Quick Mail

Please enter your name

Please type your message


Your captcha code looks wrong

Rashmin Sanghvi & Associates

Chartered Accountants

220, 2nd Floor, Arun Chambers,
Tardeo Road,
Mumbai - 400 034,
Maharashtra, India.

Landline : 02249727947

Mobile : +91 91364 23722

Fax : (+91 22) 2351 5275.

Email : [email protected]

 
Home Articles Taxation         Share :

E-Commerce Committee Report proposing Equalization Levy


Section 7 - Issues related to Value of data & User Activity in Multidimensional Business Models


7.1 Contribution of data to profitability of business


77. The Report on Action 1 (2015) lists “Reliance on data, including in particular the use of so-called big data” as one of the significant characteristics of digital economy. 48 What distinguishes the digital economy from the traditional businesses is its ability to gather large amount of data and to process and exploit this data for furthering the ends of business and generating more value, particularly by tailoring product offerings based on this data. There are various methods employed by digital platforms for collection of remote data. This may range from direct online surveys seeking specific responses from users to indirect methods like analysis of user location and online behavior. Increasingly, most of this data is generated by the users themselves, and used by enterprises in a manner that significantly contributes to the profits arising in their business models. These aspects have been examined in detail in the BEPS Report on Action 1.


78. As per the report, big data is created when the scale of data is beyond the ordinary methods of collection and is not amenable to analysis using typical database management tools. Value is created because companies are able to target their advertisements and offer goods and services according to consumer preference after analyzing the data. The use of data for improving business efficiency is not unique to digital economy, but what characterizes its use in digital economy is the massive scale at which this is undertaken. Multisided business models create value by using the data gathered from the use of a product or service by a user as an input either in improving existing products and services or in providing products and services to another group of customers. These aspects are analyzed in detail in paragraphs 137, 138 and 145 of the Report, as reproduced below:


“137. Online advertising involves a number of players, including web publishers, who agree to integrate advertisements into their online content in exchange for compensation, advertisers, who produce advertisements to be displayed in the web publisher’s content and advertising network intermediaries, who connect web publishers with advertisers seeking to reach an online audience. Advertising network intermediaries include a range of players, including search engines, media companies, and technology vendors. These networks are supported by data exchanges, marketplaces in which advertisers bid for access to data about customers that has been collected through tracking and tracing of users’ online activities. These data can be analysed, combined, and processed by specialist data analysers into a user profile.
138. In advertising-based business models, publishers of content are frequently willing to offer free or subsidised services to consumers in order to ensure a large enough audience to attract advertisers. The most successful advertising companies have been those that combine a large user base with sophisticated algorithms to collect, analyse, and process user data in order to allow targeted advertisements. While traditional advertising involved payment for display of ads for a specified period of time, with little way to monitor visibility or user response, online advertising has given rise to a number of new payment calculation methods, including cost-per-mille (CPM), in which advertisers pay per thousand displays of their message to users, cost-per-click (CPC), in which advertisers pay only when users click on their advertisements, and cost-per-action (CPA), in which advertisers only pay when a specific action (such as a purchase) is performed by a user.”
“145. In the consumer markets, many cloud services (e.g. email, photo storage, and social networks) have been provided free of charge, with revenue generated through advertising or the sale of data on user behaviour, or on a “freemium” basis in which basic services are provided for free and expanded services require payment. Other consumer cloud services, such as web hosting or hard drive backup, are sold on a monthly subscription basis. In B2B markets, cloud services are most typically sold by subscription, although “pay as you go” models are increasingly available.”


7.2 Growing Significance of Data


79. With better servers, greater computing power and the expanding internet of things, the volume of data and its significance is likely to keep on increasing. BEPS Report on Action 1 (2014) quotes the Data Driven Marketing Institute which found in its report that Data Driven Marketing Economy (DDME) added USD 156 billion to the US economy in 2012. This data can be classified three ways. Firstly, collected data, whereby data entered by a user is tracked; secondly, submitted data, i.e. data that is specifically entered by a user, e.g. on a search engine; thirdly, inferred data, data that is compiled via pooling together various strands of data from a variety of sources.49


80. Websites can collect an extensive personal profile of users within a short span of browsing time. Information such as location, specific address, name, email address and phone number is obtainable. Companies are also interested in knowing about specific shopping habits, and which keywords are used to find their site and whether or not you were interested in advertisements on their pages. Internet tools that make this possible are IP addresses, Web browser cookies, e-tags and image files called Web beacons or Web bugs. A particular type of third-party ad-serving cookie, monitors the web browsing of users to show them advertisements relating to their interests.


81. Social networking websites gather large amounts of personal information about users, including ages, friends and interests, from account signing up forms as well as from browsing habits on their sites. Some of it is collected without users being aware of it. Companies are thus able to gather data about the location of visitors, what sites users have visited, what they have shopped for, preferred modes of payment, etc. From this they can infer other personal details, such as their income, self-owned or rented house and so on. These issues find mention in paragraphs 165-166 of the 2015 Report, as under:


“165. Data can include both personalised data and data that is not personalised, and can be obtained in a number of ways. In the case of personal data, as mentioned in Chapter 3 (3.1.5 Use of data), it can be obtained directly from customers (for example, when registering for an online service), observed (for example, by recording Internet browsing preferences, location data, etc.), or inferred based on analysis in combination with other data. It is estimated that sources such as online or mobile financial transactions, social media traffic, and GPS co-ordinates generate in excess of 2.5 exabytes (billions of gigabytes) of data every day (World Economic Forum, 2012). The dividing line between personal and non-personal data is not always clear; however, as data obtained from multiple private and public sources will frequently be combined in order to create value. A recent study quantifies the value of the Data-Driven Marketing Economy (DDME) and looks at the revenues generated for the US economy. The study found that the DDME added USD 156 billion in revenue to the United States economy in 2012 and notes that the real value of data is in its application and exchange across the DDME (Data-Driven Marketing Institute, 2013).
166. Although the use of data to improve products and services is not unique to the digital economy, the massive use of data has been facilitated by an increase in computing power and storage capacity and a decrease in data storage cost, as shown in Figures 4.9 and 4.10, which has greatly increased the ability to collect, store, and analyse data at a greater distance and in greater quantities than was possible before. The capacity to collect and analyse data is rapidly increasing as the number of sensors embedded in devices that are networked to computing resources increases. For example, while traditional data collection for utility companies was limited to yearly measurement, coupled with random samplings throughout the year, smart metering could increase that measurement rate to 15 minute samples, a 35 000 time increase in the amount of data collected (OECD, 2013). This has manifested itself in particular in the concept of “big data”, meaning datasets large enough that they cannot be managed or analysed using typical database management tools. Data analytics, defined as the use of data storage and process techniques to support decisions, are becoming a driver for innovation in a number of scientific areas and are used increasingly in collaborative and crowd-based projects. In this regard, a text search performed on one of the largest repositories of scientific publications shows that articles related to data mining doubled during the last decade, as shown in Figure 4.11. The value of the ability to obtain and analyse data, and big data in particular, is increasingly well documented by market observers.”


82. Paragraph 168 of the Report (2015), that refers to the findings of the McKinsey Global Institute on ways in which data can create value, is also reproduced below for ease of reference:


“168. The McKinsey Global Institute Report notes five broad ways in which leveraging big data can create value for businesses:
i. Creating transparency by making data more easily accessible in a timely manner to stakeholders with the capacity to use the data.
ii. Managing performance by enabling experimentation to analyse variability in performance and understand its root causes.
iii. Segmenting populations to customise products and services.
iv. Improve decision making by replacing or supporting human decision making with automated algorithms.
v. Improve the development of new business models, products, and services.”


7.3 Significance of User Data in Nexus and Value Creation


83. Thus, the exploitation of user generated data for value creation and use of multisided business models has been identified as one of the broader challenges of the digital economy by the Report on Action 1, which not only give rise to BEPS concerns but also have wider implications for jurisdictional allocation of taxes.


84. As a result, users and customers become an important part of the value creation chain. The logical sequence to this inference is whether the value creation achieved in this manner should be brought to taxation in the locations from where the user data is generated. Coupled intricately with this question would be queries related to measurement of value of user generated data. Action 1 Report acknowledges that this raises fundamental questions such as:


“whether the current international tax framework continues to be appropriate to deal with the changes brought about by the digital economy and the business models that it makes possible, and also relate to the allocation of taxing rights between source and residence jurisdictions. These challenges also raise questions regarding the paradigm used to determine where economic activities are carried out and value is created for tax purposes, which is based on an analysis of the functions performed, assets used and risks assumed. At the same time, when these challenges create opportunities for achieving double non-taxation, for example due to the lack of nexus in the market country under current rules coupled with lack of taxation in the jurisdiction of the income recipient and of that of the ultimate parent company, they also generate BEPS issues.”.


It goes on to say this:


“The expanding role of data raises questions about whether current nexus rules continue to be appropriate or whether any profits attributable to the remote gathering of data by an enterprise should be taxable in the State from which the data is gathered, as well as questions about whether data is being appropriately characterised and valued for tax purposes…… If remote collection of data gives rise to nexus (or in the case of an existing taxable presence) what impact this would have on the application of transfer pricing and profit attribution principles, which in turn require an analysis of the functions performed, assets used and risks assumed.”


85. While companies in the business of collection and analysis of data have evolved means of valuing data (one such means being customer lifetime value, CLV), the valuation of data from a taxation and nexus point of view has additional challenges. What has been clearly acknowledged now is that user generated data and its valuation, should be taken into account for deciding allocation rights to tax jurisdictions.


7.4 Issues related to User activities and contribution


86. Another unique feature of many business models is their ability to get value created by users and their contributions. Paragraphs 266-67 of the BEPS Report on Action 1 (2015) introduces these issues as under:


“266. Additional challenges are presented by the increasing prominence in the digital economy of multi-sided business models. A key feature of two-sided business models is that the ability of a company to attract one group of customers often depends on the company’s ability to attract a second group of customers or users. For example, a company may develop valuable services, which it offers to companies and individuals for free or at a price below the cost of providing the service, in order to build a user base and to collect data from those companies and individuals. This data can then be used by the business to generate revenues by selling services to a second group of customers interested in the data itself or in access to the first group. For example, in the context of internet advertising data collected from a group of users or customers can be used to offer a second group of customers the opportunity to tailor advertisements based on those data. Where the two groups of customers are spread among multiple countries, challenges arise regarding the issue of nexus mentioned above and in determining the appropriate allocation of profits among those countries. Questions may also arise about the appropriate characterisation of transactions involving data, including assessing the extent to which data and transactions based on data exchange can be considered free goods or barter transactions, and how they should be treated for tax and accounting purposes. However, as discussed more generally above, the location of advertising customers and the location of users are frequently aligned in practice, such that the value of the user data is reflected in the advertising revenue generated in a country. The scale of this challenge may, in addition, be mitigated by the fact that advertising will frequently require a local presence to attract advertisers."
267. The changing relationship of businesses with users/customers in the digital economy may raise other challenges as well. The current tax rules for allocating income among different parts of the same MNE require an analysis of functions performed, assets used, and risks assumed. This raises questions in relation to some digital economy business models where part of the value creation may lie in the contributions of users or customers in a jurisdiction. As noted above, the increased importance of users/customers therefore relates to the core question of how to determine where economic activities are carried out and value is created for income tax purposes.”


87. The role of ‘users’ who neither charge the enterprise for their contributions (such as personal data, content created by them or network benefits brought to the enterprise from their presence), nor pay to the enterprise for accessing the digital / virtual platforms owned by it, pose an unprecedented challenge for taxation of profits contributed by their contributions. These multi-dimensional business models can be viewed as a combination of two simultaneous transactions – for instance, a multidimensional business model of online advertising has two limbs, one between the enterprise and the user where the enterprise provides access to the users in lieu of their contributions in the form of their personal data and content created by them, and the other between the enterprise and its customers, for services that involve display of customized advertisements shown to the users based on their personal data collected from them. In such a business model, one limb may be more like barter, while the other involves exchange of financial payments. User contributions in the form of data, content and networking benefits act as a substitute for services that an enterprise could obtain from individuals by paying them wages or contracted amounts, with the right to access and use the digital network substituting for salary or contract payments. These issues have been analyzed in the BEPS Report on Action 1 (2015) in paragraphs 258-259 as under:


“258. Similarly, users of a participative networked platform contribute user-created content, with the result that the value of the platform to existing users is enhanced as new users join and contribute. In most cases, the users are not directly remunerated for the content they contribute, although the business may monetise that content via advertising revenues (as described in relation to multi-sided business models below), subscription sales, or licensing of content to third parties. Alternatively, the value generated by user contributions may be reflected in the value of business itself, which is monetised via the sale price when the business is sold by its owners. Concerns that the changing nature of customer and user interaction allows greater participation in the economic life of countries without physical presence are further exacerbated in markets in which customer choices compounded by network effects have resulted in a monopoly or oligopoly.
259. These various developments must be understood in light of their relationship to more traditional ways of doing business. For example, while having a market in a country is clearly valuable to a seller, this condition by itself has not created a taxing right in the area of direct taxation to this point. It is also true that data about markets and about customers has always been a source of value for businesses as illustrated by phenomena such as frequent flyer programs, loyalty programs, the creation and sale of customer lists, and marketing surveys (in which customers participate for no remuneration), to name a few. The traditional economy also benefited from “network” effects in ways that are perhaps less obvious than the network effect present in social networks. Sellers of fax machines, for example, were dependent on a sufficiently broad supplier of purchasers in order to ensure that their product had value. The digital economy has, however, enabled access to markets with less reliance on physical presence than in the past. In addition, the digital economy has enabled collection and analysis of data at unprecedented levels, and has enhanced the impact of customer and user participation in the market, as well as the degree of network effects. It has been suggested that the lower marginal costs in digital businesses coupled with increased network effects generated by higher levels of user participation may justify a change in tax policy. See, e.g., Crémer (2015); Pistone and Hongler (2015). In considering policy changes to reflect customer interactions to the imposition of income tax, however, potential impact on traditional ways of doing business must be taken into account in order to maintain coherence in cross border tax policy. In addition, consideration should be given both to solutions based on income tax and to solutions focused on indirect taxes.”


88. In this context, the questions that come to fore are the extent to which nexus gets created by the activities of users, and the value of the contributions made by the users to the profitability of the enterprise. Paragraph 280 of the Report (2015) concludes that presence of users and their contribution can be indicative of significant participation of an enterprise in the economic life of a tax jurisdiction, as under:


“7.6.1.3 User-based factors 280. Given the importance of network effects in the digital economy, the user base and the associated data input may also be important indicators of a purposeful and sustained interaction with the economy of another country. A range of factors based on users could be used to reflect the level of participation in the economic life of a country, namely:
· Monthly active users (MAU). One factor reflecting the level of penetration in a country’s economy is the number of “monthly active users” (MAU) on the digital platform that are habitually resident in a given country in a taxable year. The term MAU refers to registered user who logged in and visited a company’s digital platform in the 30-day period ending on the date of measurement. A factor based on MAU presents the advantage of measuring the customer/user base in a given country both in terms of size and level of engagement. Given that little material is publicly available on the process of defining and identifying MAU, more detailed metrics would need to be developed in consultation with businesses and IT experts for the purpose of using this factor, such as how to identify a unique "user" or what level of engagement is required for a user to be considered "active". Reliability and veracity of the information would also need to be ensured, to address fraudulent accounts, multiple accounts, false information volunteered by users, and “bot”-produced data, to name a few.
· Data collected. Another factor which could be considered to reflect an enterprise’s level of participation in the economic life of a country is the volume of digital content collected through a digital platform from users and customers habitually resident in that country in a taxable year. The focus would be on the origin of the data collected, irrespective of where that data is subsequently stored and processed (e.g. data warehouse). The range of data captured by the threshold would not be confined to personal data, but would cover also, e.g. user created content, product reviews, and search histories. This core element could be coupled with proportionality tests, such as whether the volume of digital content collected exceeds a percentage of the enterprise’s overall stored digital content. Information on data collected is increasingly available, reliable and upto- date, especially if the factor is focusing on data collected that is effectively stored by the non-resident enterprise on a server. At the same time, businesses may not necessarily maintain separate and comprehensive track records of the volume of data collected and stored on a country-by-country basis. In addition, the volume of data collected (and stored) from users in a country may not necessarily reflect an effective contribution to the profits generated by the nonresident enterprise, as the value of raw data is rather uncertain and particularly volatile.”


7.5 Committee’s Observations


89. The Committee, after taking cognizance of these observations and detailed analysis provided in the BEPS Report on Action 1, and after analyzing the role and contribution made by the users by way of data, content creation and the networking benefits, considers that users are a significant indicator of both nexus and creation of value in the jurisdiction of source. In the view of the Committee, the presence of users of a digital or telecommunication network in a multi-dimensional business model signifies value creation and economic participation in that tax jurisdiction, and should give rise to the threshold nexus for taxing that enterprise in that jurisdiction, particularly, when such user contribution is relied upon for earning income from that jurisdiction. The Committee, however, notes that at this stage, quantifying such value creation can be a challenging task, and therefore considers that a simple tax rule that broadly covers such value or a significant part of it, without creating the difficulties associated with quantification of such value, may be preferable.



48. Paragraph 151 of the BEPS Report on Action 1 (2015)

49. Working paper on Digital economy by Expert Group on Taxation of Digital Economy of the European Commission, March, 2014



Back  |  Next