|Issue:||Asia-Pacific III 2008|
|Topic:||Learning from how the Internet is used|
|Title:||CEO and co-founder of|
Klaus Mochalski is CEO and co-founder of ipoque, a vendor of deep packet inspection solutions for Internet traffic management. Prior to founding ipoque in Leipzig, and after three years as an IT consultant, Mr Mochalski worked as a researcher in computer networking at the University of Waikato in Hamilton, New Zealand, at the University of Leipzig in Germany and at the San Diego Supercomputer Centre in the United States. He holds a master’s degree in computer science from the University of Leipzig.
The Web generates only ten to thirty per cent of Internet traffic, while 20 per cent of the Internet users generate between 49 and 83 per cent of the traffic with peer-to-peer (P2P) transfers. Today, Kazaa, eDonkey and BitTorrent exchange high-volume content such as movies, music and software between users. P2P has been making a shambles of the traditional recording industry, but with more acceptable pricing structures, high quality and less digital rights management, recording companies should be able to re-take control of the market.
It is a common misconception that the World Wide Web is the Internet, but there is more than meets the eye of the Web user. In fact, most of today’s Internet traffic does not result from Web browsing. To understand what is going on inside one of the biggest and most complex systems of the world and how both its builders and users will shape its future, it is crucial to get the complete picture. The advent of the Web in 1993 marked an important turning point in the history of the, then, nearly 25 years old Internet by suddenly bringing it into the public domain. The Web quickly grew in popularity and became the Internet’s main tributary. This only changed when peer-to-peer (P2P) file-sharing networks entered the stage. It started in 1999 with the music exchange platform Napster that, after its legally enforced demise, was quickly followed by a number of contenders for the Internet bandwidth throne. These mostly proprietary systems – prominent examples being Kazaa, eDonkey and BitTorrent – have been designed to efficiently exchange high-volume content such as movies, music and software between Internet users – without needing central servers. In recent years, P2P networks have been responsible for between 40 and over 80 per cent of all Internet traffic, while Web browsing has accounted for about 10 to 30 per cent. The wide ranges indicate that the matter is quite complicated. A 2007 Internet study analysed the Internet traffic in five regions of the world between August and September 2007. Comprehensive statistics about user behaviour provide a unique overview of the Internet’s current state. The study includes data about Internet telephony (VoIP), Skype, video streaming, instant messaging (IM), file hosting and encrypted P2P protocols. Three petabytes of anonymous data representing over one million users in Australia, Eastern Europe, Germany, the Middle East and Southern Europe have been analysed. The results for these different regions vary considerably. The results include these interesting facts: • Less than 20 per cent of all Internet users participate in P2P file sharing; • P2P produces between 49 and 83 per cent of all Internet traffic – with night-time peaks of over 95 per cent; • About 20 per cent of P2P traffic is encrypted; • BitTorrent is the most popular P2P protocol, followed by eDonkey; • Videos, particularly movies, are the most popular P2P content, not just by volume but also by file count, followed by music; • 10 to 20 per cent of all Internet users are using file hosters such as RapidShare for file sharing; and, • Skype is by far the most popular Internet telephony (i.e. Voice over IP, VoIP) service with a share of over 95 per cent of all VoIP traffic. The study showed there is great interest in current popular content. Movies and music accounted for the highest download volumes. Many of the top titles downloaded through P2P networks reflect chart positions. This proves that users indeed want high-quality content. Users invest considerable effort finding what they want, and pay for high-speed Internet access. Lessons for content providers For nearly a decade, the film and music industries have struggled to adapt their business models to accommodate new opportunities brought by the Internet. So far, they have mostly been reacting to Internet piracy threats, mostly from P2P file sharing and unusual competition – a former computer company called Apple being the most prominent example. Instead of simply fighting their foes, they should take a very close look and learn from them. Not only to ‘know their enemies’, but to use their own weapons against them. The rules of the game are set, and they are quite simple. First, make the access easy and attractive. Just look at popular piracy sites such as mininova.org and thepiratebay.org. Their massive and always current content is painstakingly organised in categories and subcategories to quickly find a desired file. The content industry has much better raw material. They not only have the original music or movie with multi-language support, but also all the additional information that is included to make DVDs more attractive. So they should be able to do better than the pirates. Second, get the pricing right. Sure, the zero price tag of pirated material is hard to beat, but this is not what the majority of customers expect. People have always paid for media content, and would continue to do so if premium content was available through easily accessible distribution channels at prices that do not make them feel as though they were ripped off. Look at the business model of RapidShare, the major file hosting service. It charges a monthly fee of seven Euros for a premium account with faster and unrestricted access to its content, which is not provided by RapidShare itself, but by its users. This simple model with no access to actual content generates substantial profits for this company. Add legal, higher quality access to a similarly broad spectrum of content and many Internet users would flock to such a service, even at a slightly higher price of, say, somewhere between 15 and 40 Euros. Also, the music and film industries should stop calculating their losses by multiplying the number of illegal downloads by the official price of a CD or DVD. Nobody buys tens or hundreds of discs per month. Instead, they should try to tap into the market potential out there, which they once knew quite well. This is where they should have started years ago entering the Internet world. Third, and as with RapidShare, there should be no strings attached, like poorly implemented copy protection in the form of euphemistically named Digital Rights Management (DRM). When customers buy a title, they want to own it and decide how and where to watch or listen to it – as they always have with physical media. There will always be a small part of the population that will go to great lengths to avoid paying for content. It is important to understand that these are most likely not potential customers. All the rest will prefer to avoid the hassle and get their material from a legitimate source – if the first two conditions are met. Lessons for service providers The Internet study shows that less than 20 per cent of the Internet users generate most of its traffic. For a service provider using a flat-rate tariff model, this is a challenging proposition – challenging because both the heavy hitters and the normal customers require a satisfactory service, i.e. a high quality of experience (QoE). The first group’s excessive use of resources can have an adverse impact on overall network performance. The challenge is to find the right balance between the two groups – without upsetting either one of them. How easily this can happen became apparent during the last months when it leaked to the public that the American ISP Comcast was slowing down certain file-sharing activities based on BitTorrent. A particularly unfortunate aspect of this endeavour was the use of a technology that interferes with TCP, a traditional Internet transport protocol, in a non-standard way. In response, this has already triggered much blog post on how to circumvent this kind of bandwidth management by further moving away from the TCP standard. If this marks the beginning of an arms race, it could well have an unforeseeable negative effect on the stability of the Internet as a whole. The important lesson: do not tamper with your customers’ traffic without them knowing. Interestingly, the opposite has become common practice. Instead, providers should offer managed services at different prices with QoE guarantees for customer-specific applications. For instance, many customers would like a tariff with guaranteed Skype call quality – and most would pay extra for it. Similarly, a clean P2P service without any copyright infringements would be valuable particularly for business customers. Today, though, most service providers and carriers still have a long way to go. They are running their daily business blindfolded. This was apparent to us during the data collection for our Internet studies. There was an incredible lack of information about subscriber behaviour in the providers’ networks. At best, there was a few days’ worth of measurements made on a yearly, or longer, basis. These measurements provided snapshots and, at least, allowed for a rough long-term trend observation. Providers need a much more detailed picture of their network’s traffic to prepare for the challenges of new applications, and to plan the future of their business. Insight is really the first step towards control.