The Chinese authorities have vowed to crack down on illegal advertisements on the internet and in apps, given the app advertising industry has remained largely unmonitored until now, carrying all kinds of advertisements, from healthcare products to financial products, some of which are suspected of engaging in crooked promotions and providing misleading or even wrong information, and some violate the public order and social norms. Photo: FILE

---

 

A new generation of emerging technologies such as mobile communications, big data and artificial intelligence pose an existential threat to the traditional advertisement industry and advertising studies. Computational advertising has arisen as a new approach to deliver effective brand messages to a wide range of audiences at scale.

In contrast with traditional advertising, computational advertising provides an ideal platform for monitoring, tracking, gauging and evaluating advertising campaigns and consumer behavior, thus enabling the possibility to provide accurate, personalized advertising services with quantifiable results.

In this sense, we can view the birth of computational advertising as an unprecedented opportunity for both academic disciplines and industrial practices of advertising and marketing.

 

User-centered era
Traditional advertising is production-led and sales-driven, and due to the scarcity of channels and the endowment of mass media, it is shackled by the channel hegemony. Advertisers rely on product producers and cater to the needs of customers, so they act like middlemen while penetrating through every level of marketing in terms of the social division of labor, even the upstream of production. Throughout creativity, design, production and sales, traditional advertising blends into the whole life cycle of a product, in order to achieve the effective control of production and consumption and thereby maximize profit.

Once new internet technologies have established a diverse dialogue platform for producers and consumers, a new era of service-led and demand-driven advertising will be just around the corner. The relationship between producers and consumers will become a flat end-to-end relationship, channel restrictions will be gone, and the legitimacy of advertising will fall under scrutiny. For now, however, advertisers are counting on the help of artificial intelligence to come up with massive, fresh and precise ideas to assert their presence.

However, the current advertising industry and studies have yet to clarify the social relations involved in advertising in the smart era. A majority still regard computational advertising as the inheritance and extension of traditional advertising theory and practice, using data and technology to optimize the old advertising scheme.

Some scholars have proposed that computational advertising is a mechanism of advertising based on target users and web page content, which can be calculated to find the best match and carry out precise targeted advertising. The adoption of this mechanism can greatly improve the click-through rate, reading rate and page view of advertisements. It can help users obtain high-quality information, so as to build a benign and harmonious advertising industry chain.

In my opinion, this perception is still deeply affected by the traditional production-led and sales-driven way of thinking. It is necessary to comprehensively and thoroughly figure out the basic characteristics of the service-led and demand-driven intelligent era in order to define the connotation and extension of computational advertising in a more scientific manner.

Intelligent media sides with the users (consumers), so advertisers must realize a change of perspective to better conduct brand promotion and product information distribution. Namely, they should move from being agents of production to being agents of consumption, from relying on producers to relying on consumers, from the distribution of product information to the distribution of consumer demand, from finding accurate consumers to finding exact producers, and from increasing company sales to meeting user demand.

 

Advertising studies                                         
At present, the advertising discipline’s paradigm is an integrated marketing communication based on the industrial economy of mass production and homogeneous consumption. The decision-making and research process lacks effective data support, and the research focuses more on creativity, planning, packaging, media and consumer behavior studies.

In the new era, it is possible to precisely locate the audience and measure the effect of advertising, and the advertising discipline’s paradigm should change accordingly. The old-fashioned paradigm cannot provide the best solution to the problem of advertising practice in the intelligent era. Hence, here comes computational advertising.

Computational advertising is a rapidly growing interdisciplinary research area that overlaps with a variety of established scientific disciplines, including computer science, artificial intelligence, advertising, marketing, linguistics, statistics, economics, psychology and sociology.

At the same time, many data-driven techniques have been employed in advertising research, including information retrieval, large-scale search, text analysis, statistical modeling, machine learning, optimization and econometrics.

Therefore, the discipline’s paradigm shift is by no means spontaneous and is indeed synchronized with social and historical development. Computational advertising also bears the historical responsibility of the transformation of the advertising paradigm and the promotion of the new paradigm.

The application of new information technology and theory has led to the profound transformation of the knowledge system of the advertising discipline and ways of solving relevant problems, which constitutes the external pressure of the transformation of the advertising discipline’s paradigm. The academic community of the advertising discipline has undergone unremitting inquiry of and reflection on the paradigm, seeking to find a paradigm in line with the times and thereby having internal strength.

First of all, researchers of computational advertising should strengthen their beliefs and insist on recognition of the discipline. Once members of a discipline begin to doubt their mission, the legitimacy of their discipline will naturally be shaken. Second, computational advertising should focus on the interactive consumption behavior on the internet and improve and enrich the knowledge system for solving real-life problems.

Computational advertising faces the complexity and uncertainty of ever-changing technology. It needs to learn from others and seek the best path of paradigm construction in the interdisciplinary integration.

                                                        
Possible risks
German sociologist Max Weber divided rationality into two types, namely, value rationality and means-end rationality. Value rationality believes in the unconditional value of certain behaviors and emphasizes the pure motivation and the right means to achieve the intended purpose, regardless of the results. Means-end rationality holds that actors are driven by utilitarian motives, achieve expected rational goals, only consider the maximization of effects, and ignore emotional and spiritual values. Social progress is inseparable from the development and application of advanced technology, but basic values should not be cast aside.

That said, the advertising industry and its study will discard the out-of-date logic of seeing itself as an auxiliary to industrialized production, mass marketing and the pursuit of maximized profit. In today’s world, practitioners and academics must take on a user-centered mentality to stress personalized consumer service and to better bridge consumers and producers.

When we criticize traditional advertising production and sales logic and classic advertising theory, hazards such as production waste and excessive consumption are always the first denounced. However, we should also anticipate how computational advertising might put us at greater risk.

The central challenge of computational advertising is to find the “best match” between a given user in a given context and a suitable advertisement. The information about the user can vary from being scarily detailed to practically nil. The number of potential advertisements might be in the billions. Thus, depending on the definition of “best match,” this challenge leads to a variety of massive optimization and search problems, with complicated constraints.

In this light, while avoiding the multiple harms of traditional advertising, we need to step up efforts to prevent data monopolies, information breaches, privacy violations, value misguidance and cultural pollution. Whether on the side of business and innovation, on the side of academic research and disciplinary development, or on the side of talent training and teaching practice, computational advertising should keep up with the times and pursue loftier values.

The healthy computational advertising industry and the high-quality computational advertising discipline are complementary. The future has arrived and excellent computational advertising practice and research will be expected.

 

Ma Chijie is a professor from the School of Humanities and Communication at Guangdong University of Finance and Economics.

(edited by YANG XUE)