Research Notes

Fact-checking Google’s Matt Brittin at FT-ETNO. Democratizing the Pay As You Go Model from Cloud to Content

At the annual FT-ETNO panel The Future of the EU Internet ecosystem – How should all digital actors contribute to a gigabit future? in Brussels Monday, Google’s EMEA President Matt Brittin debated with Orange CEO Christel Heydemann CEO Orange, incoming Chair to the Body of European Regulators for Electronic Communications (BEREC) Konstantinos Masselos from Greece’s Hellenic Telecommunications and Post Commission, former Vodafone CEO and now Minister of Technological Innovation and Digital Transition Government in Italy. Brittin rejected the notion of broadband fair cost recovery, saying it runs counter to net neutrality.

Note that Brittin defended Google’s tax-avoidance and arbitrage schemes, testifying  to the United Kingdom Public Accounts Committee in 2012 about how the company could generate billions of pounds in revenue but pay little to no tax in the UK. The hearing assessment was that Google is a company with short arms and deep pockets.

Strand Consult has over the last 10 years published many reports and research notes on net neutrality rules around the world. In point of fact, the relevant European Union Parliament law Regulation (EU) 2015/2120 does not contain the term “net neutrality.” Specifically the EU rules prohibit broadband providers from blocking and throttling end users. The Body of European Regulators for Electronic Communications (BEREC) bi-annual reports maintain that there is no problem with net neutrality in the EU today, and that there was not deliberate, systematic violation by broadband providers before rules were adopted.  Nothing in EU policy stops Google from helping to improve connectivity by supporting cost recovery in the broadband middle and last mile. Hence there is no inherent conflict between fair cost recovery and net neutrality.

At the FT-ETNO event Brittin boasted about Google’s infrastructure investment saying, “We carry traffic 99 percent of the way, bringing it closer to users and making it more efficient for our telco partners.” In point of fact, Google has just data centers in 5 strategic EU countries, but there are 22 EU countries where Google finds it uneconomic to invest. Brittin’s claim is false as a matter of distance and cost.

Google may have invested in undersea cables and proprietary data centers, but unlike broadband providers which must deliver content from all comers, Google only invests in infrastructure for its own content. In fact, Google’s infrastructure investment of a few billion dollars pales in comparison to its annual revenue of $257 billion. In contrast, Broadband providers spend 20-30 percent of revenue on infrastructure.

Google largely avoids investing in locations where it would be subject to regulation and social obligations like universal service. It largely avoids use of internet exchanges because it operates a parallel internet. Instead it expects broadband providers to maintain the resources to manage YouTube data (growing by 3.7 million videos per day), regardless of the growth and cost, even though only a fraction of end users access those videos.

Brittin makes a fair point that European operators have yet to make public meaningful traffic and cost data for individual networks. Such a demonstration could change the discussion in EU broadband providers’ favor. Indeed Strand Consult has documented the shortfall for US broadband providers in its report Middle Mile Economics: How streaming video entertainment undermines the business model for broadband which provides case studies of rural fiber to the home (FTTH) networks.

Notably the discussion in the US and other countries is further advanced. In the US, there is a deep discussion on digital equity; some two dozen social justice organizations told the Federal Communications Commission (FCC) that companies like Google must contribute financially to realize shared connectivity goals. Moreover policymakers in South Korea, the world’s leading broadband nation, recognize that content delivery quality is the responsibility of both the broadband provider and content providers like Google.

The global movement for broadband fair cost recovery

An effort is underway by policymakers around the world to address broadband network investment shortfalls. International organizations like the ITU, legislatures in the US and South Korea, and telecom regulators in the US, EU, Japan, and South Korea have initiated long overdue inquiry to the economics and business models of broadband. The internet has changed significantly in the last decade, and it is justified that the issue should be revisited, particularly as streaming video entertainment was not an issue when broadband policy frameworks were adopted.

Importantly these inquiries examine the responsibility of the handful of “Big Streamers” like Netflix, Google’s YouTube, Amazon Prime, Disney+, Microsoft Xbox which comprise some 80 percent of internet traffic. Policymakers want to understand how the Big Streamers should support broadband middle and last mile network costs (to which they pay little to zero today), to provide financial contribution to achieve social goals like universal service (to which they contribute little to zero today), and their role in creating harmful greenhouse gas emissions. 

Democratizing the Pay as You Go (PAYG) model from cloud to content

Presently broadband providers earn subscription fees based on a set of services performed for subscribers which include physical last mile construction and maintenance, network technology and management, R&D, account management, and customer service. There are a different and distinct set of services that a broadband provider must perform to facilitate content from third parties on their network. When content enters a broadband provider’s network, the broadband provider must provide storage, databasing, computation, transfer, migration, networking, delivery, and security. These are the same services that Google Cloud provides for a fee.

Access to Google Cloud is offered on a pay-as-you-go (PAYG) basis. Charges are based on usage or as they are incurred. No one sees anything unfair or discriminatory about this. Individuals and companies purchase Google Cloud services. The shortfalls between broadband and content providers could be resolved with simple cloud service-like interfaces. Indeed South Korean broadband providers have already implemented this for content providers which comprise more than 1 percent of traffic or with 1 million or more users. Fees amount to roughly $20/terabyte ($0.02/GB); that is the same that Google Cloud charges Europeans to send data into its network.

Google has paid such fees to South Korean broadband providers for some time. This has not slowed the rate of adoption of FTTH subscriptions, which have increased for the last three years and now stand at 86.6 percent of total broadband connections, the highest in the OECD. This fact discredits Brittin’s statement that transit fees reduced welfare in South Korea. 

Yes, it may be slightly more expensive for Google to do business in South Korea versus before when network access was free, but Google’s payment helps ensure that middle mile broadband network upgrades occur and supports broadband consumption. Moreover, Google is making more money in South Korea. It reported 2021 sales grew 32.8 percent on year to 292.3 billion won. Operating profit jumped 88.4 percent on year to 29.4 billion won, and net profit 152.1 percent to 15.6 billion won. It appears that the presence of fees has not hurt Google’s profitability, and maybe has even helped it.

Strand Consult offers a deep dive on the situation in its report Netflix v. SK Broadband. The David and Goliath Battle for Broadband Fair Cost Recovery in South Korea.

Order the report.