Failing to define the relevant market

An informative SSNIP Test in a two-sided market requires consideration of reactions from both sides of the market as they can both influence demand-side substitutability. For example, if Tencent starts to charge for QQ services, some users may leave for competing services. The smaller number of QQ users may then lead advertisers to purchase fewer ad spaces, which in turn may or may not make QQ services more attractive to existing and potential QQ users. In general, if the network service effect is reciprocal, ignoring the influence from the other side of the market can result in a relevant market that is too narrowly defined because of the positive feedback loop by which quantity reduction from one side of the market can lead to quantity reduction on the other side as well.

By definition, the SSNIP Test is a process to identify the point for greatest demand reduction when the price increase can no longer offset the loss, resulting from reduction in demand so that the price increase is no longer profitable. In the decision, the Guangdong Court repeatedly brought up the issue that many QQ users may leave QQ if Tencent starts to charge fees, but the simple fact is that regardless of how many users may choose competing services, the result will be a net gain for Tencent (from no revenue to some revenue) if only the user side of the market is addressed. Thus, the benchmark used by the Guangdong Court for its SSNIP Test is in fact meaningless. Even in a subsequent SSNIP Test in which the price is no longer zero, the result from SSNIP Test concerning only one side of the market would not reflect market reality as Tencent's decision to price its QQ services certainly would not be entirely decided by the number of QQ users that may choose competing services.

|

SSNIP Test not always suitable

An important premise for the application of the SSNIP Test is that the benchmark price is the result of sufficient market competition. On the one hand, if the benchmark price already reflects the monopoly profit, the SSNIP Test would not work well in defining the relevant market (a situation commonly known as the cellophane fallacy). On the other hand, if all competitors do not charge for their products or services at all, as the result of either government regulation or industry practice, the SSNIP Test that focuses exclusively on the price increase would not correctly reflect the state of market competition. In such a market, companies compete on the quality and range of their products or services, instead of price setting, in order to satisfy their current consumers or entice potential consumers.

In practice, many web services are offered without cost to the consumers. In the decision, the Guangdong Court recognised that free is the common, basic and only feasible model for providing many services, especially instant messaging services through the internet. As consumers in this market are likely to be very reluctant to pay for any services, the Guangdong Court concluded that, if one company starts to charge a small but significant and non-transitory price for its product or service, consumers will flock to the other that remains free-of-charge, making them part of the same relevant market. But this conclusion seems to deviate from an objective manifestation of the consumers' reaction, making the outcome of the SSNIP Test unreliable.

First, in a market with price homogeneity, the consumers are likely to develop a sense of entitlement to the current price for the products or services and may not be able to accordingly gauge their reaction to an SSNIP Test. In particular, when all products or services are free of charge, the consumers would take for granted that they would not have to pay for the products or services. When polled for their reaction to the price increase in a survey for the SSNIP Test, the consumers' response is a combination of their financial choice and some feeling of betrayal for the supposed violation of industry norm. However, if the price increase does become a reality, the feeling of violation is likely to play a minimal role in the consumers' evaluation of whether to stay with the products or services. That is to say, the result of the SSNIP Test might not be accurate here since it was made by relying on possibly exaggerated reaction from the consumers for the price increase.

Second, in a typical market with price competition, consumers are able to provide their best estimation about their response to a price increase because they make this determination every time they have to choose among competing products or services with different prices. On the contrary, the consumers in a market with price homogeneity are not experienced in how to affix a price tag to any particular product or service that one company may provide compared with its competitors. When presented with a hypothetical question of price increases in the SSNIP Test, the consumers are unlikely to make an informed answer about how they would respond.

|

Reliability depends on evidentiary support

A reliable SSNIP Test requires thorough socioeconomic analysis of the relevant consumers' behaviour in response to the hypothetical price increase. In this case, the Guangdong Court appears to have employed three pieces of evidence in an effort to show the SSNIP Test is an appropriate methodology for defining the scope of the relevant market. In particular, the Court relied on an poll conducted by China Internet Network Information Center (CNNIC) which shows 60.6% of web users are unwilling to pay for instant messaging services, a poll conducted by website eNet which shows that 81.71% of the current QQ users would switch to a competing service if Tencent starts to charge fees for its QQ services, and part of the European Commission's report on Microsoft's acquisition of Skype which cites to an internal Skype document that 75% of Skype users would cease using its service if it starts charging for it.

However, in its detailed analysis of the potential competing products for QQ, the Guangdong Court does not cite any supporting statistics to prove that demand-side substitution is likely with regard to any of the particular products had Tencent started to charge for QQ services. The Guangdong Court also fails to provide a quantitative analysis of how the reduction in usership may impact Tencent financially since the core for an SSNIP test is to examine the company's profitability after the rise in price. These three pieces of evidence are far from supportive of the Guangdong Court's conclusion that Facebook for social networking, any text, audio or video-based instant communication providers for instant messaging, and web service companies such as Baidu and Sina for provision of general online information may be considered as in the same relevant market with QQ.

|

Moving forward

Since its announcement, the Guangdong Court's decision in Qihoo v Tencent has become a seminal case in the development of antimonopoly civil litigation in China. In this decision, the Guangdong Court relied on the SSNIP Test for defining the relevant market and measuring market power. While the SSNIP test is widely accepted for this purpose, its application in the Qihoo v Tencent case is problematic. The Guangdong Court should have taken into consideration the unique characteristics of the market in which Tencent's QQ services compete and adjust the SSNIP test to meet such needs.

Fang Qi, Fangda Partners, Beijing