framing, and public relations seek to affect policy agendas, shift public opinion, and sculpt the boundaries of public debate in a subtler manner (Beyers & De Bruycker, 2013). Corporate actors pursue these tactics to gather public attention and make their preferences seem more popular and legitimate with key public stakeholder groups (Binderkrantz & Rasmussen, 2015).

One important example of such “outside” strategy of policy influence in the platform sector involves business actors actively engaging with scholars and academic organisations — and, in doing so, targeting research that could impact their commercial interests. Overall, existing literature provides rich evidence of business actors sponsoring academic conferences and professorial chairs, commissioning research projects, or engaging in diverse normative, discursive, and epistemic activities shaping academic work (Binderkrantz & Kroyer, 2012; Davidson, 2022). Specifically, platform firms have in certain circles become infamous for efforts to shape public perceptions of their role in politics and society (Goldenfein & Mann, 2023). One recent example of this is academic work in the area of machine learning and artificial intelligence (Abdalla & Abdalla, 2021). At the same time, it is important to note that the political influence exerted by business actors in such cases can be subtle. Through donations (many of which may be “unrestricted gifts” where the firm does not directly have control over final research outputs or research activities conducted with the industry money), platforms have been said to influence the scope of critiques that leading institutions, academics, and organisations are willing to advance publicly (Glaser, 2018; Young et al., 2022), and perhaps even foster relationships of dependency between them and those organisations that can be exploited strategically down the line (Bank et al., 2021). On top of this, platform firms rely on more traditional public-facing strategies (Walker, 2012), including advertising campaigns in print, television, and other kinds of established media.

Another and more novel “outside” strategy that has been highlighted in the recent platform literature involves consumer mobilisation. Culpepper and Thelen (2020) have argued that political platform power is fundamentally about the ability of major technology companies to directly reach their consumers in a sort of user-centric form of outreach and policy mobilisation. Since many platform companies have services that sit in the pockets of billions of users, potentially just a “push notification” away, platforms may choose to mobilise users to build political support. This includes tactics like “call your lawmaker to oppose this law”, click-through screens into their apps, or sending notifications that inform citizens about a proposed regulation directly (Yates, 2023). As with more classic forms of outside lobbying, the goal here is to work through consumers to depress policymaker demand for policy intervention. In this case, however, platforms use their technically-mediated linkages to customers to mobilise them politically as citizens and voters (Culpepper & Thelen, 2020). Locally tethered platform firms which can exploit specific kinds of platform “complementors” — such as application developers, sellers, workers, and others with financial incentives to keep using the platform — have been particularly effective at using platform-oriented mobilisation techniques during key policy episodes (van Doorn, 2020).

The impact of lobbying and other policy influence efforts

Across the two broad categories of direct (policymaker focused) and indirect (other stakeholder-oriented) policy influence efforts we can say that there are at least five ideal typical strategies that platform firms may be expected to deploy when faced with potentially costly new rules during a new regulatory episode (Table 1).

But the question of how to explain the actual impact and efficacy of lobbying and related policy influence strategies is far less settled. While civil society, especially watchdogs of corporate lobbying activity are inclined to be particularly critical of corporate efforts to get exert their preferences when new policies are being developed, the work of these groups understandably has a sort of selection bias: their reports and publications focus on controversial moments where corporations seem to have gotten their interests across, rather than on moments where these lobbying efforts fail or are rebuffed by policymakers. We should not expect firms to always be successful in their policy influence campaigns, as not all regulatory episodes are the same. Rather than conceiving of industry as always all-powerful, we should expect some influence strategies to work better in certain contexts and under certain conditions. Platform firms are not all the same: they can be larger or smaller, less or more embedded within domestic economies, and have different business models and preferences on different policy issues (Srnicek, 2016; Tallberg et al., 2024).

The political science literature in particular offers a range of insights regarding the potential success of business lobbying in different institutional contexts and under varying local and issue-oriented conditions. Firstly, political economists have argued that the impact of corporate policy influence efforts will be mediated by structural factors relating to the specific firm, including characteristics such as wealth, size, and specific exposure to regulatory frameworks (Bernhagen & Mitchell, 2009). Naturally, a very large company with substantial financial resources will be able to spend more on both inside and outside lobbying (Blanes i Vidal et al., 2012). Financial resources can be used to get privileged access to regulators and lawmakers, “capture” them through political donations, or pull them out of public service into lobbying. Wealthy companies can also launch expensive PR campaigns, commission research, and engage in other efforts to shape public debate (Dur & Mateo, 2013).

But the resources of business actors are not the sole factor. Scholarship on business power has long emphasised the importance of the specific institutional and political-economic context on lobbying success. For instance, firms with a large domestic economic presence (contributing significantly to domestic economic activity through employment and other means) can often leverage this as an asset in accessing influential policymakers (Drezner, 2007; Mikler, 2018). In the case of the EU’s political system, these kinds of power resources of firms will nevertheless be mediated by the complexities of the bloc’s multi-level governance system (Klüver, 2013); in general, some EU lobbying scholarship has argued that the EU’s structure tends itself towards compromises between business and government positions, rather than the “winner-take-all” outcomes in the US (Mahoney, 2008). Furthermore, given that most large platform firms are multinational companies with their global headquarters outside of the EU, they may face a disadvantage as “foreign” firms with less structural embeddedness (contributions to domestic GDP, to employment) in domestic European economies than either European tech players or competing firms in adjacent sectors (e.g. the traditional media ecosystem) that see platforms as potential threats to their business model (Flew et al., 2021; Nielsen & Ganter, 2022).

Finally, important is also a specific configuration of interests at play on a particular policy issue. This includes both specific policymaker preferences at the varying levels of lawmaking (Perarnaud, 2021; Heerman, 2023) as well as the varying configurations of industry interests. Rather than treating “big tech” and the platform industry as a monolith, we must consider the varying potential ways in which platform interests diverge across firms, including due to different business models, growth strategies, and potentially even headquarters/national allegiances. More enterprise-oriented firms like Microsoft often have different positions on tech policy issues than consumer-oriented (and user-generated content oriented) firms like Facebook (Gorwa & Peez, 2018). Smaller firms reliant on the chokepoints of larger platforms have been shown to lobby actively for new local rules that would benefit them in the long-term (Gilpin & Doctorow, 2022). Traditional business sectors (e.g. German manufacturing firms) have been shown to be important actors in shaping European tech policy through their preferences as well; their eventual preference alignment with US-headquartered cloud players is another notable interesting example that has significantly impacted the European Commission’s “digitally sovereign” cloud infrastructure projects (Lechowski & Krzywdzinski, 2022; Kemmerling & Trampusch, 2022).

Platform policy influence and the DSA

Shortly after taking up her position as the president of the European Commission in late 2019, Ursula von der Leyen announced an ambitious new strategy for a “Europe fit for the digital age”. The flagship measures of this tech policy agenda were two interlinked regulatory proposals, the Digital Services Act (DSA), which would new transparency and due process obligations on online intermediaries for user-generated content, and the Digital Markets Act (DMA), which aimed more broadly to intervene in a range of different platform-based business models through competition policy tools. Together, as the “Digital Services Act Package”, these two policies proposed significant new rules that would need to be followed by virtually all businesses offering online services in the EU, with special, stricter rules for the largest players.

In June 2020, the Commission initiated a three-month consultation process on the DSA Package. Following this, in December 2020, the Commission published a draft text of the proposal, which kicked off a complex multi-level negotiation process. For more than a year, the text was discussed, debated, and amended in the European Parliament’s Internal Market and Consumer Protection (IMCO) committee, with additional opinions and inputs from seven additional Parliamentary committees. Simultaneously, the Council of the European Union was developing its own position, which it published in November 2021. After about six months of negotiations, in April 2022, the Parliament and Council reached a provisional political agreement on the DSA. The final text of the proposal was passed by the European Parliament in June 2022, and formally came into effect in January 2024.

This long and distributed process of negotiation makes it difficult to make confident assessments of the political influence of certain stakeholder groups on the regulatory outcome. The DSA consultation alone featured responses from 2,863 different stakeholders, which gives an indication of the wide range of potential interests at play.¹ Civil society, industry, and various government actors — different parts of the Commission, different parties in the Parliament, different member states — should be expected to have had complex and nuanced policy positions that will have evolved over the period of the negotiations. There has been a growing body of work seeking to approximate actor policy preferences on certain related policy issues through analysis of their public statements or via various techniques of assumption. Nevertheless, even if the preferences of all actors involved during the DSA negotiation were somehow known in a longitudinal fashion, it would be difficult to definitively establish that changes to those preferences were linked to specific events or interventions from industry (i.e. specific meetings or specific lobbying strategies). This task is made especially difficult by the general opacity and inherent “closed door” nature of these types of policy negotiations, and the particularly important role that the highly opaque trilogue negotiations play at the very end of the EU policy-making process.

In the following empirical case study, we draw upon a combination of primary and secondary data to explore the involvement of platform companies in the regulatory process around the DSA, drawing on our typology of five platform policy influence strategies. In particular, the analysis uses the following data sources. First, we draw upon the research published by transparency watchdog NGOs active in the EU, such as Corporate Europe Observatory and LobbyFacts, as well as upon the media coverage of established media outlets such as Politico and Euractiv. These organisations closely monitor business lobbying activities in Brussels and EU member-state capitals, offering valuable insights into policy processes — even if their interpretations regarding actual influence of specific business actors in specific cases need to be approached with caution and verified through additional evidence. Second, we draw upon the EU Transparency Register and internal documents collected through Freedom of Information Requests from the European Commission and EU member state governments. These documents — such as emails, meeting minutes, memos, and other preparatory materials — provide detailed insights into interactions between policy makers and industry actors. In terms of timeframe of the analysis, we focused on reconstructing platform lobbying efforts during the main period of negotiations between the announcement of the DSA in 2019 and its eventual entry into force in fall 2022.

Access lobbying

In the period between 2019 and 2023, all of the major internationally active platform businesses maintained a significant lobbying presence in the EU policy-making landscape. Based on the data from the EU Transparency Register, Google and Meta/Facebook took the top two spots as the companies with the highest number of declared official lobbying meetings.² As the DSA negotiations unfolded, these meetings occurred at various levels — from the EU Parliament and member state governments to various parts of the EU Commission, including top-ranking officials.

In October 2020, Google’s lobbying strategy received an unusual level of public visibility when a leaked document laying out the company’s strategy to shape the DSA was published in the conservative French weekly Le Point (Berretta & Grallet, 2020). The document outlined the plan for direct policymaker influence at various EU institutional levels, including comprehensive lobbying efforts directed executives in the European Commission, key groups in the European Parliament, national member-state governments, as well as efforts to mobilise third-parties to help reshape the public narrative around the DSA (Corporate Europe Observatory, 2020). The document also described potential inside lobbying strategies that Google could deploy to create pressure against Commissioner Thierry Breton, an individual with a key role in shaping the von der Leyen Commission’s tech policy agenda, and outlined ways to create conflict between different policymaking departments of the Commission (Corporate Europe Observatory, 2020). Mentioned in it were also ideas for how the US government could potentially be leveraged to benefit Google’s position in negotiations with the EU policymakers — another classic tactic. American firms also have a long track record of mobilising their “home state” to lobby on their behalf in particularly consequential policy negotiations (Bradford, 2019), although their efforts to actually do this in the digital policy space have been markedly limited (Gorwa, 2024).

An overview of which firms were able to obtain the highest-level meetings with EU Commissioners provides another relative assessment of their influence and of the Commission’s prioritisation of different stakeholder voices (Lechowski & Gorwa, 2022). For instance, in the lead-up to the DSA Package’s consultation period, Margrethe Vestager took personal meetings with Google (January 20), Facebook (February 17), Microsoft (January 20), and the technology investment consortium Prosus (February 10). Vestager also met with major European publishers (the European Newspapers Publishers’ Association, and European Magazine Media Association, both on February 24), the Dutch travel and accommodation platform Booking.com (April 20), and the association of Nordic public broadcasters (Nordvision, April 4). HERE Technologies, a small firm trying to “foster European tech excellence” was also able to obtain a meeting with the EU’s “digital czar”.³

Documents obtained via freedom of information requests can provide some additional insight into how firm status, size, and strategic linkages to policymakers help certain firms get access to policymakers. For instance, a set of emails released by the Commission in response to requests for information about DSA-related lobbying meetings includes a note from the European Holiday Home Association (EHHA), a Copenhagen-headquartered lobbying group advocating for “the short term rental industry”. EHHA’s efforts to meet with Executive Vice-President Vestager during the DSA/DMA consultation window are politely rebuffed “due to agenda constraints”, and they are offered a meeting with a lower ranking member of Vestager’s cabinet instead.⁴ The senior policymakers in the Commission made time for the largest firms, however. Google staff disclosed 33 meetings with the Commission in 2020, including high-level meetings with the Commissioner executives Thierry Breton (4x), Margarethe Vestager (1x), Frans Timmermans (1x), Věra Jourová (7x), Didier Reynders (3x) and Nicolas Schmit (2x), and Mariya Gabriel (1x).⁵ Commission President von der Leyen seems to have met personally with only one technology firm in the lead up to the DSA proposal being published: Apple (on January 22, 2020, at Davos).⁶

The DSA proposals led to an overall increase in the staffing and expenditures of most large platform firms. Meta appears to have invested the most heavily in 2022, when the company deployed 17 full-time equivalent lobbyists and committed 8 million Euros to its EU lobbying efforts (excluding representatives in member states; see LobbyFacts, 2024b). Since 2020, Amazon has increased its expenditures from around €1.9 million to over €2.7 million (in 2022), making the firm one of the twenty largest direct lobby spenders in the EU for the first time. In the case of Google, however, a different process seems to be ongoing: while in the 2010s, during the negotiations of the GDPR, the company’s lobbying expenditure was increasing rapidly, this growth has stagnated since then (LobbyFacts, 2024a). This, however, does not seem to have impacted Google’s ability to get meetings with top EU officials. And given the high salience of the EU’s regulatory efforts for Google’s business model, it seems likely that the company’s lobbying expenditure has not simply diminished, but also been re-routed to other less direct strategies.

Rather than coming out explicitly against the DSA package, many firm consultations on the package seem to have been framed more delicately, applauding the Commission for seeking to harmonise rules and foster legal certainty.⁷ Access lobbying appears to have been focused on the most potentially costly provisions of the DSA Package –— such as mandatory interoperability, or potential rules that could prevent the ongoing use of “surveillance advertising” — tracking-oriented ad delivery done through ad auctions and processes like real-time bidding. These arguments were key for industry’s access-oriented influence efforts during the Parliament’s negotiation phase; they also worked behind the scenes during the Trilogues to try and influence the position of key member states, as exhibited in a 2022 meeting between chancellor Olaf Scholz and Amazon CEO Andy Jassy.⁸

Coalition building

A scene-setting document prepared for Vestager by her team ahead of her meeting with Facebook/Meta in early 2020 lays out some of the Commission’s key arguments around the DSA Package and their vision for its eventual success. In response to potential questions from the company about why the DSA or other reforms to the EU intermediary liability regime may be necessary, the document states the policy position that the DSA will “reinforce the single market for digital services and help provide smaller businesses with the legal clarity and level playing field they need”.⁹ This policy position, and the Commission’s signalled intent to create differentiated rules for platform companies of different sizes in the DSA and DMA, became a key rhetorical battleground for industry.

In particular, they sought to exploit the notion that a major issue for European member states was the lack of robust European competitiveness in the digital economy. Although the continent had few mega-platform unicorns, it had plenty of small and medium enterprises, or SMEs, that theoretically one day could maybe grow to challenge the dominant platforms under the right conditions (Bobier et al., 2022; Smit et al., 2022), something that was being promoted through an increasingly active European industrial policy and attempts to enshrine European “digital sovereignty” (Schneider, 2023; Lechowski & Krzywdzinski, 2022). In an illustrative public statement, Amazon European Public Policy director James Waterworth (Amazon Public Relations Team, 2021) emphasised the importance of EU regulation of consumer protection and the support of small businesses that sold via the Amazon platform.

Facebook, Amazon, Google, as well as Apple, Microsoft, and other firms also devoted significant resources to building coalitions and funding front groups well positioned to advance this specific “SMEs against the DSA” agenda. Allied for Startups provides an illustrative example. It appears to have been created in 2014 by a young American energy lobbyist.¹⁰ There is little reporting about this small organisation, although an archive of the group’s website in 2015 suggests that it was governed by a four member “continental council” that featured a mix of startup incubators and startup advocacy organisations. The organisation would organise annual convenings in Brussels and events with European policymakers (LobbyFacts, 2024b).

But in 2020, the organisation suddenly jumped into action around the DSA, as illustrated by ten press releases relating to the group’s activities on the new digital policy proposals being planned by the European Commission.¹¹ Fueled by its new corporate board of funders — Amazon, Apple, Microsoft, Meta, and Google — the organisation commissioned research looking at the impact of the DSA. One notable study argued against key provisions of the DSA, including both limits on targeted advertising as well as rumoured features of the DSA/DMA like special duties for the largest firms. Moreover, the organisation met with commission officials and began hosting a wide range of events (Stolton, 2020). Some related strategies were put in play by SME Connect, a network “dedicated to SMEs and their supporters” which received funding from Amazon, Alibaba, Meta, Google, and Uber through its “Friends of SME” program (SME Connect, n.d.).

Because these types of organisations do not need to publicly disclose their financing, transparency groups expressed concern that this front group strategy was exploiting a loophole in the EU’s lobbying transparency rules. Allied for Startups was only one of several pro-startup organisations that would become active on the DSA and DMA, and the “SMEs against the DSA” strategy was also being leveraged by large industry associations in meetings with the Commission and the European Parliament. During the Parliament’s consultations on the DSA text, several MEPs went to the media complaining that multiple industry associations were ‘‘impersonating SMEs while failing to be open about the big tech firms that led or bankrolled their organisations’’ (Rankin, 2023, n.p.).

An analysis of the DSA negotiation process shows that a few other transnational coalitions became active in terms of direct outreach to policymakers. For instance, in November 2020 Spotify policy staff reached out to the von der Leyen cabinet to share a new organisation they had launched called the “Coalition of App Fairness”, an international set of trade associations and businesses that rely on app stores to reach consumers (“with a focus on Apple”).¹² The coalition brought together some notable European tech firms (Protonmail, Deezer) with the legacy power of the European publishers (European Publishers Council, News Media Europe), as well as some notable Apple-critical players from the US (like Epic Games). These smaller players provide an example of alternative, medium tech coalitions actively lobbying the Commission for stronger rules that would improve their positions vis-à-vis the larger gatekeepers, rather than seeking to water down rules or head off certain provisions in the DSA package.

Shaping public opinion

Interestingly, although the platform industry sought to cultivate ties to policymakers and build coalitions that could push back against certain arguments around the DSA Package, it did not attempt to influence this key policy debate through strategies that sought to leverage consumer-oriented aspects of platform power. In contrast to what might be expected following the emerging literature on user-oriented mobilisation efforts (Culpepper & Thelen, 2020; Yates, 2023), Meta, Google, Amazon, TikTok, Apple, X, and other firms did not send push notifications to European consumers, display banners or other menus declaiming the potential harms of impending European regulation to their users in the EU, or otherwise seek to channel their connection to the public to influence the elite policy debate on the DSA/DMA.

Instead, the industry spread its argument that the DSA would create economic harms to the European economy in the time-honoured way: through public relations campaigns directed at the public, undertaken through traditional media channels. For example, Amazon was busy during the DSA and DMA negotiations seeking to manage its public image, seeking in particular to push back against mounting public reports as well as regulatory inquiries into poor working conditions in the company’s logistics operations, and allegations that its platform down ranked third-party sellers and copied their most popular products via its in-house AmazonBasics brand (van Dorpe, 2019). The company spent more than 8 million Euro for a print campaign in German newspapers in 2021 alone — more than double its inside lobbying expenditures in Brussels for that year (see Duffy, 2022; LobbyFacts, 2024c).

Similarly, Facebook ran almost 7 million Euro of newspaper ads related to the DSA package in Germany in the first 9 months of 2021, an estimated 2.5 million of which emphasised how small businesses relied on Facebook pages to reach their customers (Corporate Europe Observatory, 2022). Google appears to have also run a small media campaign in Germany on “ZukunftHandel” (the future of retail), arguing that the firm’s products help, rather than hurt, small retail shops in the dying city centre high streets.¹³

Some of the most intensive public relations campaigns seemed to have not targeted the general public, however, but rather specific constituencies of tech policy experts actively involved in the DSA debate. Meta and Amazon appear to have run targeted ads on Facebook and Twitter that would be served to lookalike audiences built from the followings of the critical Politico journalists Laura Kayali and Mark Scott (Corporate Europe Observatory, 2022), as well as individuals in Belgium profiled by the platforms to have “interests” in “politics and current events” and “government policy”.¹⁴ Rather than seeking to persuade these audiences more generally that the platform company in question was a generally positive political actor, instead the companies appear to have made narrower arguments that bolstered some of their core lines of argument against the DSA — for instance, that “more than 50% of small businesses surveyed that used Facebook apps said doing so helped them attract new customers in the pandemic”, or that “Amazon has helped many small businesses prosper”.¹⁵

To strengthen their epistemic position and make arguments about the potential negative implications of certain provisions in the DSA Package, industry also deployed the classic lobbying technique of directly funding research and organising events relating to the proposal. A prominent example can be seen in the Google-funded study published in mid-2020 by the European Center for International Political Economy (ECIPE), which sought to model the economic impacts of the (at that point speculative) DSA. Although the actual draft text of the DSA had not yet been published by the Commission, the paper speculated that the DSA might lead to ‘‘a loss of about 85 billion EUR in GDP and 101 billion Euro in lost consumer welfare’’ (Narayanan & Lee-Makyiama, 2020, p. 2). This figure, and argument of prospective economic harm that was echoed consistently in 2020, as representatives from many of the major platforms spoke on panels in Brussels discussing the new digital policy’s outlook (Corporate Europe Observatory, 2020).

Discussion

The Digital Services Act is a high-stakes regulation which promises not only to have a significant impact on the operations of the largest platform businesses, but which may also potentially serve as a sort of template for other jurisdictions like the GDPR has done (Bradford, 2023). As substantial reporting and civil society observations have shown, platform companies from all jurisdictions have been involved to varying degrees in trying to shape the regulation and minimise the eventual impact that it has on their business models.

Empirically researching lobbying and other policy influence activities presents significant challenges. Part of this relates to the issue of observability: much activity happens outside the public domain, with lobbying happening “inside” policymaker offices in closed-door meetings. When lobbying registers exist, as they do in the EU, there are often weak enforcement mechanisms ensuring that firms actually disclose all of their relevant meetings promptly and accurately (Alter-EU, 2009), so researchers and civil society inherently are working within certain structural limits to what they can know. Even the use of creative techniques, filing freedom of information access requests, is inherently limited by the ability of bureaucracies to redact the most sensitive (and potentially interesting!) information (Walby & Larsen, 2012).

A second challenge relates to measuring the impact and outcomes of policy influence efforts. To paraphrase a statement often cited in advertising circles, it is possible that “half the money spent on policy influence by platforms is wasted — but do they know which half?” In this research, we have not attempted to empirically assess the potential effects of various targeted digital advertising or print PR campaigns run by specific platform firms on the overall preferences of the elite or public targets in question, or tried to document whether these individuals may then have contributed to the DSA negotiations in some tangible way. There is a staggering potential range of different actor preferences on a policy like the DSA across different firms in different sectors, different types of civil society organisations, different parts of the European Commission, Parliament, Council, and member states, all of whom are actively seeking to shape the final text. We have not sought to systematically map these preferences and causally isolate how they, and thus the DSA’s final form, were shaped through contact with platforms, industry coalitions, or other (potentially tech-mediated or tech-funded) actors.

Such research projects may prove to be impossible: the politics and social phenomena at play here are incredibly complex, and the devil regarding the real “effects” of industry lobbying will likely be in the details of implementation and enforcement. We have provided an exploration of strategies that some major platform companies deployed when seeking to influence the DSA Package, offering an overview of these different tactics and an argument that these strategies need to be examined through an eye towards the local institutional context.

As far as we can tell, consumer-facing influence has not been a major part of the political playbook deployed by leading platform firms seeking to shape the DSA. This observation contrasts with the expectations of some recent work that has highlighted this strategy as the central tactic that the platform industry uses to buttress their power (Culpepper & Thelen, 2020). There do not appear to have been push notifications sent to European citizens via Instagram, Facebook, or Whatsapp declaiming the DSA; we did not find reports of display banners or other forms of targeted user outreach proclaiming the advantages of the web-tracking business model and of services these firms provide to consumers. Instead, we see a more classic business power toolkit, with access lobbying and coalition building appearing to be the core of Amazon, Meta, Google, Spotify, and other firms’ efforts, with some tailored outside strategies sprinkled in on top.

Why might this be the case? Evidently, the European context provides interesting structural wrinkles that make it a different political battleground than the US context. Firstly, there could be a material explanation: do platforms think — or perhaps expect, due to internal polling or other types of research — that their preferences are not fully aligned with their general consumers in Europe? European consumers may feasibly be more pro-privacy, pro-labour, and (perhaps) more anti-platform than those in the US (see e.g. Farrell & Newman, 2019). Have EU citizens been primed against tech arguments by a European policy discourse suffused with other countervailing narratives, such as strategic, economic, and (geo)political arguments that portray “big tech” firms as unaccountable foreign actors, rather than consumer allies (Pohle & Theil, 2020; Lechowski & Krzywdzinski, 2022)?

There could also be an institutional reason why platforms chose not to work through consumers on the DSA Package, instead opting to try to leverage small and medium enterprises as their key allies that could be mobilised against policymaker preferences. The European system is a complex multilevel one, and key policymakers at the European Commission are not directly linked to electoral constituencies in the way that national level elected officials are. Did platforms think that consumers would not be as effective allies on the DSA package? Or perhaps that they wouldn’t be able to claim the support of European consumers in the same way as they can in other jurisdictions, given other factors like the presence of European consumer rights organisations that were actively mobilising against them? More work will be required to help us better understand how different institutional playing fields condition industry interests, their ability to successfully deploy certain policy narratives, and meaningfully build coalitions with consumers and other policy stakeholders (Seidl, 2020). We hope that further research will continue to build on the emerging literature on platform influence, and present us with more empirical insight into the conditions under which these strategies are more or less effective at helping platforms achieve their political goals.