Following decades of exciting developments in laboratories around the world, pervasive computing technologies are starting to reshape our world in real ways. Computers, visible and invisible, are everywhere: traditional signs and billboards are being replaced by digital signage, projections, and interactive surfaces. Communication is digital and ubiquitous—the majority of the world’s population uses mobile phones. Physical computing and robotics make physical environments digital and interactive. It is apparent that analogous for developments on the Internet, advertising will be the business model driving pervasive computing. These new computing technologies are powerful tools for advertising and they will supersede traditional advertising in the near future. Pervasive computing is the disruptive technology for advertising.
This book presents the core principles of Pervasive Advertising and makes accessible to practitioners and researchers alike research findings in pervasive computing, modern advertising, and human-computer interaction. The major opportunities addressed include: symmetric communication, the long tail, powerful experiences, personalization, context adaptivity, audience measurement, and automated persuasion. The major challenges addressed include: calm and engaging advertising, privacy, and ethical persuasion. This book presents and discusses innovative applications based on digital signage and mobile advertising, venturing into new territories like music and scents.
This book is intended as a guide for:
- Advertisers who want to understand how new technologies revolutionize their field.
- Pervasive computing engineers and researchers who want to understand what business models will drive their technologies in the real world.
- Anyone who is interested how pervasive advertising will shape the future of urban life.
Opportunities of Pervasive Advertising
Pervasive computing is about to change the nature of advertising in major ways. The properties of pervasive computing as being interactive, automated, address the senses, networked, and ubiquitous are reflected in novel properties of advertising when applied to this field. We extract six important properties of Pervasive Advertising compared to traditional advertising. These properties are symmetric communication, the long tail, experiences, personalization and context adaptivity, audience measurement and persuasion.
Power to the People (Symmetric communication)
Classical advertising follows a mass media approach where few advertisers push their advertisements to the masses. This unidirectional communication model causes an asymmetric distribution of power. All the power is with the advertisers who decide which ads to show where and when. The audience has at most the possibility to ignore, protest against, or vandalize the resulting ads. With some people, this asymmetric power distribution leads to the feeling of being at the mercy of advertisers.
As pervasive computing is interactive, networked and automated, it offers the opportunity to transfer a lot of power towards the audience. The unidirectional communication model is fundamentally changed by allowing the audience to express themselves towards advertisers and other audiences. Companies must face customers at eye height. This will be to the benefit to both customers and companies, as both have a much tighter relationship and companies can learn from their customers much faster. Practical examples are the audience being enabled to choose content they like and even submitting their own content. Eventually, this may lead to a democratization of advertising and the appearance of public spaces (see the chapter of van Waart).
Me, too (The long tail)
Pervasive computing is highly automated by definition, and many things that needed manual intervention in classical advertising will become automatic. This significantly lowers cost and effort of individual advertising campaigns. Starting a new campaign may be as easy as filling out a few fields on a website and may cost only a few cents. This price decline enables very small companies and even individuals to launch their own tiny local campaigns. It is important to remember that advertising as the communication of sponsored messages is not only interesting to big companies. Even a small restaurant or market stall must advertise, as must anyone who wants to sell an old bicycle. Even someone who wants to surprise his wife at an airport or make a present for a birthday may be interested in showing something in the environment, as might be someone who wants to impress his friends.
The ‘Wow’ Effect (Experiences)
Pervasive computing offers powerful media that can address all the senses. Large, bright displays surrounding us give powerful visual impressions, but also the hearing, feeling, haptics, and even smelling can be addressed. As Norman (reference) explains, there are three levels to interactive computer systems. The lowest level is visceral, that is, related to the initial appearance. This can be described as “the first impression”. The second level is behavioral, related to the look and feel. It can be described as “how it feels”. The third level is reflective, e.g., what we think others think about us when we use it. As most traditional ads are not interactive, they do not go beyond the first level. Pervasive advertising however has to address all three levels properly. It has a look and feel, and also makes us reflect when we interact with it. These properties make it a much more powerful tool for advertising. As pervasive computing is all around us, these experiences can follow and surprise us wherever we go. Furthermore, pervasive advertising is digital, making very easy to create new experiences all the time. Together, this enables to create a Wow-effect over and over again. Analog posters will look relatively pale compared to the intense and memorable experiences that can be created with pervasive advertising.
Just for me, just for now (Personalization and Context Adaptivity)
Personalization and context adaptivity are at the core of pervasive computing and provide natural powerful tools for advertising. In personalization / user modeling, computers learn models of preferences and behavior of groups or individuals. This fits naturally with the target groups as a core concept of marketing. It is important to remember that in marketing, target groups are used in two different ways, for the development of the product and creation as well as placing of the advertisements. For the development of the product, the dilemma of marketing traditionally was that the properties the target group is defined by have to be measurable, restricting them basically to demographics and other easily measurable criteria. Pervasive computing allows to measure many more things, such that target groups based ,e.g., on actual behavior. The main potential of pervasive advertising however lies in the possibility of measuring all kinds of things in real time, building user profiles, and adapting advertisements on this real live data. This is explained for example in the chapters of Partridge and de Carolis.
In addition to personalization, adaptation to the context becomes much more fine-grained because of automation and better sensors. Traditionally, it was a huge effort to post different signs, e.g. depending on the weather. Using pervasive advertising, however, such things as advertising ice cream when the sun is shining and umbrellas when it is raining become trivial. It can be assumed that when advertisements are much better adapted to the context, e.g., showing products in the context of the hometown and current weather of the audience, they are more effective.
Did you see me? (Audience Measurement)
Audience measurement has always been an integral part of advertising, mainly because “if you cannot measure, you cannot improve”. Any advertising campaign is driven by goals, and one can only set goals on factors that are measurable. Limited measurement capabilities also limit the scope of what one can try to achieve. Pervasive computing provides powerful sensors to measure, e.g., the actual behavior of people. To see the immense opportunities this provides one only needs to have a short look at the web. Because clicks are easily measurable, whole new business models and paradigms have emerged. Ads are often paid for on click-through basis. Campaign success is completely transparent using tools such as Google analytics, enabling advertisers to optimize and cancel campaigns based on live data. Even the fully-automatic optimization of advertising campaigns is possible, using tools such as Google Website Optimizer, which can automatically run statistical tests on user behavior in different versions of campaigns and optimize the campaign accordingly.
Pervasive advertising allows for taking this whole approach to the real world. User behavior, e.g., whether people looked at an advertisement, can easily be measured using computer vision and, e.g., face detection. Tracking when audiences interact with the ad is basically for free, and even such things as eye tracking may soon be ubiquitous. This will allow advertisers to set much more detailed goals (e.g. 50% of bald men between 40 and 60 should have read the first sentence in this text block advertising hair implants), and optimize their campaigns in a rapid loop. If one for example notices that many people frown and turn away after seeing a specific part of an ad, one might try to find out why and change things.
Persuasion
Persuasive technology is defined as “using computers to change what we think and do” (reference). Persuasion is opposed to coercion, because no formal pressure is used. Of course, persuasive technology is not limited to advertising, but has important applications in healthcare and education. They can for example help people quit smoking or persuade them to learn. Fogg distinguishes macrosuasion, where the whole intent of a product or service is to change intentions and behavior, and microsuasion, where persuasion is used to accomplish small steps for a product or service with a different intention. Computers have the benefit over traditional media as being interactive, and over human persuaders as being more persistent, allowing anonymity, processing large amounts of data, using multiple modalities, being scalable and pervasive. Computers can take roles as tools, media, or social actors. As tools, they can persuade through reduction of complexity, tunneling the user into predefined action sequences, tailoring and personalization, suggesting at the opportune moment, simplifying self-monitoring of users, giving users a feeling of being surveilled, and operant conditioning. As a medium, they can persuade through putting the user into simulations. As a social actor, they can persuade through providing positive feedback, social support, acting as a role model, or providing social cues such as being attractive. In order to persuade, it is important that a system is perceived as credible, that is, trustworthy and with expertise.
Although changing attitudes and behavior are not the only objectives of advertising, as for example in advertising one also tries to evoke emotions, it should come as no surprise that persuasive technology is ideally suited for advertising. For example, an advert might try to persuade a customer to buy a specific product over another one, or to build a generic preference for certain kinds of products.
Challenges
Pervasive advertising is coming, no matter whether we like it or not. The fundamental direction is being determined now, and what direction we choose will determine the future appearance of urban spaces for the years to come. This is a crossroads, where one direction may lead into a world full of pervasive spam, people being spied upon, and subconsciously being manipulated to buy things they do not need. However, there is a choice to take the direction into a positive future of advertising. This is a world where advertisements strike a balance between being calm when we do not need them, and being engaging and inspiring when we want to play with them. Our privacy is well protected, and we can inspect our user profiles and change or delete data as we may like. Some ads try to persuade us in a positive way, but unethical persuasion strategies are avoided, used persuasion strategies are overt and we are educated to reflect on persuasion strategies being used. Thus we believe that the three main decisions that must be made are in the areas of calm and engaging advertising, privacy, and persuasion.
Calm vs. Engaging Advertising
The issue of calmness and attention preservation has been at the core of pervasive computing from the very beginning. In the seminal paper on calm computing by Weiser and Brown (Weiser, 1998), it is proposed that when our whole environment is filled with computers, they better remain calm. It was proposed that key to this is that information should slide effortlessly between the center and periphery of our attention. This idea has strongly influenced research for decades, and the underlying paradigm was that systems should remain invisible, predict the requirements and wishes of users from data obtained with various sensors, and then ‘magically’ perform some actions like suppressing phone calls or switching on the light. Over the years, it became clear that it is very difficult or maybe impossible to predict what users want from observing them. Rogers noticed this and proposed the seemingly opposed paradigm of engaging computing. Computers should provide great experiences and engage users more in what they currently do.
We believe that pervasive advertising should be both calm and engaging. Although this might seem like a contradiction, it is not. Calm advertising means that advertisements should be easy to ignore. Engaging advertising means that ads should provide engaging experiences when one is actively engaging with them. This can be achieved at the same time. A pervasive ad could take the appearance of calm mildly waving water when nobody engages with it, and convert to an engaging mini-game once somebody pays attention.
Privacy
Like calm computing, privacy has been an important topic for pervasive and context-aware computing from the beginning. Most systems center on fair information principles (see chapters of Geiger and Haddadi) of notice/awareness, choice/consent, access/participation, integrity/security and enforcement/redress. A variety of systems (like PAWS) have been proposed to implement these principles in technical systems (reference).
In pervasive advertising, there is a huge incentive for advertisers to collect as much user data as possible. Thus, it is crucial that user privacy is being protected. This can happen either through industry self-regulation, lawmaking, or both. How well privacy protection will be guaranteed will determine trust of users towards advertising companies. Such trust is hardly gained and easily lost, and guaranteeing user privacy is one of the foremost challenges for pervasive advertising.
Persuasion
In persuasive technology, it is said that the biggest threats of unethical use are that novelty may mask the persuasive intent, the positive reputation of computers may be exploited, computers can be proactively persistent, computers control the interactive possibilities, they can affect emotions but are not affected by emotions, and computers cannot shoulder responsibility.
It is said that intentions of persuasion can be ethical or unethical, as well as methods. Deception and coercion are always unethical, while operant conditioning and surveillance raise a red flag. Furthermore, it is unethical to persuade vulnerable groups like children. The method proposed to analyze ethics is stakeholder analysis, where all stakeholders are listed as well as what they have to loose, it is then evaluated which stakeholder has the most to gain or loose, and ethics are determine by examining gains and losses in terms of values. Finally, the values and assumptions that are brought to the analysis should be acknowledged.
Persuasion is an integral part of advertising, and ethical use of persuasion is an important challenge for pervasive advertising. We propose, that for advertising any intentions to persuade audiences against their own interests are unethical. Similarly, persuasions of vulnerable groups are unethical, as are any persuasion methods that are deceptive, use coercion, operant conditioning, or surveillance.
Publications
![[PDF]](http://www.florian-alt.org/academic/wp-content/plugins/papercite/img/pdf.png) |
F. Alt, J. Müller, and A. Schmidt, “Advertising on Public Display Networks,” Ieee computer, vol. 45, iss. 5, pp. 50-56, 2012. [Bibtex] [PDF] |
| |
F. Alt and S. Schneegass, “A conceptual architecture for pervasive advertising in public display networks,” in Proceedings of the 3rd workshop on infrastructure and design challenges of coupled display visual interfaces, 2012. [Bibtex] [PDF] |
| J. Müller, F. Alt, and D. Michelis, Pervasive Advertising, Springer London Limited, 2011. [Bibtex] |
|
| |
F. Alt, D. Michelis, and J. Müller, “Pervasive advertising — technologien, konzepte, herausforderungen,” in Media cultures, 2012, pp. 331-338. [Bibtex] [PDF] |
| |
F. Alt, D. Michelis, and J. Müller, “Pervasive advertising technologies,” in Media cultures, 2012, pp. 121-128. [Bibtex] [PDF] |
| |
J. Müller, F. Alt, and D. Michelis, “Introduction to Pervasive Advertising,” in Pervasive advertising, 2011. [Bibtex] [PDF] |