In 2023, it has been called "the explosive year of generative AI" by global consulting giant McKinsey.
In the past year, AIGC technology has developed rapidly, and more and more AI tools have sprung up, making it possible for AI to be applied in depth in all walks of life. Especially in the field of brand marketing, all of a sudden, every ordinary person's life is full of AI-created advertisements, and even behind the invisible to consumers, brands are also actively exploring the interface between AI technology and business.
In 2023, 8.3% of the winning entries at Cannes Lions Lions International Festival of Creativity, known as the "Oscars of advertising", mentioned artificial intelligence in their synopsis, more than double the 4% in 2022.
From the generation of brand content that inspires creative sparks, to the analysis of data that digs deep into consumer behavior, to the development of personalized marketing strategies, AI technology is being integrated into business practices at an unprecedented rate.
As a year-end summary, "Top AI Player" invited 7 industry leaders who are involved in both "AI" and "brand marketing" to share their 2023 "AI + Brand Marketing" annual cases.
@海辛, experimental filmmaker and AIGC artist
Recommended case: McDonald's & Karen X Cheng, using AI technology to innovate Spring Festival marketing
Case background: To welcome the 2023 Year of the Rabbit, McDonald's partnered with Karen X Cheng, a Chinese digital content creative blogger with millions of followers on Instagram, to launch a series of interactive Lunar New Year activities, including Karen X Cheng's creative design, experiential AR filters, immersive metaverse experiences, and imaginative AI commercials created using NeRF technology.
NeRF (Neural Radiance Field) is a technology that utilizes neural network models to represent and render three-dimensional scenes, which can produce highly realistic three-dimensional images. In an effort to recreate Karen's childhood holiday memories, the brand said the AI ad was one of the first to use NeRF technology. Viewers can scan the QR code on the screen to experience the stunning 3D scenes created by NeRF technology in the form of interactive AR filters.
Reason for recommendation: Karen X Cheng's collaboration with McDonald's uses NeRF technology, which uses neural networks to achieve video editing and effect enhancement, presenting viewers with a visual experience that is both fun and innovative.
Specifically, the application of Nerf technology not only allows the dynamic elements in the video to be seamlessly integrated with the real scene, but also allows the advertising production team to make more refined adjustments in post-production, and through AI analysis of the audience's attention distribution, the key elements in the picture can be accurately adjusted to ensure the effective delivery of advertising information.
Based on the above, I recommend this case study because it is not only interesting, but also represents a cutting-edge application of AI technology in brand marketing.
P.S. On January 23-24, the "Content Good Business" 2024 New List Conference will be held at Hyatt Wangjing, Beijing, and Hai Xin will attend as a guest speaker to share more AI observations
@Tudouren toudou_man, AIGC artist
Recommended case 1: "Mai Mai Museum Series", using AI to create Mai Men heirlooms
Case background: In April 2023, McDonald's officially announced the launch of the "McMc Museum Series", which consists of a group of brain-opening creative works, including a bronze burger, heirloom treasure jade-style fries, and blue and white porcelain-style cola, etc., all of which are imaginative designs by AIGC artist "Potato Man".
The first round of release of the work received nearly 900,000 views on Xiaohongshu, and as the content continued to spread, Tudou reached a cooperation with McDonald's China.
Reason for recommendation: If you share it from my perspective, I can talk about my creative experience. In fact, the "McDonald's Museum Series" was created in January 23, and was originally inspired by two interesting information, one is that a post about positive energy creed published by McDonald's on Instagram was wildly retweeted, and the other is that domestic McDonald's fans proudly call the brand "McGate" and call themselves "McMen".
Inspired by these two messages, I decided to combine my personal art style with McDonald's classic products, integrate elements of ancient Chinese cultural relics, and create this series of works with the theme of "Maimen Holy Relic", hoping to resonate with "McDonalds".
The tool I used throughout the creative process was Midjourney + Photoshop. In the early stage of AI generation, although Midjourney has shown amazing expressiveness in capturing the subject's creativity and creating an atmosphere, it often needs to do a lot of prompt word tests in real creation, iterate round by round and screen effective images, and then I will use Photoshop to complete post-optimization, such as composition optimization, color correction, smearing and modifying details, adding McDonald's LOGO and personal copyright chain logo, etc. At this point, a complete digital work of the potato man has come to an end.
Recommended case 2: Miaoya camera, the first AIGC product in China to break the circle, understands technology and marketing
Case background: Miaoya camera is the first AIGC product to break the circle in China, since it was officially launched in July this year, it has quickly broken the circle on the C-end with the "9 pieces of 9" one-click high-definition AI photo, which has attracted widespread attention. During the period of rising product traffic, Miao Ya Camera also cooperated with Ctrip Travel and "Alibaba Cloud", the official partner of the Hangzhou Asian Games, to launch a series of co-branded photo templates that closely follow hot spots, accurately grasping young consumers' pursuit of personalization and freshness.
Reason for recommendation: The vast majority of AIGC marketing in China is based on gimmicks, and it is very boring to change the skin on the previous form and content. AIGC itself has a certain threshold for use, but in the past year, under the guidance of marketing accounts, ordinary people's cognitive understanding of AI is fast, cheap and simple. This also exacerbates the fact that in terms of marketing actions, people can't go too deep, and still rely on traditional graphics and video forms for packaging.
In contrast, the launch of the Miao Ya camera is very "wonderful", the technology, traffic, products, consumer needs are very clear, in my opinion, the second duck camera is one of the few cases that successfully commercializes AIGC's fresh technology, and effectively cuts into the C-end consumer case.
@李3水, W's Wild Dog Chief
Recommended case 1: "Ye Xiaohui" Women's AI Art Exhibition, which cares more about whether AI is male or female than whether AI is good or bad
Case background: "Ye Xiaohui" Women's AI Art Exhibition is an art exhibition co-founded by W.Ai Nodao and internationally renowned artist Wang Xiaohui, which opened at the Wang Xiaohui Art Museum in Shanghai on July 15, 2023. Over the course of two months, the exhibition brought together thousands of digital artists from home and abroad to create original works in various forms and media, including text, images, videos, and interactive installations, with the theme of "Love, Conception, and Hope".
Reason for recommendation: Frankly speaking, with the rapid iteration of technology, using AI to make art is not new. But looking back at "W Ye Xiaohui" today, it can still be said that it is forward-looking.
First, it is the first AI art exhibition with a female perspective. So far, AI in the field of art and artistic expression, has not stood in the perspective of women to consider its creative logic and language model adaptation path, and "W Ye Xiaohui" is the first AI commercial art exhibition in the category of female perspective, female theme, and female culture.
Second, it is also the first non-artist-led exhibition, not a top-down, elite-dominated art exhibition, but a bottom-up exhibition in which art exchanges and works are presented by unprofessional to professional people, all of whom are from W's community, whether they are students who have taken our AI courses in the W Wild Temple and Nojima community, or some ordinary followers of W on social media platforms.
Thirdly, it is an exhibition that is constantly upgrading and iterative. Different from the static display of traditional exhibitions, "W Ye Xiaohui" has new works submitted and iterated every day during the two-month exhibition period, and expands the influence of the crowd according to the topic every day. This dynamic nature makes the exhibition like a digital creature, constantly iterating, so I think it is an AI art exhibition that is really changing, or pretentiously speaking, growing alive.
Recommended case 2: Make your holiday a masterpiece, Microsoft uses AI to draw ordinary people's holiday memories
Case background: Recently, Microsoft teamed up with advertising agency McCann to launch a creative campaign called "Make your holiday a Masterpiece". At the event, Microsoft invited people to share their holiday stories, which were then translated into creative prompts by cross-media artist Ellie Pritts, and transformed into one-of-a-kind digital artwork using Microsoft Designer's AI image generation technology.
In addition, the audience can participate in this event and create their own artwork through Microsoft Designer's AI creation capabilities. Microsoft's event page also lists the production formula to lower the barrier to entry for users.
Recommended reason: Microsoft uses the AI creation function of its product Microsoft Designer to invite passers-by on the street to share their New Year's holiday memories, and through the form of keywords, Microsoft Designer's AI technology can customize paintings that match these memories.
These paintings not only create a festive scene with a sense of technology and human care, but also have a particularly strong randomness and application, showing the depth of possibilities for virtual technology to transform physical reality.
@八云, founder of the Marketing Lab
Recommended case: Taimo Coffee's crowdfunding has set a record, and "AI Reviews" help brands improve the efficiency of drainage and conversion
Case background: Timemore is a new coffee machine brand, and its electric grinder "Sculptor", which focuses on the design of the powder, successfully raised more than $5 million on Kickstarter, breaking the crowdfunding record for coffee brands.
Recommended reason: For AI marketing cases, I focus on the "C" of AIGC, that is, how to generate content and help businesses or brands get traffic and value on social networks. In the case of Taymour Coffee, a typical internet-based brand, it has set a crowdfunding record on Kickstarter, not only because of its unique burr-ground movement design that establishes differentiation, but also because of its AI-generated "comment section creation" on social platforms like Reddit.
The specific process is probably to collect coffee keywords through AI vertical marketing robot tools, and simulate real personalities to reply to comments. These AI comments are not only highly integrated into the language style of the community/platform where they are located, and fit the context of the upstairs and downstairs in the comment area, but also intimately @Official Account, attach a Youtube review video, and provide a purchase link.
Prior to this, the maintenance of similar comment areas generally only stayed at the brand account, and most of them were either formed template copywriting, or relied on a large number of views and replies from the operation staff. Time more's method is to go out to the content ecology where the official account is located, and cut the comment area between the public and private domains, which has the lowest cost and the lowest user trust cost, and at the same time, combined with AI generation, improves the overall drainage efficiency and saves time and labor costs.
In some of the projects I participated in as a consultant or coach this year, especially in industries that rely on social traffic to obtain clues, the refined operation of each link has become an inevitable trend from live broadcast room to private message, private message to capital retention, capital retention to store, and even the retrieval and active reach of short video comments and liked users. If similar AI tools can be widely used in such marketing scenarios, it should have a more practical effect with the platform's intelligent customer service system.
Another consideration in recommending this case is that on the basis of effect, this kind of AI marketing may be controversial, and it is also worth interesting and observing whether the attitude of users and platforms treats AI reviews differently from traditional water armies.
@赵圆圆, the big V of self-media
Recommended case: Netizens' self-made Balenciaga AI fashion show exploded all over the Internet, why not LV or PRADA, but Balenciaga?
Case background: In March last year, Youtuber "demonflyingfox" posted an AIGC video titled "Harry Potter by Balenciaga", which caused viral and parody on the Internet. In the video, the main characters in "Harry Potter" not only wore Balenciaga's fashion, but the classic lines also became Balenciaga's version.
Up to now, the total number of views of this video has exceeded 10 million. At the same time, many foreign classic film and television IPs such as "Breaking Bad", "The Matrix" and "Game of Thrones" have also been revitalized into Balenciaga AI fashion shows, forming a wave of Balenciaga AIGC craze without brand participation in the whole process.
The production process of these AI magic classic IPs is not complicated, with a total of four steps: first, use ChatGPT to generate prompt words for character descriptions, the second step, write the description into complete prompt words, and hand it to Midjourney to generate images, the third step, use ElevenLabs (an AI voice generation tool) to generate AI voiceovers that imitate the tone of the characters, and the fourth step, use D-ID (a tool for making virtual digital humans) to make the photos generated by Midjourney "move".
Reason for recommendation: Why not LV or Prada, but Balenciaga? Balenciaga has suddenly become enlightened in recent years and has taken the route of Internet celebrity. Its cool design, popular colors, minimalist shape, special materials, with the blessing of celebrities, hipster platforms, and social media momentum, have become a cool representative in the minds of young people.
So what kind of BGC there is, what kind of UGC, the folk Harry Potter AI catwalk is logical. In many cases, the brand itself does not have the attribute of Internet celebrity, and it expects PGC or UGC to produce a hit, which is very screwed. With the emergence of AI, many brands are using it as a hot hype, rather than making AI a replacement for real daily communication work (which is indeed more difficult).
When the folk gods used AI to make a brand-oriented PGC content, it actually helped the brand open up its minds, and it turned out that AI can still be used like this!
@王喆, co-founder of Tezan
Recommended case 1: Nestle 818 Pet Fan Festival, using AI to open up brains and improve efficiency, brand promotion nodes can still play like this
Case Background: The Nestlé 818 Fan Festival is a series of marketing activities organized by the Nestlé Group to thank and give back to consumers, aiming to strengthen the connection between brands and consumers and enhance brand loyalty by providing a variety of offers, interactions and experiences. This year's "Nestlé 818 Fan Festival", Nestlé Group launched an 11-day "Delicious Play Out of Bounds" Fan Festival event, which used AIGC technology to create the "Nestlé Food World", leading fans on all platforms to travel around Nestlé's huge food world through the concept of online cloud travel.
Recommended reason: The Nestle 818 Pet Fan Festival Campaign involves multiple business lines, classic products from different origins, in the case of a tight project production cycle and difficulties in shooting in multiple places, Praise AIGC Studio borrowed the "brain hole" of AI, innovatively generated multiple natural scenes, and reshaped the creative production workflow in a "human-machine collaboration" way: using AIGC technology to quickly and Generate rich concept maps in batches, communicate and determine the creative direction as soon as possible, and efficiently generate realistic scenes and key elements at home and abroad, saving the time and cost of material shopping, modeling and rendering.
In the end, 12 high-quality and difficult materials were produced in three weeks of preparation time, and the construction period was shortened to 50% of similar projects. The materials have been applied on multiple public and private domain platforms, and the novel AI-generated materials have also sparked a check-in boom on social media platforms.
In my opinion, this case is a typical example of AIGC's new technology to assist the production of large-scale marketing content, which meets the needs of small budgets, lightweight production, and high-quality output, and serves the big promotion nodes, providing a new solution for the content needs of "quality + efficiency".
Recommended case 2: Tmall Double 11 super release, "Taobao Design + AI" to create a super brand joint poster
Case background: As one of the highlights of the Tmall 11.11 event, the super brand joint poster attracts consumers' attention in a novel way every year. This year, Tmall joined forces with the top 50 super brands to create a joint marketing event for the opening of Double 11, in which a set of joint posters released by super brands and Tmall were designed and created by AI.
Recommended reason: This case is another innovative joint marketing event after Tmall and 50 super brands, following the "Double 11 AI Brand Float Parade", which needs to complete the poster design of 16+ TOP brands in a relatively short period of time.
In order to meet the different theme demands of "super brands" in a personalized way, the Tmall platform integrates new technologies with old skills in design, and flexibly uses AI and conventional design software in each link.
The posters are all based on the "Signal Tower" that fits the brand image and tonality, echoing the theme of the "Super Release" event. Extend from the creative and brand theme to build different styles of surreal scenes. The picture creates a comfortable imagination space for consumers, which can be said to be both unexpected and in line with the brand tonality, intuitively showing the brand's imagination.
During the design process, from the general direction of using Midjourney to produce creative vision, to the repair of parts of the conventional design software, to the integration of details and materials with Stable Diffusion, the whole process reflects a deep understanding of the overall creative concept and a mastery of new and old design techniques.
@罗锴, co-founder of Tomato Egg MCN
Recommended case 1: WPP, the world's largest advertising group, applied AI to recognize videos
Case background: On November 15, Microsoft announced the launch of the preview version of the multimodal model GPT-4 Turbo with Vision for the Azure OpenAI service. GPT-4 Turbo with Vision is a large-scale multimodal model developed by OpenAI that not only supports image analysis, but also provides text answers to image-related questions. Combined with the Azure AI Vision video retrieval feature, GPT-4 Turbo with Vision is able to extract keyframes from the video, enabling the AI model to summarize the video content and generate corresponding responses.
Recommended Rationale: There is no need to send all frame data to a large multimodal model (LMM) when analyzing a selected set of images in a video to answer video prompts, especially for long videos that may be longer than the 128k token context window of GPT-4 Turbo with Vision. Azure AI Visual Video Retrieval technology not only saves time and money by providing relevant frames to the AI model, but also ensures that the answers generated are accurate.
Satalia, an artificial intelligence center under WPP Group, is using this AI recognition video technology for content analysis and optimization. Applications include creating detailed ad summaries, predicting the impact of video content, and recommending improvements to meet audience expectations and platform specifics. This combination of artificial intelligence and human creativity ensures that the content is not only visually appealing, but also emotionally resonant.
In addition, I recommend this case study because WPP is the world's largest advertising group, and its application of this technology will have a profound impact on the industry. In the future, the technology will be widely used in the project needs of more than 60 subsidiaries of WPP, including Ogilvy & Mather, J. Walter Thompson, GroupM, Wunderman, J&C Advertising and Weida Public Relations Consulting.
Recommended case 2: Meituan refines a proprietary model and opens the brand AIGC design exploration
Case background: Internet platforms often create some virtual characters that are close to the user's image, create a product scene atmosphere, and let users have emotional resonance. However, the traditional 3D character design method has problems such as difficult design implementation and high cost.
To this end, the Meituan platform design team independently trained a series of algorithm models with specific generation capabilities with the help of AIGC, namely the Meituan platform character LoRA model, which covers a set of classic user roles preset by the designers and the image characteristics of each role.
Reason for recommendation: Like WPP Group's introduction of AI recognition video technology, Meituan's proprietary model is also a good landing case. In addition to facilitating quick design and optimizing the design process, the proprietary model also helps unify the style and allow users to form a consistent perception of the tonality of the Meituan brand.
Enter 2024 and look back at 2023.
It can be said that the past year has been both "the year of the explosion of generative AI" and the year of the integration of AI technology into brand marketing practices. In March 2023, Coca-Cola released the commercial "Masterpiece (Masterpiece)" to make the sky-high famous paintings in the museum "move", behind the god-level transition and brain-opening creativity, it is inseparable from OpenAI DALL-E2 generated image model and ChatGPT technical support, which seamlessly synthesizes AI-enhanced animation with live-action action.
The famous painting "The Maid Who Pours Milk", which also sparked heated discussions in the industry, was also created by DALL-E2 and launched in September 2022. Where advertising created or co-created by AI was often seen as a "gimmick", today, with the rapid development of generative AI technology, traditional content production is being disrupted by new technologies, from text and images to audio and video, and brand content production is no exception.
In the past year, from abroad to China, whether it is used as a tool to participate in the production process of brand marketing, or as a concept and traffic to bring voice to brands on the content side, it can be said that AI has become a new engine for brand innovation and growth to a certain extent.
At the same time, it is not difficult to imagine that the practical application of AI in the field of brand marketing will in turn accelerate the commercialization process of AI technology and products. Entering 2024, as AI technology continues to mature and market acceptance increases, we have reason to believe that AI will play a more important role in the stage of brand marketing.