What is artificial intelligence?
A transformative force that promises to leapfrog marketing productivity
Artificial intelligence is a new technical science that studies and develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence. A branch of computer science that attempts to understand the nature of intelligence and produce a new class of intelligent machines that respond in ways similar to human intelligence.
In the summer of 1956, the first human artificial intelligence seminar was held at Dartmouth College in the United States. A group of young scientists jointly researched and discussed a series of related issues of simulating intelligence with machines. Among them, John McCarthy and others proposed the term "artificial intelligence", which was the origin of the concept of artificial intelligence and also marked the official birth of the emerging discipline of "artificial intelligence".
Since the birth of artificial intelligence, the theory and technology have become increasingly mature, and the application fields have also continued to expand. It can be said that artificial intelligence is promoting the fourth leap in the development of human society, and it is expected to become a technological transformation force that will once again dramatically improve productivity after steam, electricity, and information technology. This transformation is expected to be realized within 20 years.
In terms of production efficiency, artificial intelligence will promote technology and innovation in the current world, greatly improve production efficiency and accelerate economic growth. In terms of employment, on the one hand, the existing employment structure will be destroyed and deconstructed, and artificial intelligence technology will subvert some positions and cause unemployment; on the other hand, while destroying, it will also be reconstructed. New technologies will bring new production methods and new service models, which will inevitably give birth to a large number of new businesses, new demands, and new occupations.
The history of artificial intelligence
Artificial intelligence originated in the 1950s and entered a booming period after 2010
Artificial intelligence model classification
From Analytical AI to Generative AI
Artificial intelligence models can be divided into two types: "Analytical AI" and "Generative AI". At the end of 2022, the generative AI ChatGPT series will bring great shock to the human world with its powerful content generation capabilities and "smartness" close to human level.
China's artificial intelligence industry
Industry prospect and industry chain
The industrial structure of artificial intelligence will include three parts: the basic layer, the technical layer, and the application layer. The direction of R&D and the products and services provided by enterprises at different levels are different. After 2020, as the capabilities of the basic layer and technical layer are consolidated, the application layer is also gaining momentum. It is expected that more commercially available large-scale models and application products will appear within 3 years. The scale of generative AI commercial applications in the Chinese market is growing rapidly. According to the "China AI Digital Business Outlook 2021-2025" report released by the Zhongguancun Big Data Industry Alliance, it is expected that by 2025, the scale of generative AI commercial applications in China will reach 207 billion yuan, with an average annual growth rate of 84% in the next five years.
According to Gartner's "Prediction 2021: The Impact of Artificial Intelligence on Humans and Society", 20% of all content will be created by generative AI by 2023. Generative AI is expected to generate 10% of all data by 2025.
The industrial structure of artificial intelligence will include three parts: the basic layer, the technical layer, and the application layer. The direction of R&D and the products and services provided by enterprises at different levels are different. After 2020, as the capabilities of the basic layer and technical layer are consolidated, the application layer is also gaining momentum. It is expected that more commercially available large-scale models and application products will appear within 3 years.
artificial intelligence related technologies
Machine learning, natural language processing, computer vision, knowledge reasoning, etc.
It is difficult for general enterprise organizations to distinguish which AI technologies have real business value. According to Gartner's "2022 Artificial Intelligence Technology Maturity Curve" report, as more and more enterprise organizations believe that the risk has dropped to an acceptable level, the adoption of AI technology has entered a stage of rapid growth, including composite artificial intelligence, decision-making intelligence and edge artificial intelligence, which are expected to become mainstream adoption innovations within 2 to 5 years. Early adoption of these innovations can drive significant competitive advantage and business value, and mitigate issues associated with AI model vulnerabilities.
AI's "Smart Emergence"
More powerful and closer to "human" general artificial intelligence
After the development of AI is strong, it will become an artificial general intelligence AGI (Artificial general intelligence) that is closer to the ability of "human", and can complete more complex tasks like human beings. Let's take ChatGPT-4 as an example:
- Competent for more complex tasks: In various tests designed for humans, ChatGPT-4 scores higher than most human candidates. Based on the input of a picture, complex physics problems can be solved, which means that AI can replace real people to complete repetitive work based on experience
- Multi-modal comprehension ability: image analysis is possible, OpenAI CEO showed at the press conference that GPT-4 can convert sketches into front-end page codes, which means that combined with the Intelligent Internet of Things (AIOT), AI can have a huge breakthrough in interactive applications with real people
- Integrate more applications: ChatGPT is an open ecosystem, with networking + plug-ins, it is possible to integrate more applications. The first batch of plug-ins include travel, shopping, flight accommodation, food delivery, shopping, and collaborative office, which can complete various tasks. Just as the emergence of the iPhone subverted the mobile phone, the emergence of ChatGPT has brought artificial intelligence into the age of the AI iPhone.
State of the Artificial Intelligence Market
In the domestic market, major Internet companies, technology companies, and colleges and universities are intensively building and releasing AI large models, and the battle is in full swing
AI masters say-core player views
AI is the most important industry opportunity of the era and will bring far-reaching impact
At first, we thought (artificial intelligence) was a once-in-a-lifetime opportunity for the Internet, but the more we thought about it, the more we felt that it was a once-in-a-lifetime opportunity, similar to the industrial revolution that invented electricity. Internet companies should solidly implement the underlying algorithms, computing power, data, and more critical scenarios, instead of eagerly displaying semi-finished products in order to boost stock prices. "The company's total revenue has achieved solid growth, reflecting the improvement of revenue structure and operating efficiency. Tencent is investing heavily in building artificial intelligence capabilities and cloud infrastructure to embrace the opportunities brought by the basic model. It is believed that artificial intelligence will become a multiplier for business development." (Tencent Shareholders' Meeting on May 18, 2023)
——Ma Huateng, founder of Tencent
In the process of rapid development of artificial intelligence technology, it is indeed possible to appear in a direction that is not good for human beings. To prevent loss of control, countries with advanced AI technology need to work together to formulate rules from the perspective of a community with a shared future for mankind. To participate in the formulation of the rules, one must first go to the poker table, to have the right to speak, and to have tickets to the global competition. (20230526 Zhongguancun Forum)
I am not worried that the big model will reduce human job opportunities and make life worse. AI cannot replace people, but will create more opportunities and freedom for us (20230518 Tianjin World Intelligence Conference).
——Li Yanhong, Founder, Chairman and CEO of Baidu
Artificial intelligence will have a profound impact on the global economy, society and life. First of all, artificial intelligence will change the traditional manufacturing and service industries, improving production efficiency and service quality. Secondly, artificial intelligence will accelerate the process of automation and intelligence, reduce labor costs, and improve production efficiency and competitiveness. Finally, artificial intelligence will bring more innovation and employment opportunities, and promote social and economic development. An important feature of the development of China's AI field is that more and more companies are beginning to apply artificial intelligence technology to business scenarios, thereby improving efficiency and innovation capabilities. This includes not only e-commerce giants like Alibaba, but also many small and medium-sized enterprises
——Jack Ma, Founder of Alibaba Group
A true era of artificial intelligence has begun. GPT was weak artificial intelligence before, and 3.5 created the era of general artificial intelligence and strong artificial intelligence at the same time. In the future, it may develop into super artificial intelligence. GPT cannot be regarded as a toy, a chat tool, or a search engine. The super brain behind it represents the advent of the era of super artificial intelligence. The power of large models has only just begun.
With the advent of the new industrial revolution, any industry deserves to be reshaped by artificial intelligence. If a product does not have the blessing and changes of AI, it can only be a big knife and a spear, and it is likely to be attacked by people's dimensionality reduction.
——Zhou Hongyi, Founder, Chairman and CEO of 360 Company
AI masters say-core player views
Head players do bottom-level computing power & large models, while small players do applications and verticals
In the future, there will be turmoil in AI large models, not just Microsoft. The direct contribution of artificial intelligence software platform companies to human society may be less than 2%, and 98% is the promotion of industrial society and agricultural society. The popularization of AI services requires 5G connections. In addition to our company's ability to be the underlying computing power platform for AI, the application platform is not an option for us. Our company knows nothing else, so we have to focus on this aspect and strive to build a computing power platform that meets social needs. (Spark Awards Symposium, February 2023)
1. High-quality data is the premise and foundation of artificial intelligence, and high-quality data output should be used as the standard for job completion;
2. Focus on investment and dare to invest, success is only a matter of time;
3. Artificial intelligence should focus on investment and not fully bloom. First, fight the battle of annihilation vertically, and then expand horizontally after winning the flag. (GTS Seminar on Artificial Intelligence Applications in May 2023)
——Ren Zhengfei, Founder and President of Huawei Technologies Co., Ltd.
Friends often ask me, what does Xiaomi think about the big model and AIGC? My unified reply is as follows: Xiaomi has been working in the AI field for many years, and has AI labs, Xiaoai classmates, autonomous driving and other teams. For large models, we will of course go all out and embrace them firmly. We are developing some interesting technologies and products, and we will show them to you when we polish them. (Weibo on April 14, 2023)
——Lei Jun, Founder of Xiaomi Technology
The essence of the advertising industry is to generate good ideas based on good strategies and consumer insights, and then communicate with consumers through good channels. In the past two decades, the media has been digitized to the maximum extent, but the synergistic effect has not yet appeared. Consumers can be classified into thousands of circles, but the marketing side cannot provide thousands of creative ideas. This is the bottleneck of marketing productivity. With the advent of AI, the spring of creativity is also here. The future will be an era of "thousands of people, thousands of channels, and thousands of ideas", and marketing productivity will explode.
In the era of the rapid development of artificial intelligence and the contention of large models, if humans do not have unique insights and cognition, as well as efficient problem-solving capabilities, it is very likely that large models can do what they do. I believe that humans still hold the ultimate power of evaluation criteria, machine learning still needs human feedback, and what we want to create is a beautiful world where humans and machines work together. For marketing practitioners, we must continue to think deeply about the power of brands, expand customer needs, and allow our marketers to have more job opportunities, rather than being satisfied with temporary results.
——Wu Minghui, Founder, Chairman and CEO of Minglue Technology
AI Through the Eyes of a Marketing Scientist
Multiple feelings in the face of AI: excitement, excitement, anxiety, anxiety, stress
On June 8-9, 2023, Miaozhen Academy of Marketing Science held two seminars on AI applications in Beijing and Shanghai. More than 30 marketing scientists participated in the seminars and expressed their views on AI. In the face of the development of AI, the intuitive feelings of marketing scientists are multiple and mixed. The most expressed words are the positive feelings of "excitement" or "excitement", but also accompanied by more negative feelings such as "anxiety", "stress" and "uncertainty".
Many scientists said that AI has brought a greater unemployment crisis to the marketing industry and made the industry more complicated. But at the same time, it also brings the spring of marketing science and new productivity changes.
AI Through the Eyes of a Marketing Scientist
Looking forward to AI can bring: industry revolution, stronger brand, better life
Regarding the expectations of AI, most marketing scientists hope that AI will improve the efficiency of marketing, liberate the productivity of marketing, make the work of marketers more efficient, and empower Chinese brands through technology. At the same time, in life, they expect AI to bring consumers a better life and better experience.
What artificial intelligence means for marketing
AI Breaks the Bottleneck of Insight and Creativity, Bringing an Explosion of Marketing Productivity
Artificial intelligence is reshaping the productivity and production relations of the marketing industry.
Marketing productivity is determined by three elements: insights, creativity, and media effectiveness.
✓ Among them, insight capabilities include the collection and analysis of data and information on consumer needs, market environment, industry trends, and competitive situations, as well as the ability and efficiency to finally obtain analysis results.
✓ Creative ability includes the ability to create and generate text, pictures, videos, audio, websites, etc., innovative ability, and production efficiency.
✓ Media includes various media platforms for marketing communication such as mobile, computer, TV, social media, search, e-commerce, offline, etc., and the ability and efficiency of brands and advertisers to use these channels for marketing.
Today's AI model in the marketing industry has broken the development bottleneck of data insight and creative production, and can increase the production capacity and efficiency of these two links by ten times, or even a hundred times. It can be asserted that AI will bring about an explosion in marketing productivity.
What artificial intelligence means for marketing
All marketing processes and tasks are worth doing again with AI
In the new production relationship, all marketing workflows and tasks are worth doing again with AI. Including: the process and method of marketing insights, the production mode of marketing content, the way media platforms serve and communicate with customers, etc.
Evidence 1: The Level of Marketing Creativity of AI
The development of artificial intelligence technology may bring about a new round of industry changes and human replacement, among which the level of creativity of generative artificial intelligence has sparked extensive discussions.
This research focuses on the field of advertising marketing, using double-blind experiments, recruiting experts and consumers to evaluate the difference between artificial intelligence and human copywriting ability.
Evidence 1: The Level of Marketing Creativity of AI
Introduction to the experiment process
Evidence 1: The Level of Marketing Creativity of AI
Research results: Recognizability of AI copywriting and matching working years
Studies have shown that experts can tell whether a copywriter is human or not.
→ Experts correctly differentiate 59% of the time (62% correctly distinguish human copy, 57% correctly distinguish AI copy), 15% are unsure, 27% misclassify. Experts do not have enough confidence in the accuracy of the judgment, and most of them choose the "probably" option. There is no significant difference in distinguishability between ChatGPT and Wenxin Yiyan.
Ordinary consumers can't tell whether the copywriting is written by humans or artificial intelligence.
→ Among consumers, the proportion of artificial intelligence copywriters who are considered human-written (definitely human + probably human) is 55%, even slightly higher than 54% of human copywriters. Copywriting written by AI is mistaken for human writing 55% of the time. It can be seen that artificial intelligence copywriting has reached a state where consumers cannot distinguish, and in a sense it has passed the Turing test of copywriting.
From the perspective of matching working years, the corresponding working years of human copywriters are significantly higher than those of artificial intelligence.
→ Here, the human copywriting mainly comes from the entries of college students. The equivalent work experience given by experts is 3.83 years (standard deviation 2.19 years); the corresponding experience period of artificial intelligence copywriting is 2.47 years (standard deviation 1.58 years). Experts commented that the equivalent experience years of artificial intelligence copywriting is significantly lower than that of humans, and there is no significant difference between ChatGPT and Wenxin Yiyan.
Evidence 1: The Level of Marketing Creativity of AI
Research results: Professional ability evaluation of AI copywriting
The comparison found that the professional ability of human copywriters is significantly higher than that of artificial intelligence copywriters. There is no significant difference in professional ability between ChatGPT and Wenxinyiyan.
In terms of various professional ability dimensions, experts’ evaluation results of human copywriting and artificial intelligence copywriting show that there is no significant difference between human copywriting and artificial intelligence copywriting in the scores of “clear and complete” and “easy to understand”, and they are indistinguishable; AI copywriting’s scores in “creativity”, “consumer insight” and “business insight” are significantly lower than human copywriting, indicating that there is still a gap between artificial intelligence and humans in advanced capabilities such as creative innovation and handling complex affairs.
The scores of human copywriters and two kinds of artificial intelligence copywriters are very close in terms of consumer perception; in terms of different tasks, human copywriters perform better on a few tasks than artificial intelligence copywriters; but there is no significant perception difference in most tasks, and even for some tasks, consumers perceive artificial intelligence copywriters better.
Subdivided into specific sub-dimensions of consumer perception, human copywriters have low scores on "easy to understand" and "relevance", but higher scores on "unique", "trust" and "like".
Demonstration 2: Consumer AIGC cognition and reaction
The impact of AIGC on the marketing industry depends not only on the breadth and depth of the application of AI in marketing practice by companies on the "supply side", but also on the cognition and acceptance of AIGC by consumers on the "demand side".
Consumer attitudes toward AIGC are not necessarily positive. Studies in the academic community have found that consumers have a tendency to "Algorithm Aversion" (Castelo et al., 2019; B. Dietvorst et al., 2018; BJ Dietvorst et al., 2015). In addition, in December 2022, the visual website Artstation broke out in an anti-AI-generated image campaign, and spread to all major AI-generated content fields. Therefore, it is particularly necessary to study the cognition and response of consumers on the "demand side" to AIGC.
The School of Management of Fudan University will carry out research in 2023, focusing on exploring consumers' cognition and response to AIGC from the demand side, and analyzing whether in social media content, when the different creative subjects of advertising copy are "AI vs. human", will it lead to differentiated responses from consumers.
Demonstration 2: Consumer AIGC cognition and reaction
Consumers cannot accurately tell whether the subject of copywriting is AI or human
• Consumers have absolutely no way to tell exactly whether an ad copy is created by an AI or a human.
Consumers are completely unable to accurately distinguish the creative subject of the advertising copy. Regardless of whether the ad copy they see is actually AIGC or UGC, they are more likely to think that the copy was created by a human (57%).
From the perspective of recognition rate, the average correct recognition rate of consumers for copywriting creation subjects is only 50.1%; when consumers see the copywriting is actually UGC, 57.4% of consumers make a correct judgment; when the copywriting they see is actually AIGC, only 42.8% of consumers make a correct judgment.
• Consumers with strong writing skills and rich experience in using AIGC are more able to make relatively accurate judgments.
Further analysis shows that compared with male consumers (p = 0.03), consumers with strong writing ability are more able to make accurate judgments on copywriting subjects than consumers with weak writing ability (p = 0.006).
In addition, consumers who are more familiar with AI technology and more proficient in using AIGC tools have a higher probability of making accurate judgments on copywriting subjects (p < 0.001).
Demonstration 2: Consumer AIGC cognition and reaction
Consumers' quality evaluation of AI copywriting is not significantly different from that of human copywriting
• Consumers' evaluation of the content quality of AI-generated ad copy is not significantly different from that of human-written copy.
Studies have shown that when consumers do not know the subject of the creation exactly, their evaluation of the content quality of AI-generated advertising copy is not significantly different from the content quality evaluation of human-written advertising copy.
Even in some dimensions, consumers have a higher evaluation of AIGC's advertising copy. Specifically: when consumers see AIGC (compared to seeing AIMC), they have a higher willingness to know about the products recommended in the advertising copy (p = 0.003), and they will be more willing to read the comments below the post (p = 0.045); when consumers see AIGC (compared to seeing UGC), they will believe more in the content of the product introduced in the post (p = 0.034) and are more willing to forward the post (p = 0.046).
Demonstration 2: Consumer AIGC cognition and reaction
Consumer feedback is more positive when AIGC recommends low-involvement, search-oriented, and experience-oriented products
• When the product recommended by the AIGC ad copy is a low-involvement, search-oriented product, it will get more positive consumer interaction responses (likes, comments, forwarding)
Compared with trust-based products (such as health care products and nutritional products), when companies use AIGC to recommend search-type products (such as computers, mobile phones, cameras and other electronic products), consumers will want to know more about the product and show a stronger willingness to buy, like and comment;
Compared with high-involvement products (such as luxury goods, maternal and child products, medical beauty, automobiles, etc.), when companies use AIGC to recommend low-involvement products (such as daily necessities, snacks and beverages, etc.), consumers will want to know more about the product;
Compared with recommending trustworthy products (such as health care products and nutritional products), when companies use AIGC to recommend experiential products (such as travel, hotel accommodation, beauty salons, etc.), consumers show a stronger willingness to comment.
Demonstration 2: Consumer AIGC cognition and reaction
Use "man-machine collaboration" to co-create, eliminate consumers' "algorithm aversion" to AI content
• When consumers are reminded that the main body of advertising copywriting is AI, or once consumers think that copywriting comes from AI, their attitude towards AIGC-recommended products and their willingness to interact with AIGC released by companies will be significantly reduced, showing a strong algorithm-averse tendency.
Specifically: When consumers think that the content of the advertising copy is generated by AI (compared to when written by humans), they will think that the content is less credible, the point of view conveyed is less convincing, and they are more willing to refute the content. At the same time, they do not want to know about the product, and their purchase intention decreases significantly.
Overcoming consumers' algorithmic aversion tendencies is critical to promoting the rapid and widespread application of AI in marketing scenarios.
Evidence 3: AI replaces professional marketing skills
19% of jobs have at least 50% of their daily work exposed to GPT
→ On March 17, 2023, OpenAI, OpenResearch and the University of Pennsylvania research paper "GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models" (early research on the impact of large language models on the labor market) shows that: an estimated 80% of workers belong to occupations where at least 10% of tasks are exposed to LLM, while 19% of workers are exposed to LLM in more than half of their tasks in occupations marked as exposed.
AI can also make career substitutability judgments
→ In April 2023, the Miaozhen Academy of Marketing Sciences will carry out a comparative study on the dual system evaluation of experts and AI, systematically evaluate the substitutability of artificial intelligence for various majors, occupations, ranks, and professional skills related to advertising and marketing in China, and obtain specific and quantitative conclusions. The purpose is to answer the question of "the substitutability of artificial intelligence for professional skills in the advertising and marketing industry", and provide clear and effective guidance and suggestions for personal professional skills improvement and industry change trends.
Evidence 3: AI replaces professional marketing skills
Substitutability of majors related to advertising and marketing
The research shows that among the 22 majors in advertising and marketing, 12 majors are considered to be more than 50% substitutable by experts and AI
Including recording art, animation, e-commerce, painting, photography, visual communication design, digital media art, fine arts, journalism, history, business administration, drama film and television art design. ChatGPT not only gives the ratio of artificial intelligence substitutability, but also gives specific substitutable content of each major and irreplaceable content in the short term.
Evidence 3: AI replaces professional marketing skills
The Substitutability of Occupations in the Advertising and Marketing Industry
The study shows that 11 of the 19 occupations are rated as more than 50% substitutable by both experts and AI
Including editors, painters, graphic design, media placement, project process management, advertising production, digital marketing, new media operations, media planning, art, and copywriting.
Evidence 3: AI replaces professional marketing skills
Substitutability of different positions and professional abilities in advertising and marketing related industries
The research shows that the more senior managers are in the substitutability of artificial intelligence for the ranks of the advertising marketing industry, the less substitutable they are
This is because artificial intelligence cannot yet replace the performance of managers in terms of team management, strategic thinking, leadership, decision-making ability, social relations and resources.
In terms of the substitutability of artificial intelligence for professional skills in advertising and marketing, cognitive ability is more likely to be replaced by artificial intelligence, while the substitutability of soft ability is lower
The consistency between expert evaluation and ChatGPT evaluation is high.
Evidence 3: AI replaces professional marketing skills
Research Conclusions and Recommendations
• The impact of artificial intelligence on the advertising and marketing industry further highlights the irreplaceability of soft skills
• Enterprises need to develop new capabilities of human-intellectual collaboration as soon as possible. There are two new models that can be used for reference
Artificial intelligence as a colleague, the human-intellectual collaboration model includes:
① Collaborative. AI and advertising marketers work together to complete the work of advertising production and delivery. For example, artificial intelligence can work together with advertising marketers for data analysis and advertising production.
② Interactive type. AI interacts with advertising marketers to help them better understand user needs and behavior patterns and achieve better advertising effects.
③ Assisting type. AI can be used as an assisting tool for advertising marketers to help them better complete advertising production and placement,
For example, AI can assist ad marketers in creating creatives.
Artificial intelligence is the leader, and the human-intellectual collaboration model includes:
① Intelligent decision-making type. AI can help advertising agencies make better decisions. For example, based on user data and trends, AI can predict market trends and help advertising companies make better plans and decisions.
② Intelligent management type. AI can help advertising companies with their administrative work. For example, monitor the status and effect of advertising, and help advertising companies better manage advertising.
③ Intelligent control type. AI can be used as a control tool for advertising delivery. For example, according to user data and behavior patterns, the goal of advertising delivery is controlled to achieve better advertising effectiveness.
• There may soon be a new model of human-intelligence collaboration, a new flatter organization, human collaboration with AI to complete more work, a new division of labor model, and reduced marketing outsourcing. Large international companies can establish a centralized international marketing layout and reduce employees in target countries.
AI+ Marketing Applications in Six Fields
Marketing digitalization can be divided into six major areas, each of which can use AI+ to improve marketing effectiveness
China's practical experience in the field of digital marketing leads the world, and Chinese brands have embarked on a new growth path centered on digital marketing. The scope of marketing is very broad, and the application and practice of AI in marketing should also follow the division of marketing digitalization.
From 2019 to 2022, Miaozhen Academy of Marketing Science will continue to iterate the "Marketing Digital Transformation Mountaineering Map" every year to help enterprises recognize the overall goal of marketing digitalization. We divide marketing into six areas: advertising, social media, content, e-commerce, user growth, and innovation management, and describe the digital path of enterprises in the six marketing areas. Enterprises can use the mountain climbing map as an index, and based on their current marketing capabilities, choose an applicable path and implement digital transformation execution strategies. Enterprises in each path can use AI capabilities to accelerate digital capabilities and improve marketing efficiency.
Advertising: As a paid mass communication, advertising is the most expensive form of marketing investment. The digitalization of advertising is firstly the digitalization of the advertising delivery model, including the delivery of digital media, the adoption of digital transaction methods and delivery methods; the second is the assetization of data in delivery, online process, and intelligent decision-making involved in advertising digitalization;
Content: The digitalization of content operations can optimize content insight and production efficiency, improve the distribution efficiency of high-quality content, improve the overall content marketing effect, and strengthen brand equity through cross-channel management of brand consumer experience.
Social media: Social media marketing is based on the distribution of massive content, stimulates consumer interaction, and builds an effective communication channel between brands and consumers. The core of social media marketing digitalization, covering talents, self-owned media operations, social media platform operations and digital insights of consumers;
E-commerce: The online channel of enterprise sales is naturally digital. It provides accurate orientation for enterprises in e-commerce, provides personalized experience, and even redirects people, and provides data support for improving conversion effects. It can also access brand-owned consumer digital data through members.
The production system provides data empowerment for marketing in other fields, and feeds back to cross-e-commerce platform operation tools to improve the overall operation efficiency of e-commerce;
User growth: User growth is a key link in direct connection to consumers, especially data compliance and supervision are becoming increasingly stringent, which is being valued by more and more brands. The digitalization of user growth can legally obtain user authorization and data, and through refined layered operations, directly connect and reach consumers through all channels, and provide each user with more complete and personalized services and user experience;
Innovation management: agility, innovation, and user-orientation are important factors driving the development of enterprise performance. Enterprises can only obtain new driving forces through agile innovation. Better integrate and use new technology means, promote effective collaboration among all parties, and improve innovation success rate and innovation efficiency.
AI+ Marketing Applications in Six Fields
The priority application scenarios of AI+ in various fields, the available scenarios are different
AI+ Advertising Application Scenarios
The AI priority scenarios for advertising are advertising intelligence, advertising creative generation, and advertising copywriting
AI+ advertising practice case
Advertisers actively embrace AI technology as a highlight of advertising creativity
• Application direction: video ad creative
• Application case: Coca-Cola's March 2023 "Masterpiece" creative advertisement
• Case description: This advertisement is produced using AStable Diffusion, and the real shooting, 3D production and AI painting are completed together. Integrating famous art paintings with reality, it shows that in a fictional art gallery, the art collections are discussing how to deliver "water of inspiration" to tired students-cold and refreshing Coca-Cola. The combination of real scene shooting, digital special effects and AI technology makes famous paintings of different styles come to life, conveying to the public the diverse and creative "Real Magic" brand spirit of Coca-Cola across time, region and culture.
• Application direction: image ad creative
• Application case: Fliggy's subway advertisement in April 2023
• Case description: The advertisement is based on the theme of "What to do on this May Day", and the AI painting function is used to create tourist attractions. Compared with ordinary landscape photos,
While adding innocence and romance to AI painting, it has also caught up with a wave of AI enthusiasm. AI maps imaginary travel destinations. flying pig
The official WeChat public account published an article "One hour, Fliggy used AI to make 1,000 print advertisements!" Telling the story of AI print posters. At the same time, the Weibo launched the topic #飞猪AI广告打热论论#.
• Application direction: video ad creative
• Application case: KFC's May 2023 advertisement "First Experience and AI Dialogue"
• Case description: In the video advertisement, Zhu Yilong, the brand spokesperson, asked AI: "What is the taste of love at first sight? What is the taste of elation? What is the taste of never forgetting?". After denying AI's answer, Zhu Yilong shared his understanding with AI, introduced KFC's new KFC product, and finally expressed "a taste that AI can't feel, created by humans." This advertisement uses AI elements, which aroused heated discussions and attention.
AI+ advertising practice case
Media and agency companies use AI tools to optimize services and innovate advertising products
• Application direction: advertising creative generation
• Application case: Meta will launch an artificial intelligence sandbox tool in May 2023
• Case description: Meta's artificial intelligence sandbox, including three tools: text change, background generation and image display. Text variations can generate multiple versions of ad copy, allowing advertisers to choose the version that best suits a particular audience; background generation helps advertisers enter text to create images that can be used as backgrounds for their advertising materials. Texts such as "blurred streets of New York City" and "blurred snow mountain abstract high quality" can generate corresponding background images; the image display allows advertisers to automatically adjust the size of the image on the Meta platform. If you want to post an ad on Instagram's main feed, Stories and Reels, you only need to provide a picture, and the AI tool can automatically adjust the picture's aspect ratio.
• Application direction: AI conversational advertising
• Application case: Microsoft, Google introduce advertising in generative AI
• Case description: On March 29, 2023, Microsoft stated that "it is trying to insert advertisements in Bing chats; on April 20, 2023, the Financial Times reported that Google will introduce generative AI into its advertising business in the future. Currently, Google has used AI in its advertising business to create simple prompts (Prompts) to encourage users to buy products.
• Application direction: intelligent advertising, precise advertising
• Application case: Focus Media's intelligent marketing and precise delivery
• Case description: Focus Media stated in its 2021 annual report that the company continues to improve its digital and intelligent marketing capabilities. Currently, it has realized cloud online push, and has connected most of the screens to the Internet of Things, realizing remote online monitoring of the screen's playback status. At the same time, through continuous in-depth analysis and research on big data and cooperation with platforms such as Tmall, the company continues to help customers achieve accurate and effective delivery according to the consumption needs and brand preferences of consumers in different buildings and communities.
• Application Direction: Industry Efficiency Tools
• Application case: BlueFOCUS AIGC tool matrix
• Case description: In March 2023, BlueFocus officially announced the release of the "AI Xiao Creation" tool. In April, Blue Label stopped four outsourcing expenses related to creative design, plan writing, copywriting, and short-term employees, which caused heated discussions in the industry. BlueFocus focuses on building proprietary applications: BlueFOCUS AIGC tool matrix empowers, improves efficiency, and reshapes marketing scenarios. Focusing on industry model and application layer innovation, a number of projects have been implemented in creative material production, public relations dissemination, virtual human content, virtual live broadcast and other businesses, and pilot cooperation with customers.
AI+ content application scenarios
AI priority scenarios for content are content tagging, image generation, and text generation
AI+ content practice case
Songs, TVC, posters, anime, scenes, models...
AI+ content practice case
Songs, TVC, posters, anime, scenes, models...
• Application direction: simulation model
• Application case: Dutch company Lalaland.ai AI model
• Case description: In March 2023, the American jeans brand Levi Strauss & Co announced that it will cooperate with the Dutch artificial intelligence company Lalaland.ai
Cooperation, through artificial intelligence to generate digital models of different body shapes, skin colors and ages to display its products, and it is expected to start testing in the second half of this year. AI model customization process: first select parameters such as hairstyle, body shape, size, skin color, etc., and specify pose, emotion and other options. Then, on the software page, carry out targeted clothing design based on the AI model, or import clothing files into the AI model without actual sample cutting. After the final design is verified, a 3D clothing sales page can be generated.
• Application direction: brand personalized content generation
• Application case: Start-up company Typeface brand content model + content review algorithm
• Case description: Artificial intelligence technology can improve the accuracy of marketing through big data analysis and machine learning. Founded in 2022, Typeface's start-up product is to create personalized text and image content for large brands based on corporate brand positioning and audience goals. Unlike general content generation for the general public, the requirements of big brands for brand tonality and content controllability cannot be achieved by directly calling Stable Diffusion. (There are brand image safety issues such as inaccurate, plagiarized or offensive content), Typeface helps big brands solve the above pain points through personalized training of brand-specific models and content review algorithms.
AI+ social media application scenarios
The AI priority scenarios of social media are social media graphic content generation, social media streaming automation, and social media intelligent insights
AI+ social media practice case
Integrate AI into social media marketing activities, intelligent insight, public opinion management and control
AI+ social media practice case
Personalized interaction with consumers through social media data analysis
AI+ e-commerce application scenarios
E-commerce’s AI priority scenarios improve efficiency for AI models, smart digital human customer service, and smart operations
AI+ e-commerce practice case
The platform integrates AI to develop operation tools and optimization algorithms
• Application Direction: Intelligent Operation Improves Efficiency
• Application case: E-commerce SaaS service provider Shopify integrates ChatGPT
• Case description: Shopify is a SaaS company that provides e-commerce platform services for global e-commerce retailers, providing e-commerce website establishment, maintenance management, logistics and transportation, platform payment and other e-commerce operation processes. On March 1, 2023, Shopify took the lead in integrating ChatGPT, upgrading its intelligent customer service (chat robot), helping consumers make personalized recommendations, and helping merchants save interaction time. In addition, ChatGPT can also provide platform product review data analysis, title and keyword optimization, marketing copywriting, intelligent website development and programming, and other functions, effectively improving sellers' operational efficiency and consumer experience. It is expected that once ChatGPT is introduced into the country in the future, it will bring about huge changes in the field of e-commerce and e-commerce agency operations. For example, when searching for products, there are multiple results, and there is only one answer. ChatGPT will recommend more suitable products based on user buying habits and behavior data, enhance the matching between the platform and consumer needs, increase user stickiness to the platform, and build competition barriers for service providers.
• Application case: smart product recommendation
• Application case: what is worth buying - algorithm optimization
• Case description: What's Worth Buying continues to optimize the "Thousands of People, Thousands of Faces" recommendation and search algorithm. Through efforts such as enriching user behavior sequence characteristics, adding more positive and negative feedback, and optimizing fusion formulas, it enhances the diversity of system recommendations and the accuracy of searches, presents users with richer consumption decision-making content, and improves users' content acquisition experience. From the fourth quarter of 2022, "what is worth buying" will try more content such as pictures generated by AIGC and product highlights. Research and try to build a conversational user decision-making scenario, introduce a conversational robot assistant to provide highly matching product and content recommendation results, and improve the efficiency of user consumption decisions.
AI+ e-commerce practice case
AI virtual human helps improve the efficiency of live e-commerce delivery scene
Application direction: virtual anchor
Application case: silicon-based digital live broadcast of silicon-based intelligence
Case description: Silicon Intelligence is the only AIGC technology company that Tencent has invested heavily in the field of artificial intelligence. With the goal of "siliconization of carbon-based life", we focus on the innovative application and production scenarios of digital human + AIGC, and realize digital human application solutions that are closer to humans, personalized and good at learning. The digital human live broadcast platform and silicon language short video platform are widely used in local life live broadcast, e-commerce live broadcast with goods, cross-border e-commerce and other fields. It is applicable to Douyin, Kuaishou, Video Account, Taobao, Tmall, Pinduoduo, JD.com and other domestic video and e-commerce platforms, and supports overseas video platforms.
Application Direction: Intelligent Operation Improves Efficiency
Case Case: Qingmu Co., Ltd.'s e-commerce intelligent system
Case description: Aoki Co., Ltd. cooperated with the Ali data team to jointly explore data application solutions, create and optimize educational data models, and improve the efficiency and effectiveness of e-commerce operations. At present, a data application system has been established for data-based commodity operation, data-based traffic operation, full-link digital marketing, data-based crowd operation, and data-based customer management. System integration can be carried out according to the individual needs of brand owners, providing full-link comprehensive IT solutions, and self-developed multiple IT systems:
AI+ User Growth Application Scenarios
AI priority scenarios for user growth are intelligent customer service, customer analysis, and service evaluation
Practical cases of AI+ user growth
AI Robot Service Helps Brands Reduce Costs and Increase Efficiency
• Application direction: intelligent customer service
• Application case: AI dialogue robot of Songuo Travel
• Case description: Songuo Travel is a well-known motorbike sharing brand in China, with nearly 50 million users and a daily order peak of 3 million. With the growth of business, in the process of serving users, in terms of service scale, service cost, corresponding speed, etc., problems that cannot be solved by conventional human resource solutions are encountered. With the help of AI technology and business integration capabilities, a multi-round dialogue robot is created to achieve cost reduction and efficiency increase. The original large number of manual customer service links (such as checking user identities, checking orders, applying for refunds, etc.) realize self-service processing, accurately identify user inquiries around the clock and respond in seconds, improve the problem-solving rate of robots and the proportion of intelligent consultation, solve the problem of "single-round interaction" that has plagued many years, and reduce costs and increase efficiency for service teams. Realize the release of 10 manual customer service personnel, saving about 2 million labor costs per year.
Practical cases of AI+ user growth
AI supports customer insight, service, evaluation, and directly writes research reports
AI+ Innovation Management Application Scenarios
AI priority scenarios for innovation management include intelligent demand analysis, new product opportunity discovery, and virtual sample design
AI+ product innovation practice case
Use AI to design new products, generate virtual products, and discover new products
• Application direction: virtual sample design
• Application case: Nike AR shoe fitting software Fit app
• Case description: In the application of AI artificial intelligence, AI design samples can be seen on many Nike sneakers. In addition to the design output of shoe shape and concept, AI can also automatically configure colors and detailed lines and patterns. Based on the virtual models and samples rapidly generated by AI, it has achieved cost reduction and efficiency increase in reality. Nike launched the Fit app, an AR shoe-testing software based on computer vision, that allows consumers to choose the shoes that best suit their foot size through precise intelligent calculations. Relevant data will also be retained for reference in the next consumer purchase decision. The technology relies on technology from an Israeli computer vision startup that Nike acquired.
AI+ product innovation practice case
Using AI to design product packaging is already a very common application
• Application direction: exploration of new product concepts, discovery of new product opportunities
• Application case: Miaozhen system new concept exploration platform
• Case description: The new concept exploration platform developed by Miaozhen System monitors the whole network public opinion of new products and new concepts in the industry for brands, and quickly understands industry trends and opportunities. Combining the dimensions of people, time, context, and interests, it helps brands imagine new concepts and identify new opportunities. Not only can it generate the most potential product concept combination, but it can also directly generate promotional copy through text training of big data.
AI empowers individuals, realizes economies of scale, and improves marketing efficiency
Marketing is a representative of the service industry, and AI will bring about a revolution in the service industry.
In the previous industrial revolution, through scientific management, thousands of small workshops were integrated for standardized production, forming a large-scale production and manufacturing effect, thereby replacing individual labor with large factories, which completely affected the manufacturing industry.
Today, the service industry, including marketing, relies on the ability and experience of individuals and lacks scale effects. The largest accounting firm or law firm in the world is far smaller than manufacturing companies.
With AI technology, enterprises in the future may produce "super employees", whose individual abilities and experience will be greatly empowered by AI to complete previously unimaginable work tasks.
AI will make it possible to build service industry factories and realize the mechanization of mental work. Enterprises in the service industry, including marketing, will rely on AI technology to form a "service assembly line" to achieve economies of scale.
To formulate an AI strategy, an enterprise must first decompose the mental work internally based on the marketing scenario, determine which links can be completed by AI, and then restructure the workflow and collaboration mode to collaborate with AI to improve productivity and improve marketing efficiency.
——Wu Minghui, Founder, Chairman and CEO of Minglue Technology
How can companies use AI to improve their marketing capabilities?
The implementation path of AI applications can be divided into four stages: full use, mind sharing, intelligent transformation, and organizational evolution
How can enterprises adopt suitable AI models?
Make decisions based on three elements, based on training data, different models are competent for different tasks
training data is important
• Regardless of the type of model, natural and high-quality training data is crucial to the model's effect. To be competent for professional domain tasks, training on general data, dialogue data, domain data (such as industry knowledge base), and effect feedback data is required.
• It is recommended to use Finetune to achieve the effect of LLM+local knowledge, which is better in terms of cost efficiency, enhanced model generalization ability, and cumulative effect.
• A general-purpose large model may be competent for vertical domain tasks only after being trained on specialized domain data.
How can enterprises adopt suitable AI models?
Build industry or company models based on the general large-scale model capabilities and according to application scenarios and enterprise conditions
→ Enterprises of different sizes and types have different ways of implementing AI applications and adopting models. Whether to use a general solution needs to be decided according to its own situation.
→ For the use of 2B marketing business scenarios, it is recommended to start from work tasks and integrate products with large models to call appropriate-scale models.
→ The general model and the industry model will coexist and develop for a long time
→ The training optimization of large industry models in vertical fields benefits from fine-tuning based on general large models. That is, two supervised fine-tuning (Supervised Fine-Tuning, SFT): instruction fine-tuning (Instruct Tuning) and reinforcement learning based on human feedback (RLHF).
→ Perform finetune with data sets that are close to the business, and use rich "domain professional models" to better complete complex tasks in professional fields.
→ Cooperate with adaptive model service partners with data security, high cost-effectiveness, and good results to reduce development costs and empower their own marketing.
How Can Businesses Gain the Competitive Advantage of AI?
Competitive advantages come from differentiated customized models, strategic investment, data knowledge accumulation, business scenario selection, etc.
