Court case: Deutsche Telekom vs. Meta: Should Big Tech pay for the massive traffic they send to broadband networks?

Today’s internet access business models don’t support the underlying cost of broadband infrastructure. When the internet started, there was no video. Traffic between internet service providers was symmetric, and the killer app was email.

Things could not be more different today; traffic is asymmetric, disproportional, and overwhelmingly video. The internet ecosystem lacks more than $2 trillion for connectivity. New business models are needed to recover broadband infrastructure costs and ensure continued investment and innovation in networks.

Overview of the case: Deutsche Telekom vs. Meta

A lawsuit brought by Deutsche Telekom (DT) against Meta’s German subsidiary was recently ruled in DT’s favor, obliging Meta to pay €20 million in contract fees. The case offers helpful insight to the growing global discussion on broadband cost recovery, an opportunity to review theory and practice of two-sided markets, and how parties could exploit market power and asymmetric regulation. More largely, the case sheds rare and needed light on the murky unregulated world of interconnection and suggests that greater attention is needed. Meta is expected to appeal to the Düsseldorf Higher Regional Court. Strand Consult explores the case with this note as part of its Global Project on Business Models for Broadband Cost Recovery and offers a larger library of information on the topic, including a ground-breaking report about similar case brought by Netflix against SK Broadband in South Korea.

German media reported that DT contracted with Meta to dedicate 24 private interconnection points with 50 ports and 5,000 gigabits/s data rates at 7 locations for the exclusive use of the Meta services Facebook, Instagram and WhatsApp. Meta paid a bandwidth-dependent fee of around €5.8 million euros per year for the service. The contract began in 2010, but Meta requested reduce price after a decade. It asked for 40 percent discount; DT countered 16 percent.

Meta terminated the transit agreement at the end of 2020, effective March 1, 2021. DT confirmed the termination and offered to keep the ports open “for the benefit of consumers “ until a new agreement could be reached. A series of email exchanges over the following months suggests that DT billed Meta as before. Meta did not pay, claiming it would use the dedicated infrastructure without fee under “settlement-free peering.”

DT thus sued Meta in December 2022. The Bonn Regional Court concluded that it was not competent to rule and referred the case to the Cologne Regional Court without a hearing in Feb 2023. 

Meta’s defense was that free transit is the global standard, that it only serves the data that DT’s customers request, and that request for payment to terminate the traffic to end users is an abuse of market position.

The court did not see it that way. Contrary to the Meta’s s view, it is irrelevant “whether the interconnection constitutes ‘peering’ and not ‘transit’ from a technical point of view,” wrote the court. “The data was sent in the same way as before the termination.”

Indeed, settlement-free peering is predicated on mutual agreement and exchange of relatively equal of traffic. Upon cancelling the contract, Meta effectively demanded premium treatment of its data for free; whereas consumers only purchase best efforts access to the public internet. DT said it uses settlement free peering on a contractual basis with predefined criteria.

More largely the charge of dominant position cuts both ways; DT may have large market share in Germany for broadband, but this pales to Meta’s dominance in social media and mobile communications.  The companies are “mutually dependent”. Meta’s market power “countervails” DT, noted the decision.

Indeed, the case follows a dispute concluded in the Supreme Court of South Korea in December 2023 which Meta won against the Korean telecom regulator. The case demonstrates that Meta has multiple options in how to route its traffic to the broadband provider. The regulator charged that Meta knowlingly made a decision to harm its customers by choosing a roundabout route and reducing the speed of its traffic 2.5-4.5 times. The court upheld that Meta has a right to exercise different routing options. Translated to the German situation, if Meta was not satisfied with DT’s offer, it could avail itself to best efforts service.

The economics of the parties

Two-sided markets literature suggest that DT and Meta have incentives to maximize end user experience, and indeed the 2010 contract is a testament to that. DT provides two distinct sets of services to two different markets. For end users, DT offers access technology at the user location, account management, customer service and so on. To Meta, DT offers a dedicated set of infrastructure (hardware and software) in multiple locations and access to a large customer base.

Indeed, Meta has profited handsomely during the period, accordingly to its financial statements. Its average revenue per user in Germany increased ten-fold from the time it negotiated the agreement. Meanwhile DT’s mobile ARPU has fallen to $10 per month. At this rate, it will not be long until Meta’s ARPU exceeds DT’s.

Moreover, the period reflects a time when DT made major upgrades from 4G to 5G and investment in fiber, enhancements that enable Meta’s applications for photos, videos, live streams, games, podcasts and so on. Meta’s payment of €5.8 million per year amounts to €0.33 per Facebook user on DT’s network. The fee is a pittance, and it’s hard to see how such modest payments can close the EU infrastructure gap of €200 billion.

The case included some technical points, for example that Meta didn’t need all the infrastructure resources which were provisioned and thus wanted a discount. However, if that was the case, it’s odd why Meta only renegotiated after 10 years.  Engineers could debate the optimal level of provision for peak traffic, but this does to diminish that provisioning network resources is not free.

The point remains that it is unfair to raise the price on all DT customers to accommodate the needs of Meta. That DT and Meta should contract to provision the network is utterly reasonable and logical. This was also the essence of the case of Netflix and SKB in Korea.

Notably the DT-Meta contract stipulated that the parties would resolve disputes in court. In a free market, parties could block, withhold, and degrade content to extract rents. However, Meta has the advantage in that it can withhold payment because of asymmetric rules on DT. A broadband provider risks a public relations/regulatory firestorm even with the perception that it could violate the no blocking/no throttling rules of the EU open internet law (so-called net neutrality, though the term does not appear in statute).

However, Meta is not bound by any such conditions and can block, throttle, degrade, and/or prioritize its content at will and without disclosure.  Hence it is likely that Meta calculated that it could use the pandemic and net neutrality rules to its advantage and drag out the process, hoping that DT would give up.

That we can learn anything about such private agreements owes to the fact that DT sued and endured years without a customer payment. Most operators can’t. Powerless in the face of Big Tech entities like Meta, most operators never attempt to negotiate in the first place, even though they have similar cost challenges.

The case intimates that the market for interconnection is touch and go, and parties play fast and loose with uncodified terms like peering, transit, public, and private. It appears that for the massive amount of internet traffic from a handful of Big Tech entities, there are but a handful of contracts with only the largest broadband providers. Fees are de minimis versus Big Tech revenues and can’t even be found on financial statements.

Big Tech’s assertion of the sweetness and light of “settlement-free peering” is rather a market failure. The fact remains that tens of thousands of broadband providers globally have no contracts at all. These providers do not exchange equal amounts of traffic with Big Tech companies; rather they only receive large amounts of Big Tech traffic. In a competitive market, such exchanges would be subject to formal contracts. The de facto status quo is that Big Tech need not pay or negotiate for the use of other networks, even though they enjoy disproportional traffic.

Policy observations

Regulators are catching on to Big Tech exploiting the veil of ignorance on net neutrality to free ride on broadband networks. The Brazilian telecom regulator observes as much a recent proceeding that Big Tech entities being classified as value-added service does not inoculate them from responsibility. The status quo threatens the larger regulatory and social objectives which regulators must deliver.  As such the movement for fair share is an outgrowth of net neutrality, not an opposition.

Naturally policymakers are concerned that regulators and courts would be flooded with complaints if more broadband providers were enabled to contract. However, such outcomes are not inevitable, and much contracting could be automated.  Software already facilitates millions, if not billions, of contracts for data and network usage in cloud computing.

Nevertheless, there could be advantages to regulatory intermediaries playing a role to recover costs from Big Tech and other digital service enterprises, particularly to recover cost for public spending on universal service and affordability. As such, different regions may pursue different solutions or even a blend of methods.

The case demonstrates the persistent double standard which Big Tech enjoys and suggests that net neutrality rules should count for all players or none at all.  In the meantime, courts remain backstop for justice if your enterprise has the resources.

Learn more.

For more than 25 years, Strand Consult has researched and published reports and analyses on the economics of broadband and telecommunications networks, chronicling the rise and evolution of network technologies, their business models, and regulation. Strand Consult studies these challenges around the world and assembles this information to for its audience.

Strand Consult launched its Global Project for Business Models for Broadband Cost Recovery to inform the policy development with evidence-based assessments, policy research, and transparency efforts. This information helps operators and policymakers understand the problem at local and global levels and evaluate the pros and cons of different solutions.

Effective broadband policy ensures that all people can access the internet and that networks evolve to serve a wide range of services. The Covid pandemic increased the urgency for universal broadband to enable remote work, education, healthcare, and other essential service. More largely, the internet increasingly drives the economy and productivity and the delivery of government services. However, without modernized policy and business models, gaps in broadband investment, rollout and affordability will persist.

Strand Consult’s Global Project for Business Models for Broadband Cost Recovery provides a library of valuable reports and research notes as well as education and training for policymakers about this critical challenge and opportunity. Contact Strand Consult to learn more.