From advertising philosophy to flywheel effect: an AI-driven marketing revolution
Deep thinking in the wave of change
On the afternoon of July 23, I attended the [[AI Driven × Sustainable Innovation] Industry Trend Lecture] (https://www.cisanet.org.tw/Course/Detail/5705) hosted by the Industrial Development Bureau of the Taipei City Government and executed by the Information Software Association of the Republic of China. The organizer invited a total of four experts to share their views. Among them, Dentsu Marketing Communications Group Strategy and Innovation Chief [Shao Yiwen] (https://www.facebook.com/freda.shao) is not only my friend for many years, but also a schoolmate of the National Taiwan University Engineering Institute.
It is a pleasure to listen to her share her in-depth insights into the marketing changes in the AI era as Chief Strategy and Innovation of Dentsu Marketing Communications Group. This lecture made me deeply realize that we are in an era of unprecedented change. Not only technology is evolving rapidly, but the entire business ecosystem is undergoing fundamental reconstruction.
With thirty years of experience in the marketing industry, Yiwen deeply analyzed the challenges and opportunities currently facing the marketing industry from the perspective of strategic innovation. She is not just talking about how to use AI tools, but how to redefine the nature of marketing, reconstruct business processes, and create sustainable competitive advantages in the AI era. This kind of strategic thinking made me realize that AI brings not only an improvement in efficiency, but also a fundamental change in business model.
After listening to the entire lecture, my biggest feeling is: this is not a simple technology application sharing meeting, but a forward-looking thinking on the future development direction of the marketing industry. At the current intersection of multiple trends such as AI, big data and sustainable development, every practitioner needs to rethink their value positioning and competitive strategies.
Philosophical reconstruction of the nature of advertising
At the beginning of the lecture, Yiwen asked a seemingly simple but extremely profound question: “Why do people watch advertisements?” This question pointed directly at the core philosophy of marketing and made me rethink the fundamental meaning of advertising. Her answer is concise and powerful: Human nature hates making choices, and the core value of advertising is to help consumers make decisions faster.
This interesting insight suddenly enlightened me. In our daily lives, we face countless choices every day, from what to eat for breakfast and which means of transportation to taking, to which brand of product to buy and which restaurant to choose for dinner. Every decision requires the consumption of cognitive resources, and human cognitive resources are limited. When there are too many choices, we often fall into difficulty choosing, which not only wastes time, but also causes psychological pressure.
Excellent advertising is like an efficient decision-making aid. It helps consumers quickly select the most suitable options through clear value proposition, emotional resonance and trust building. This not only saves consumers time costs, but more importantly reduces their psychological burden and makes the decision-making process easier and more enjoyable.
In the AI era, this core value has not been weakened, but has been strengthened like never before. AI technology can analyze large amounts of user data and deeply understand individual preference patterns, consumption habits and life scenarios, thereby providing more accurate and personalized recommendations. But Yiwen also reminds us that technology itself cannot directly affect consumers’ emotions and decisions. This still needs to be achieved through creative expression and emotional connection.
I just took this opportunity to re-examine the evaluation criteria for advertising creation. In the past, we often focused on the creativity, visual impact or brand tone of advertisements, but Yiwen reminds us that the most fundamental evaluation criterion should be: does this advertisement really help consumers shorten their decision-making time? If an advertisement is exposed just for the sake of exposure and does not solve the consumer’s choice difficulty, then no matter how brilliant the creative idea is, it will not be considered a truly successful advertisement.
This interesting point of view also made me think about the future direction of content marketing. In the era of information explosion, consumers do not need more information, but better information screening and decision-making support. Excellent content should have the function of decision-making assistance and can help consumers find the answer that suits them best among complex choices.
Three-dimensional redefinition of brand value
Yiwen put forward three levels of in-depth analysis of brand building, each level clarifying the core value of the brand in the new era. These three levels do not exist independently, but are a complete system that is interrelated and progressive.
The first level is to help companies use money correctly. Although this statement sounds very straightforward, it accurately outlines the business logic of brand investment. When a brand establishes sufficient awareness and trust, the efficiency of a company’s marketing investment will be significantly improved. Consumers’ familiarity with the brand reduces their cognitive costs and search costs, allowing each marketing budget to have greater impact.
From an ROI perspective, this is a classic compound interest effect. A strong brand in the market is like an asset that continues to increase in value. Every marketing investment will have an amplifying effect based on the existing brand assets. On the contrary, companies that lack a brand foundation need to build consumer awareness from scratch for every marketing campaign. This is not only costly, but the effect is often unsustainable.
The second level is about the sustainable investment benefits of the brand. Yiwen pointed out that the longer the brand management lasts, the lower the subsequent marketing expenses will be, because consumers have established trust and loyalty to the brand, and companies can interact with consumers through multiple channels instead of relying entirely on paid advertising. This reminds me of Customer Lifetime Value Value) concept, high-quality brands can establish long-term customer relationships, thereby obtaining a more stable and predictable revenue stream.
What impressed me most was the third level. Yiwen mentioned that brands should bear the responsibility of changing consumer behavior and promoting social progress. She pointed out that in some cases, the influence of brands may even exceed that of governments, which gives brands a greater social mission. Today, when sustainable development is becoming increasingly important, brands must not only create commercial value, but also create social value.
This perspective also made me rethink the boundaries and meaning of corporate social responsibility. Traditional CSR is often seen as an additional burden for enterprises, but Yiwen’s analysis allows me to see that in the new era, social responsibility has become a core component of brand value creation. In particular, the younger generation of consumers are paying more and more attention to the values and social influence of brands; in other words, they are willing to pay a premium for brands that can promote positive social change.
The redefinition of brand value at these three levels made me realize the complexity and systematic nature of brand building. It is not only a tool for marketing communication, but also a core component of a company’s business strategy, and is also related to the company’s long-term competitiveness and sustainable development capabilities.
Search ecological reconstruction in the AI era
In the lecture, Yiwen particularly emphasized the profound impact of AI search on brand exposure. We are indeed at a critical moment in the reconstruction of the search ecosystem. Recently, I have often talked about this part with students. Traditional Google search is gradually being supplemented or even partially replaced by AI assistants such as ChatGPT and Claude, and this change will naturally have a fundamental impact on the visibility and influence of the brand.
When consumers use AI tools for search and consultation, whether a brand can be recommended by AI largely depends on its authority and credibility on the Internet. This not only includes traditional SEO indicators, but more importantly, whether the brand information has been reported by trusted media, whether the website content is continuously updated, and how long the user stays on the website and the depth of interaction, etc.
An important point mentioned by Yiwen is that 80% of online content is now synthetic data, including real data provided by companies and data generated by AI. In such a huge ocean of information, which content is trustworthy and which sources are authoritative have become key issues that both consumers and AI systems need to solve.
This also got me thinking about an important shift in content strategy. In the past, content marketing mainly focused on attracting user attention and improving search rankings, but in the AI era, the credibility and authority of content have become more important. Enterprises need to pay more attention to cooperation with authoritative media, establish a verifiable professional reputation, and continuously maintain and update their digital assets.
At the same time, this also means that the threshold for brand building has been raised to some extent. New brands entering the market need to spend more time and resources to establish authority, while brands that already have a certain degree of popularity and trust have a greater first-mover advantage. This change may intensify the market’s Matthew Effect, that is, “the strong get stronger and the weak get weaker.”
Another trend worthy of attention is the change in consumer search behavior. Compared with traditional keyword searches, AI assistants support more natural conversational queries, and consumers can ask more complex and specific questions. Having said that, this is also equivalent to requiring brands to not only consider keyword optimization when creating content, but also think about how to answer various in-depth questions that consumers may raise.
The double-edged sword effect of generative AI
Yiwen gave an in-depth analysis of the opportunities and challenges brought about by generative AI in the lecture. This analysis made me deeply aware of the double-edged sword characteristics of technological progress. On the one hand, AI tools have greatly improved the efficiency of content creation and significantly lowered the creative threshold; on the other hand, they have also brought new challenges such as quality control, legal compliance and brand consistency.
From an opportunity perspective, generative AI gives everyone basic creative capabilities. Image design that used to require professional designers can now be quickly realized by ordinary people through AI tools. This not only reduces creation costs, but also significantly shortens the time cycle from idea to finished product. This is undoubtedly good news for small and medium-sized enterprises with limited resources.
But Yiwen also reminds us of potential risks. The first is compliance issues. AI-generated content may involve legal risks such as copyright infringement and portrait rights infringement. She specifically mentioned that it is relatively safe to use images of celebrities who have been dead for more than 80 years, but for content that is still within the copyright protection period, you must be extremely cautious when using it, otherwise you may face legal action.
Second is the challenge of brand consistency. Different people using the same AI tools often produce content with very different styles. If there are no unified creative standards and quality control mechanisms within the enterprise, it is easy for brand image confusion to occur. Consumers may not be able to clearly identify the uniqueness of the brand, which is very detrimental to brand building.
The third risk is the tendency of content homogeneity. When everyone uses similar AI tools and prompt words, the generated content will often show similar characteristics. This homogeneity not only weakens the brand’s differentiation advantage, but may also cause consumers’ aesthetic fatigue.
Yiwen’s analysis made me realize that although AI tools have lowered the threshold for creation, they have not reduced the importance of creative strategies. Enterprises need to establish a more complete content management system while embracing new technologies, including creation specifications, review processes, quality control and other links. Only in this way can the advantages of AI tools be truly utilized while avoiding potential risks.
The transformation dilemma and solution of marketing agents
As the strategic innovation chief of a large media agency like Dentsu, Yiwen has personal experience and profound insights into the transformation challenges of the agency industry. She candidly analyzed the impact of AI on agents’ business models, as well as the industry’s strategic thinking in response to changes. Yiwen mentioned that Dentsu Marketing Communications Group has about 1,200 employees in Taiwan. When AI tools can automatically complete many tasks that originally required manual work, companies must rethink their human resources structure and business processes. This is not only based on cost considerations, but also a re-examination of the entire business model.
Traditional agency business processes are often very complex, and may take several months from ideation to final execution. This process involves the collaboration of multiple departments, including strategic planning, creative design, media placement and effect monitoring. The introduction of AI technology allows many standardized and repetitive tasks to be automated, thereby greatly shortening the entire project cycle.
But Yiwen also pointed out that this change brings new opportunities. First, agents can internalize work that was originally outsourced to production companies, improving service integration and quality control capabilities. Secondly, AI tools lower the technical threshold and allow more employees to participate in work that originally requires professional skills, which enhances the overall capabilities of the team.
Most importantly, AI technology allows agencies to more effectively demonstrate their core values: deep understanding of consumer behavior, keen insights into market trends, and systematic thinking in creative strategies. When standardization work is taken over by AI, human creativity, judgment, and strategic thinking will become even more valuable.
Yiwen’s analysis made me realize that the transformation of agents is not only a technological upgrade, but also a reshaping of value positioning. In the AI era, agents need to shift from executors to strategic partners, and from cost actuaries to value creators. For agents, it is not only necessary to master new technologies, but also to improve their overall strategic thinking and innovation capabilities.
AI opportunities and challenges for small and medium-sized enterprises
Yiwen pays special attention to the opportunities and challenges faced by Taiwan’s small and medium-sized enterprises in the AI era. This analysis has important guiding significance for the majority of small and medium-sized enterprise owners. She pointed out that small and medium-sized enterprises usually face three core pain points: not enough people, not enough money and not enough time. The popularization of AI technology provides new possibilities for solving these pain points.
From an advantage point of view, small and medium-sized enterprises have flexibility and agility that large enterprises do not have. Short decision-making chains, low communication costs, and high execution efficiency give SMEs a natural advantage in adapting to new technologies. When AI tools can make up for the lack of manpower and resources, small and medium-sized enterprises can often achieve transformation and upgrade faster than large enterprises.
Yiwen suggested that small and medium-sized enterprises start with data collection and systematically sort out the data accumulated in the past in terms of customer interactions, sales records and service feedback. Although this data may not be as vast as that of large enterprises, it is often more accurate and targeted. Through the analysis of AI technology, small and medium-sized enterprises can have a deeper understanding of their customer base and discover new business opportunities.
The second is the deep cultivation of vertical scenes. Compared with large enterprises that need to consider the coordination of multiple business lines, small and medium-sized enterprises can focus on specific niche markets or business scenarios and conduct in-depth AI application experiments. This kind of focus not only reduces the complexity of technology introduction, but also makes it easier to achieve significant results.
But Yiwen also reminded small and medium-sized enterprises to pay attention to potential challenges. The first is the pressure on initial investment costs. Although the cost of using AI tools is constantly decreasing, considerable investment is still required to establish a complete AI application system. Therefore, small and medium-sized enterprises need to choose the appropriate entry point according to their own circumstances and avoid investing too many resources at one time.
Secondly, there is the challenge of talent and knowledge. Although AI technology has lowered the threshold for use, it still requires corresponding knowledge background and operational skills. In other words, SMEs need to invest time and resources in employee training and building internal AI application capabilities.
The most important challenge comes from the strategic planning level. AI is not a panacea. Enterprises need to clearly know what problems they want to use AI to solve and what goals they want to achieve. Blindly following the trend will only waste resources and may even be counterproductive.
Yiwen’s suggestion is that small and medium-sized enterprises should start with small scenarios, learn and improve in practice, and gradually build their own AI application capabilities. Adopting this incremental approach not only has lower risks, but also makes it easier to gain successful experience and lay the foundation for subsequent expansion.
Employee transformation and corporate culture reshaping
Yiwen particularly emphasized the importance of employee experience in the AI transformation process. This perspective gave me a deep understanding of the humanistic dimension of technological change. She pointed out that AI transformation is not only a technological upgrade, but also a comprehensive change of organizational culture and human resources strategy.
In terms of recruitment, Yiwen shared an interesting detail: Dentsu will now give priority to job seekers with GitHub accounts. Although this is only a small indicator, it reflects the importance companies attach to employees’ AI literacy. GitHub is not only a gathering place for programmers, but also an important platform for knowledge workers in the AI era to learn and share. Job seekers with GitHub accounts often indicate that they have some exposure to and understanding of new technologies.
After hearing this, I immediately logged into my GitHub account to view some of the projects I had been working on. Ting Yiwen mentioned this short story, which made me realize the changes in the company’s talent evaluation standards. In the AI era, traditional academic qualifications, experience and other indicators are still important, but employees’ learning ability, adaptability and innovative thinking have become more critical. Enterprises need to establish new talent identification mechanisms to find talents who can collaborate with AI and use AI to improve work efficiency.
In terms of employee training, Yiwen emphasized the importance of AI literacy education. This includes not only how to use AI tools, but more importantly, understanding the boundaries of AI’s capabilities, potential risks, and ethical requirements. Employees need to know what can and cannot be done, what is safe or dangerous.
Especially in the field of content creation, legal issues such as copyright, portrait rights and privacy protection have become more complex. Yiwen mentioned that companies need to establish an internal AI tool certification mechanism. Only certified tools can be used by employees, and this must be done within the company platform to ensure data security and compliance with regulations.
This also made me realize the responsibilities and challenges that enterprises face in the process of AI transformation. On the one hand, companies need to provide employees with opportunities to learn and grow and help them adapt to new technologies; on the other hand, they must also establish a sound management mechanism to ensure the safety and compliance of technology use.
Yiwen also mentioned the special considerations of large enterprises in the use of AI tools. Due to the need for data security and trade secret protection, many large companies still prohibit employees from using AI tools such as ChatGPT, and instead develop their own internal systems. Although this approach increases costs, it can better protect the core assets of the company.
This also made me think about the differences in AI strategies among companies of different sizes. For example, some large companies have the resources to build their own AI systems, but small and medium-sized enterprises may need to rely more on public platforms. How to enjoy the convenience of AI while protecting corporate secrets is an issue that every enterprise needs to seriously consider.
Three types and ability requirements of future talents
At the end of the lecture, Yiwen conducted a forward-looking analysis of the future development direction of marketing talents. She believes that in the AI era, there are three main types of marketing talents, and each type has its own unique value and development path.
The first is Performance Optimizer, which is the performance optimizer. The core ability of this type of talent is to carry out in-depth optimization and improvement based on the simple tasks that AI can already complete. They must not only be able to use AI tools, but also be able to analyze the results produced by AI, find room for optimization, and continue to improve results. This role requires data analysis skills, business understanding and optimization thinking.
Yiwen particularly emphasized that if a person cannot find room for optimization and cannot add value to the output of AI, then he is likely to be eliminated. This also reminds us that it is not enough to just learn to use AI tools. The key is to be able to create incremental value based on AI.
The second type is Trusted Solution Builder, which is the builder of trusted solutions. The core ability of this type of talent is to integrate different AI tools and technical resources to build end-to-end solutions. They need to understand the capabilities boundaries of various tools and know how to organically integrate different systems such as ChatGPT, Google Forms, and data analysis tools to provide customers with complete solutions.
The key to this kind of role lies in the word “believable”. In an environment where AI tools are plentiful and of varying quality, customers need not only powerful solutions, but also solutions that are reliable, secure, and compliant with regulations. Talents who can provide such credible solutions will have huge market value.
The third type is Asset Builder, that is, asset builder. The core mission of this type of talent is to help companies continue to build and accumulate brand assets, rather than conduct one-time marketing activities. They need to have long-term strategic thinking and be able to connect every marketing activity with the company’s long-term brand building goals.
Yiwen uses the metaphor of a skyscraper to illustrate this concept: Every company hopes that its brand will be as stable and towering as a skyscraper, rather than as gorgeous but short-lived as fireworks. But as we all know, the construction of skyscrapers requires time and continuous investment. Every content creation must consider whether it can add bricks and tiles to brand assets and whether it can enhance consumers’ awareness and favorability of the brand.
The division of these three talent types also allows me to see new directions for career development in the marketing field. Each practitioner needs to choose an appropriate development path based on his or her own interests and abilities, and continue to improve corresponding professional capabilities.
Implementation path and strategy suggestions
Yiwen also shared the implementation path of enterprise AI transformation in the lecture. These suggestions are of high reference value for enterprises that are considering or have already started AI transformation.
The first is to build AI literacy among all employees. Yiwen emphasized that before starting any AI application, companies must ensure that all employees have basic AI literacy, including an understanding of AI capabilities and limitations, awareness of ethical and legal risks, and an emphasis on data privacy and security. Blind use of AI tools may not only bring risks, but may also affect work effectiveness.
The second step is data preparation and sorting. Since the effectiveness of AI depends largely on the quality and completeness of data, companies need to systematically sort out existing data assets and accurately assess the strengths and weaknesses of the data. For areas with relatively complete data, AI application experiments can be prioritized; for areas with a lot of missing data, data collection and sorting work needs to be strengthened first.
The third is tool evaluation and selection. There are many AI tools on the market. She suggested that enterprises need to establish their own evaluation system and choose the combination of tools that best suits their business needs. Yiwen recommends that companies focus on four aspects: brand safety, quality control, audit process and compliance management. Only after these basic issues are solved can other higher-order functions be considered.
第四是场景化应用。 Enterprises should not pursue large and comprehensive AI systems, but should start from specific business scenarios and conduct targeted application experiments. Successful small-scenario applications can not only bring direct business value, but also provide employees with opportunities to learn and grow, laying the foundation for subsequent expansion.
The last step is continuous optimization and iteration. AI technology is developing rapidly, and enterprise-related applications also need continuous improvement. This includes not only technical upgrades, but also continuous improvements in process optimization, personnel training, and improvement of management systems.
Yiwen specifically reminded that enterprises should clarify the goals and expected effects of AI applications. If a company does not know what problems it wants to solve and what goals it wants to achieve with AI, then the initial investment is likely to be wasted. Therefore, she suggested that enterprises start by selecting small, accurate, and quantifiable application scenarios, and then gradually expand after gaining successful experience.
Future Outlook and Thoughts
After listening to this lecture, I have a deeper understanding and clearer understanding of the marketing changes in the AI era. In the past two years, I have learned a lot of theoretical things in the doctoral class, and Yiwen’s sharing not only allowed me to see the opportunities brought by technological progress, but also made me aware of the challenges and risks in the process of change. What impressed me most was her analysis of Flywheel Effect. In the AI era, successful companies need to build their own business flywheel: more customers bring more data, more data supports better AI models, better AI models provide better services, and better services attract more customers. Once an enterprise establishes this positive cycle, it will form a strong barrier to competition.
This made me think that AI is not only a tool, but also a new business logic. Under this logic, data has become the most important means of production, algorithms have become the core production tools, and human creativity and judgment have become the scarcest resources. Companies need to rethink their value creation models and redesign their organizational structures and incentive mechanisms.
Yiwen mentioned at the end of the lecture that Kevin Kelly once predicted that no one will discuss AI in thirty years, just like few people discuss digitalization now, because AI will be completely integrated into our lives and work. This prediction makes me feel both excited and pressed for time. What is exciting is that we are at the starting point of a great change and have the opportunity to participate in and witness the creation of history; what is urgent is that the window for this change may be shorter than we think, and we need to learn, adapt and act faster.
As a marketing lecturer and consultant, I am deeply aware of the profound changes this industry is undergoing. The advancement of information technology has provided us with more powerful tools and broader possibilities, but it has also placed higher demands on our professional capabilities and strategic thinking. In this era of change, only those who can embrace change, continue to learn, and have the courage to innovate can stand out from the competition.
Yiwen’s sharing made me more convinced: in the era of AI, human value is not replaced, but redefined and improved. We need to learn to collaborate with AI and leverage AI’s capabilities to amplify our own value while leveraging uniquely human creativity, empathy, and judgment. These are not only the needs for personal career development, but also the requirements for the development of the entire industry.
The road ahead may be full of challenges, but it is also full of opportunities. As Yiwen said, we all have the advantage of early entry. The more we know, the greater our opportunities will be. In this era of change, let us embrace new technologies with an open mind, improve our abilities with professionalism, explore new possibilities with innovative thinking, and jointly create a better business future.
Further reading
- [Content Marketing Frontier] Building Taiwan’s AI Competitiveness ─ Participate in the 2025 Taiwan Artificial Intelligence Annual Conference
- Make good use of generative AI to help content marketing
- [Content Marketing Frontier] Master the opportunities of combining content marketing and artificial intelligence ─ 2023 Taiwan Artificial Intelligence Annual Conference
