Consent-based video optimisation

Enabling a new role for mobile operators in the video value chain
A Disruptive Analysis thought-leadership paper
Commissioned by Yospace

February 2012

Author: Dean Bubley Contact: [email protected]

Disruptive Analysis Ltd, February 2012

Consent-Based Mobile Video Optimisation

Disruptive Analysis
Introduction

Dont Assume

For mobile operators, video is both a threat and an opportunity. Its use drives the purchase of data-plans by smartphone and tablet owners, but its growing popularity threatens to clog up cellular networks with extra traffic. It is possible to reduce the impact of Internet video fairly quickly, by using techniques such as optimisation to compress video streams in the network, or pace their delivery, which reduces some traffic load. But this can alienate content companies, which dont like to see their material being altered - especially by operators who then try to sell them enhanced QoS. It can also create issues where customers have paid for HD-quality content, which the network downgrades without permission or notification. In addition, few implementations only squeeze down video just when there is actual network congestion, which could be seen as reasonable by regulators most attempt to optimise across-the-board. Partly linked to this, content publishers and aggregators are increasingly now using applications or new video formats which self-optimise to counter network congestion or unavailability. Techniques such as adaptive bitrate streaming are becoming popular, as are various forms of encryption and DRM (Digital Rights Management). These approaches may well also have the side-effect of bypassing the optimisation boxes or conflicting with them. The net effect is antagonism between stakeholders, rather than cooperation. At the same time as this so called OTT (over-the-top) story is playing out, some operators are thinking once more about developing their own video offerings. Despite indifferent success with mobile TV in the past, there are some signs of renewed interest in developing channels or video-based apps to provide content both to their own subscribers, and possibly to mobile users on other networks. Certain of the constraints that stopped mobile TV from gaining success in the past have been removed, and different use-cases and are taking shape. If operators can make their new video offerings inherently more network-friendly than other sources, then they can develop new business models and revenue sources while also providing their mobile data customers with an all-round better mobile video experience. Depending on the operator, it may also form the basis of future partnerships with content publishers and advertisers to create more integrated and interactive experiences, or paid-for enhanced delivery. Done properly, it should also be able to cope with the broadening universe of new smartphones and device types, while requiring minimal intervention to existing billing or network policy infrastructure. This paper examines some of the key trends in mobile video the tensions between the players, the realities of new video formats and delivery mechanisms, and the shifting roles of network operators and the traffic-management strategies required to run their mobile broadband networks. Note: This document has been commissioned by Yospace, but represents the independent & consistent viewpoint of Disruptive Analysis (as at time of publication). Disruptive Analysis has retained full editorial control over the content and stance.
Disruptive Analysis Ltd, February 2012 Consent-Based Mobile Video Optimisation 2

Disruptive Analysis
Network congestion & pressure points

Dont Assume

The rising traffic levels on 3G and 4G networks are well-documented. Mobile operators frequently report high rates of growth in data volumes, while studies and forecasts from network vendors such as Bytemobile, Ericsson and Cisco highlight the role of video content and forecast huge future growth. While some of this data can be challenged (especially the predictions, given recent introductions of capped dataplans), there is nevertheless a real problem in some networks, and an imminent issue in others. That said, there is a broad diversity of both causes of network congestion and its potential mitigation. Some issues come as a result of signalling traffic from short data bursts or connections, while others stem from large Internet video downloads. Some problems relate to capacity limits in the radio network, others are driven by backhaul or core network constraints. And despite the efforts of many network vendors trying to sell extra intelligence such as application-detection and throttling/QoS measures, most of the real impact on lowering network load has come from billing and charging. Encouraging end-users to change their behaviour for economic reasons, rather than using network smarts to do it for them, seems to have been effective. Part of this reflects the difficulty of really understanding applications in the network: categorisations of data into broad buckets such as video, web and P2P are often arbitrary, overlooking mashups and combinations of services. Video is not an application per se but a component of many. A mobile web page may include a video-based content clip or advertisement, a web conferencing system might include telepresence, and an aircraft maintenance application might incorporate video-based instructions and examples. Given the strategic importance of mobile data, it is unsurprising that numerous vendors, across the value chain, have touted approaches to fixing the problem of traffic management, often with video as a specific focal point. There is no one size fits all method for treating mobile video traffic each of these has its own specific requirements. Some will be encrypted, some will be tagged to prevent caching, some will be streamed and some will use adaptive bitrate approaches. Increasingly, strategic combinations of these approaches will be used, rather than individual tactics. So it is important that solutions fix the right problem, in the right way and ideally also support current and incremental revenues for the operator.

Non-consensual optimisation
At the moment, many mobile operators around the world have implemented so-called transparent content and video optimisation. Typically, this involves a dedicated box sitting between the operators core network and the Internet itself on what is called the Gi interface. Sometimes, the optimiser is just a function of operators gateway node itself (the GGSN). In this scenario, the optimisation acts as a proxy between the user and the original content source on the Internet. It takes in content (in this instance video) on the northbound side, processes it, and then sends it on to the end user. The user is

Disruptive Analysis Ltd, February 2012

Consent-Based Mobile Video Optimisation

Disruptive Analysis

Dont Assume

typically unaware that he or she is seeing the optimised content rather than the source itself. There are various functions that mobile video optimisation boxes can perform. They can transcode data, for example from one codec to another. They can compress it, in various ways (eg dropping resolution, or frame-rate), delay it, or potentially even modify it for example by inserting advertisements. Each vendor has its own specific use cases and subtleties, but the general story is similar: (Realistically) attempt to reduce network congestion by sending less video traffic to the user than might otherwise have occurred (Potentially) reduce costs of upgrade or opex needed for the radio and/or backhaul network (Hopefully) improve user experience by reducing or mitigating the incidences of stalling (Optimistically) offer the operator a chance to monetise (i.e. charge) the OTT video content providers for sending streams over the network

It is important to point out that this form of optimisation does indeed reduce the overall amount of traffic on the network, so it is by that definition successful. However, the actual effects on perceived quality are relatively unpredictable especially to the original content provider. The user may end up watching a lowerquality stream so that buffering pauses are reduced but the video publisher does not know this, and neither do its advertisers. The user may have actually paid for an HD stream, and been prepared to wait for it to download into the buffer before watching it yet the network may have assumed it was being viewed in real-time and unnecessarily reduced the quality. Disruptive Analysis believes that this form of non-consensual optimisation has serious flaws. Ideally, both the user and the content publisher will have agreed to what is being done to the data flow (after all, the user is paying for it), and they will also be able to see what impact that optimisation has on the experience. It is interesting to note that various of the optimisation vendors are now evolving their approach, to differentiate amongst themselves and add value to their boxes, and to better fit with the realities of the video and content world. Many have developed additional use cases for their kit, such as attempting to measure and control video stalling by observing the behaviour of TCP/IP connections. Others vendors have been discussing ways to monetise third-party video content charging video publishers for enhanced QoE or prioritisation, for example. As yet, such notions have seemed superficially attractive, but it is not clear that the business-model promises match up to the practicalities of the technology.

Searching for new video business models

So, it is undoubtedly true that most of the volume traffic on mobile 3G/4G networks is video-based, although on some networks the bulk of it is destined for laptops with 3G modems or perhaps tablets, rather than smartphones. However, such traffic typically derives from Internet-based applications such as TV companies direct-to-user smartphone apps, or other players such as Facebook and YouTube. A growing amount of video is driven by communications rather than content, such as through apps like Skype.
Disruptive Analysis Ltd, February 2012 Consent-Based Mobile Video Optimisation 4

Disruptive Analysis

Dont Assume

Numerous suggestions have emerged for mobile video pricing and business models in recent years, with operators (and their vendors) hoping to derive payments from users, content providers, advertisers and other sources. Although many of these concepts seem elegant at first glance, they often hide complex technical or organisational difficulties in implementation. Among the options considered have been: Capped and tiered data plans, which are not video-specific but just include a bundle of MB/GB quota for the user to consume as they wish. Zero-rating of certain traffic from particular applications or sources. This could either allow an operator to package data in innovative bundles (1GB a month, plus extra free YouTube), or partner with specific companies to bundle the data transport into another price plan. ($x per month for mobile Netflix, and the video doesnt count against your data plan!). So-called personalised data plans which allow / disallow certain types of application. (Social network data plan 10 per month, Full web access (exc video & VoIP) 20, Full web + video 30) Quality-of-service guarantees or prioritisation, charged either to the user (2Mbit/s data plan with 3 hours HD-quality TurboVideo where available) or to the content provider / publisher (Gold content delivery service with Priority QoS) Deployment of competitive on-net CDNs (content delivery networks) that rival existing global players such as Akamai, allowing improved download speeds and lower latencies, with operators deriving revenues from the content firms directly. Hijacking of mobile video streams, with the insertion of additional advertising sold by the operator. Creation of network-aware push video channels which pre-cache content on the users device, when there is no cellular congestion, or when WiFi is available. Development of mobile advertising networks by the operators, using customer information to assist in the targeting of users, or measurement of activity. Bundled connectivity for mobile video / entertainment devices. Consider a video version of the Amazon Kindle as an archetype, where the operator has concluded a wholesale deal with the device/content vendor, and the end customer doesnt need a separate contract.

Some of these approaches are complex because of the many layers of technology integration needed for example, prioritising traffic ideally requires links right down into the radio network and base stations, something which adds hugely to the difficulty of the solution. Others need careful work to fit with billing or customer-case systems, or thoughts about various scenarios such as WiFi offload. Yet more face legal minefields around contentious issues such as Net Neutrality. Another model is starting to re-emerge as well: the re-establishment of operator-run mobile video applications or channels.

Disruptive Analysis Ltd, February 2012

Consent-Based Mobile Video Optimisation

Disruptive Analysis
Operator-owned content? Again?

Dont Assume

Conventional wisdom suggests that mobile operators arent particularly good at creating, sourcing or aggregating their own content. Early attempts at mobile TV largely fell flat outside of Japan and Korea (remember Europes attempts with DVBH?), while mobile music has typically been dominated by companies such as Apples iTunes or Spotify. There have been some minor successes in areas such as usergenerated content (for example, 3s original SeeMeTV) but these have largely been steam-rollered by the YouTube juggernaut in recent years. The original difficulties experienced by MNOs can be traced to diverse factors, such as: The need for specialist infrastructure and devices for non-cellular broadcast networks like DVB. Rise of flat-rate (or very large) mobile data plans, coupled with easy-to-use 3G USB dongles and smartphones, meaning that a wide choice of open Internet video became feasible on ordinary devices, at exactly the same time that operators more-constrained mobile TV offers became available. Poor fit of monthly subscription-based TV offers with prepaid mobile top-up accounts. Difficulty and expense of creating custom content. Low reach meaning a lack of interest from advertisers. Greater scale effects from open web-based video sites, compared with those just targeting a small sub-set of mobile users belonging to a single operator. Increased ease of pre-loading video content on mobile phones via flash memory, Bluetooth or WiFi.

So taken together, many operators have been left nursing expensive, little-used content portals and under-achieving business units. Users have instead opted for smartphones with generic data-plans and the convenience of video-based apps supplied by Internet or broadcast TV firms. Against that background, it sounds like a return to operator-controlled video content provision seems a mistake: have we not learned lessons from experience? And yet, there is a strong argument to suggest that, perhaps, things are different this time. Maybe some (but perhaps not all) operators can exploit video content, taking advantage of these trends rather than fighting against them. Not only that, but they might also simultaneously solve the problem of network congestion from video-based content. A couple of options in particular stand out: Many fixed/mobile hybrid operators now have sizeable IPTV footprints among their home broadband subscribers. There is a strong argument to mobilise those services Youve got TelcoTV at home, now watch it on your phone for an extra $5 a month! which could also help drive mobile subscriptions to that operators quad-play package. Many operators are starting to develop their own apps businesses, creating smartphone software for a range of phones both for their own subscribers, and in some cases other operators as well. Video players for both their own

Disruptive Analysis Ltd, February 2012

Consent-Based Mobile Video Optimisation

Disruptive Analysis

Dont Assume

and close partners content fit nicely into this model and do not require additional network investment. It may also be reasonable to call such TV streams managed services which could mean different regulatory treatment, such as permitting differential QoS. As with fixed networks, there is a fair argument that Net Neutrality laws do not apply to pure onnet mobile data streams on a broadband connection, that do not transit the public Internet. It is common for ADSL connections to have dedicated capacity set-aside for the telcos own VoIP and IPTV services, in parallel to the public Internet access part of the connection. This means that MNOs might legitimately be able to prioritise their own mobile video streams with higher QoS, and/or lower charges. Recent indications from operators such as Telefonica suggest that this approach is becoming a reality. Around the world, many operators are now starting to mobilise their TV or other video content. Time Warner Cable and Comcast have TV anywhere strategies including mobile app-based video players, SingTel in Asia has spoken about OTT opportunities. Both Orange Group and its UK joint-venture with T-Mobile (Everything Everywhere) have been talking up mobile video delivery in recent months.

Creating differentiation for MNO-owned video

To make operator-driven content work, the telcos need to differentiate themselves from the Internet-based publishers and, potentially, to such a degree that those firms eventually decide to partner to improve user experience, rather than continue operating entirely independently. The way that operators can create a sustainable advantage is if they can, simultaneously create and manage two separate things: Network-optimised content Content-optimised networks

To date, most of the solutions pitched to operators have been in the latter category. The optimisation solutions discussed above, intended for installation in the core network and gateway to the Internet, are good examples of this, but they are not the only approach. There is also: Content delivery networks (CDNs), either the operators own on-net solution or those from a third party like Akamai. Direct content peering, where the operator has direct links to specific publishers. WiFi offload. Radio network and backhaul innovations of various types (eg intercepting video streams during congestion and showing a video busy signal instead). Throttling of specific traffic types, apps or websites, using deep packet inspection boxes for policy enforcement. Overlays of additional network capacity.

Clearly, such solutions generally work independently from the content and application sources (except for direct peering). While that is appealing to network operations and architecture groups that do not want to deal directly with upstream publishers, there is a second important trend.
Disruptive Analysis Ltd, February 2012 Consent-Based Mobile Video Optimisation 7

Disruptive Analysis

Dont Assume

Figure: Multiple options for operators to control mobile video traffic

Source: Disruptive Analysis

Network-optimised content is an umbrella term for ways to reduce the impact of video or other data on the mobile network, or to enhance the performance from the users point of view. There are various approaches here, going right from the device/silicon up to actions by specific publishers and application developers, as well as the architecture of the video player apps themselves. France Telecom is reportedly working with Google to help it create more benign apps. This can involve minimising unnecessary signalling load, managing background tasks and network polling, or by reducing outright data volumes. Many content and device vendors are working on similar initiatives independently either to deal with apparent network congestion when it happens, or to improve battery life and user experience. However, in Disruptive Analysis opinion, the optimum benefit comes when operators and device/content suppliers collaborate and share best-practice, ideally supported by tools such as congestion APIs to identify real-world network problems. Although the MNOs might be wary of disclosing such sensitive-seeming operational data, in reality is measured by various means anyway. In fact, there are already moves in the industry towards this goal, such as the IETF project called Conex1 (Congestion exposure), and the optimisation use-cases suggested by the Alliance for Telecommunications Industry Solutions (ATIS)2. It also goes without saying that operators own video and app businesses need to be at the forefront of creating network-friendly and network-optimised content. One critical element of that is the adoption of adaptive bitrate streaming.

1 2

https://tools.ietf.org/html/draft-ietf-conex-concepts-uses-03 http://www.atis.org/PRESS/pressreleases2011/091311.html
Consent-Based Mobile Video Optimisation 8

Disruptive Analysis Ltd, February 2012

Disruptive Analysis
The rise of adaptive bitrate streaming

Dont Assume

Adaptive bitrate streaming (ABR) is a fast-growing way of delivering video content over both fixed and mobile networks. Although there are different versions, the basic premise is quite simple: Encode a given video stream to play out at different speeds Typical encoding rates are 128kbit/s, 256 kbit/s, 512 kbit/s, 768 kbit/s, 1.25Mbit/s and 2Mbit/s . Divide the video up into chunks typically 2-12 seconds long, depending on the specific protocol. Have the video player measure the flow of inbound chunks to determine the congestion level of the network. If the network is busy, have the player request a lower-resolution chunk next. Revert to higher-resolution chunks when the network speeds up again.

In other words, the video application watches out for how busy the network seems, and adjusts its own behaviour accordingly. With ABR, it becomes essentially selfoptimising, adjusting to the available network bandwidth on a basis determined by the segmentation policy. In the mobile industry, the most popular form of ABR is called HTTP Live Streaming, or HLS. Originally developed by Apple but now beingstandardised3, it uses HTTP files (the main language of the web) to deliver video, rather than a dedicated (and often more expensive) streaming connection. Use of HTTP also means that it is suitable for distribution via CDNs, as well as working with proxy servers or firewalls. The separation between the control aspects of the HLS-like protocols to the binary delivery of the video data means that streams can be manipulated on a per-user basis in a scalable manner for example with inserted advertisements. Microsoft and Adobe also have ABR solutions, but HLS seems to be gaining most relevance to telcos especially as it is now not just in the iPhone / iPad, but also in later versions of Android from 3.0 Honeycomb onwards. It is also being adopted by major content providers such as the BBC, which announced its use for its 3G iPlayer in December 2011.

A new consent-based optimisation approach

Bringing a number of the elements together, mobile operators need to differentiate their content offerings via a combination of optimised networks and content: Content based on adaptive bitrate streaming Congestion-aware network policy management Flexible billing and charging systems On-net or federated off-net CDNs

Note: HLS is in the process of becoming a fully-open standard as MPEG-DASH, an ISO-lef initiative to combine the various vendor-specific protocols into a single unified protocol. Furthermore, Apples implementation has been submitted as an IETF RFC: http://tools.ietf.org/html/pantos-http-live-streaming07 under the name Pantos HTTP Live Streaming. In the meantime is becoming widely adopted on various platforms Disruptive Analysis Ltd, February 2012 Consent-Based Mobile Video Optimisation 9

Disruptive Analysis

Dont Assume

Taken together, this means that operators could create, package and sell video content which is: Network-friendly and congestion-aware, optimised from the radio, right through the backhaul, to transit and interconnection domains. Delivered from the operators own servers and therefore potential exempt from some Net Neutrality concerns for example, potentially being prioritisable in terms of QoS and appropriate for differential charging (eg zero-rated against the customers data quota). Configured to offer maximum quality from the perspective of the publisher given the prevailing constraints on the network.

The last point is important. Since the operator and content publisher / OVP work collaboratively, they are able to define exactly what happens when the network is busy, to give the best end-to-end perceived result. They can use ABR streaming to back-off to a lower throughput, but still deliver the content from the operators servers, based on a pre-agreed and pre-tested lower resolution or frame rate. This is what Disruptive Analysis refers to as Consensual Content Optimisation a significant change of philosophy from todays antagonistic use of transparent (ie non-consensual) compression or control of video streams. Some notion of artistic integrity is therefore built in to this process, and importantly the network could also give a notification to the user about what is happening. (The network is busy right now we apologise for the temporary lower video quality and will resume normal service as soon as possible). There is a lot of research on behavioural psychology to suggest that people accept temporary inconvenience as long as they are kept informed. In other words, Network Congestion and Content Quality are jointly optimised, with the user being kept fully informed. This approach could either work for the operators own content, or potentially also be used with partners such as broadcast TV channels or other media outlets. In the longer term, it may even be possible to work with video providers such as YouTube on a collaborative basis. In the longer term, it may also be able to enhance such a proposition with extra features such as exploitation of WiFi offload (if available), or pre-emptive downloads of content during network quiet periods to be cached on the device.

The role of ABR-capable OVPs

However, if operators want to mobilise their existing fixed-IPTV content, they will first need to convert it to mobile-suitable formats, especially if they want to use HLS. This is a relatively complex process, especially as different variants will be needed for the various smartphones and tablets all of which have different screen sizes, playback capabilities and user behaviour. Encoding and storing many versions of high-quality files is difficult and depending on the approach and technology used, can be expensive. There are some other tricky aspects as well, such ensuring the device can effectively synchronise key frames when it switches between the streams. By a similar token, if operators want to offer consensually-optimised video hosting capability to third parties, they will also need to deal with content from a variety of new sources, probably re-formatting (or normalising) it to ensure device compatibility.
Disruptive Analysis Ltd, February 2012 Consent-Based Mobile Video Optimisation 10

Disruptive Analysis

Dont Assume

There is also another element in the mobile video ecosystem Online Video Platform (OVP) providers, which can fulfil a role here. These are companies that work on behalf of video publishers or aggregators, taking raw video output and converting into forms useful for transmission towards the user. In a sense, these are content dispatch networks rather than content delivery networks, facing towards the companies which upload video, rather than users who download it. In this context, the publisher would be the video arm of the telco itself, or its partners. And the CDN could either be the operators own on-net CDN capability, which are increasingly being deployed, or it could be an more traditional Internetbased player such as Akamai or Limelight. In fact, these companies in particular are also offering managed CDN services to operators on a white-labelled basis. An OVP would ingest content on the operators behalf (perhaps from the fixed IPTV business unit), integrate with the relevant CDN, generate the various chunks of video used for streaming, optimise the encoding bitrates, perform content management and tagging tasks, deal with advertising insertion, manage workflow tasks and so forth. It seems probable that mobile operators internal teams would not wish to get embroiled in the minutiae of media publishers businesses and IT systems, so this type of function may be outsourced.

Conclusions
Mobile operators have two problems when it comes to video: How to limit network congestion from video, and thereby minimise the costs of upgrade and operation. How to derive new revenues, either by extending current mobile data business models or developing new ones.

In Disruptive Analysis view, there is only a limited amount of scope by MNOs merely addressing content-optimised networks. Although developments such as CDNs, video optimisation boxes, DPI and the like can improve efficiency and offer some incremental revenue opportunities, that is not enough for meaningful change or sustainable growth. In particular, many of the notions for monetising so-called OTT video by exerting controls, implementing QoS or operating toll-gates dont stand close scrutiny. There are too many exceptions, work-arounds and practical difficulties. Indeed, we already see independent video publishers using adaptive bitrate streaming to deal with the effects of congested networks that is, they are creating network-optimised content, but without the direct involvement of the operators. Operators themselves are the only parties that can potentially blend contentoptimised networks with network-optimised content to create a full end-to-end solution. They need to either create their own content/apps, or work collaboratively with video content publishers for consent-based optimisation, avoiding confrontation and the risk of an arms race in their attempts to manage traffic load on their networks. In particular, operators need to consider once more entering the mobile video market directly, but in ways that exploit new content formats, different business models and
Disruptive Analysis Ltd, February 2012 Consent-Based Mobile Video Optimisation 11

Disruptive Analysis

Dont Assume

on-device intelligence, as well as new network technologies. One of the innovations near the top of the list should be adaptive bitrate streaming, although this brings with it new challenges in terms of content encoding and management. In the past, operators mobile video initiatives focused on developing standalone silo channels or portals. In order to gain greater success and blend the content/network optimisation approach described above they should instead look at mobilising existing IPTV offers, or moving more decisively into their own OTT-style video applications. These could be bi-optimised (network + content format) when onnet for their own subscribers, using CDNs and other network-side intelligence to minimise costs and improve QoE. But they could also be supplied off-net to other telcos subscribers, providing scale economies and incremental reach for advertisers. Over the next few years, Disruptive Analysis expects ABR and consent-based optimisation to grow in popularity (along with telco-run CDNs) as solutions for managing and monetising mobile video.

Figure: Mobile video network/policy solution maturity, 2011-2015

Source: Disruptive Analysis

Disruptive Analysis Ltd, February 2012

Consent-Based Mobile Video Optimisation

12

Disruptive Analysis
About Disruptive Analysis

Dont Assume

Disruptive Analysis is a technology-focused advisory firm focused on the mobile and wireless industry. Founded by experienced analyst Dean Bubley, it provides critical commentary and consulting support to telecoms/IT vendors, operators, regulators, users, investors and intermediaries. Disruptive Analysis focuses on communications and information technology industry trends, particularly in areas with complex value chains, rapid technical/market evolution, or labyrinthine business relationships. Currently, the company is focusing on mobile broadband, operator business models, the Future of Voice, smartphones, Internet/operator ecosystems and the role of governments in next-generation networks. Disruptive Analysis attempts to predict - and validate - the future direction and profit potential of technology markets - based on consideration of many more "angles" than is typical among industry analysts. It takes into account new products and technologies, changing distribution channels, customer trends, investor sentiment and macroeconomic status. Where appropriate, it takes a contrarian stance rather than support consensus or industry momentum. Disruptive Analysis' motto is "Don't Assume". For more detail on Disruptive Analysis publications and consulting / advisory services, please contact [email protected] . For details about Future of Voice workshops & publications, please see www.futureofcomms.com Website: www.disruptive-analysis.com Twitter: @disruptivedean Blog: disruptivewireless.blogspot.com Quora: Dean-Bubley

Intellectual Property Rights / Disclaimer

All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the publisher, Disruptive Analysis Ltd. Every reasonable effort has been made to verify research undertaken during the work on this document. Findings, conclusions and recommendations are based on information gathered in good faith from both primary and secondary sources, whose accuracy it is not always possible to guarantee. Disruptive Analysis Ltd. disclaims all warranties as to the accuracy, completeness or adequacy of such information. As such no liability whatever can be accepted for actions taken based on any information that may subsequently prove to be incorrect. The opinions expressed here are subject to change without notice.

Disruptive Analysis Ltd, February 2012

Consent-Based Mobile Video Optimisation

13