This article is from the WeChat public account: One-Symbol Studio (ID: One-Symbol), author: Jin Xin YOYO, the original title: "Sora has been upgraded again?
Just 40 days after the last stunning appearance of Sora, on the morning of March 26, OpenAI once again demonstrated its powerful video generation capabilities. Different from the realistic video scene of "blowing up everyone", this time OpenAI has taken its creativity to the extreme, and by teaming up with major visual artists, it has generated imaginative and more difficult videos.
Among these videos are vlogs with the "Yellow Balloon Man" symbol throughout, "pseudo-documentaries" of lifelike cross-species creatures, and fashion shorts of "Undersea Fritillary Fashion...... It can be described as a hundred flowers blooming, which makes people dizzy.
Also detonated the circle of friends is Suno. It can quickly create a piece of music that is not annoying with arbitrary text and parameter settings. The era of "everyone can become a director" has just arrived, and the era of "everyone can become a musician" is not far off.
On the other hand, it is the AI intelligent assistant KIMI, which satisfies the domestic "AI backwardness anxiety", which came out to show off the muscles of "2 million words long text processing", and at the same time triggered a carnival of concept stocks in the secondary market.
In just 2 months, as a cross-border practitioner of advertising and AIGC, I have deeply experienced the rapid development of AIGC applications at home and abroad, as well as the huge impact on advertisers and content creators.
However, at the level of practical commercial application, there are still many difficulties that need to be broken through in AIGC. Next, I will talk about the current market challenges and opportunities based on the specific customer needs in the field of advertising and marketing.
1. Advertisers are actively embracing AI, starting from wanting to save money, which will lead to price wars for suppliers
If the technological development of AIGC can bring productivity breakthroughs to many advertising companies, then for advertisers, what they are actually concerned about is not whether you as a service provider can reduce costs and increase efficiency, but whether they themselves reduce costs and increase efficiency.
The advertiser doesn't care whether you use human or AI for content marketing, he only cares about which supplier has a lower cost under the premise that the quality standards are similar, because this directly determines how much marketing cost he can save. Therefore, from this starting point, it is inevitable that suppliers will be involved in cost wars and price wars.
Cutthroat market competition has always existed, and nothing will change just because AI is coming.
Second, advertisers will be more strict in their evaluation of AIGC technology, but this is not necessarily what advertising companies are good at
According to iResearch's online survey data in December 2023, more than 93% of advertisers have used AIGC technology to assist in creative content production, followed by some transactional and strategic marketing efforts.
In other words, customers generally accept the use of AIGC to assist in the production of advertising content, especially after the whole year of "AI tool tuning" last year, Party A and Party B's technical understanding of AIGC has gradually converged, and may also have a general understanding of the strength and application direction of mainstream suppliers.
There must be a star financing company that makes TO C application products. Because TO C is just a cloak for building momentum in the market and penetrating the mind, it may be that the real profit model of others is still TO B. Moreover, there are indeed many suppliers of AI applications, ranging from unicorns to self-employed individuals.
Even Party A, which I think is a little more flexible, may already be studying and formulating an AIGC cost reduction, efficiency increase and content quality evaluation system, including in the bidding and bidding links, and will look at the technical strength of each supplier very objectively.
This is actually a challenge for advertising companies that have excelled in creativity, strategy or media resources in the past.
To put it bluntly, there is a huge AI talent gap in the entire industry. Not all designers can learn to program and achieve transformation and upgrading, and AIGC engineers may not have received systematic aesthetic training. Therefore, this industry has high requirements for people's interdisciplinary compound ability.
3. Advertisers' rigid needs for the accuracy and quantity of content > thirst for creative content
We all know that now is the era of the rise of social e-commerce and live broadcast e-commerce, and advertisers are more and more inclined to do "effective marketing". Therefore, the customer can't wait for all parties B, C and D to "sell goods" to him, instead of engaging in any "brand advertising". This is especially true for many up-and-coming brands.
At the same time, this can also explain why e-commerce advertising has always occupied the majority of the advertising market.
Because sales can't be faked, users don't need to listen to "nonsense" when they buy things. Therefore, advertisers have a more urgent demand for content accuracy and content production quantity, but creativity is a problem to consider in the back row.
This requires us, as a service provider, to focus on our advantages: how to reach consumers with high frequency through "saturated content" + "matrix army" under the "platform algorithm mechanism", and then influence purchases.
99% of customers will ask the advertising image to accurately show the original appearance of the product, and they will be very concerned about the slightest mistake.
In terms of the cost of model training, at least 100~200 product real-life materials are needed to practice 90% similarity AIGC maps, and the resulting training cycle and cost are not low, and a SKU has to be restarted.
From the advertiser's point of view: if I can give you so many real-life pictures at one time, what do I need you to do to produce so many AIGC images?—— unless the customer's demand scenario is a large number of pictures, so many that tens of thousands, hundreds of thousands. And such customers are not available in all industries.
Meituan is a typical example, and its own AI character library is really good. I summarize the following two reasons:
First, it is because Meituan has many categories and marketing activities, and it is in high demand. The content center of a character library can be trained to allow each business line to access materials and use them freely, so as to meet the scenarios that can be used infinitely in one training, so as to achieve the purpose of reducing costs and increasing efficiency by internal designers;
Second, their solution is more in line with the actual capabilities of the current technology: only the characters that appear in the picture are trained, and the product is still used with merchant materials, so it can effectively avoid the problem of low product restoration after Lora model training.
Fourth, AIGC text + full marketing + new media, a highly effective marketing combination
At present, it is technically very mature to complete the training of marketing text models based on media characteristics, industry attributes, brand data and product selling points.
Taking my company as an example, the internal platform has been applied in multiple industries. Especially in the scenario of full marketing + new media. Because it serves the long-tail marketing chain population, it is lighter in terms of service model, such as open API interface, model training, product customization, training guidance and operation services. This is also different from the traditional advertising agency delivery model.
For example, dealer employees can use our platform to quickly generate new media tweets, short video scripts and other content, and restaurant employees can produce Moments content and creative images to maintain regular customers in the private domain, etc.
From the perspective of brand owners, they can realize full-staff marketing management and data monitoring from internal employees, dealer employees, KOCs, KOLs and even to the user layer through the background, which is very necessary for large-scale marketing campaigns of millions or even tens of millions.
I don't need to say much about the value of full-staff marketing: for example, the rejuvenation of the Wuling Motors brand and the acquisition of customers through live broadcasts mostly rely on the planting of modified car owners and the self-broadcasting of dealers.
This is also the reason why brands nowadays are more and more inclined to do all-staff marketing. All-hands marketing is the greatest resource integration for advertising campaigns, and AIGC tools minimize the cost of content creation for these people.
Employees don't like to use their brains, so we help them set the framework and leave it to AI to solve the problem of ability.
Fifth, if it is really more creative than creative, AIGC user co-creation may be more in line with the marketing trend
Let's take Nongfu Spring and Wahaha packaging, which have been very popular recently.
I was also very ashamed to rub the hot spot of Nongfu Spring (the left side of the picture below is designed by the author), and I caught a small fire, which was very accidental. In this regard, I want to learn from Wahaha's AI packaging designer, and the effect on the right is really amazing.
The picture on the left is Nongfu Spring, and the picture on the right is Wahaha
My colleague Li Yichen is even better, he is the director of the first AI Village Gala, because the artist co-creation videos during the Spring Festival have been widely disseminated, which has brought him more resources.
In fact, whether it is in the field of product packaging or video production, in the eyes of brand owners, these are still at the level of "technology" and belong to the single-point grasp of marketing.
There can be many kinds of grippers: for example, there will be many special activities every year, by linking brands and designers, and holding various co-creation design competitions to achieve interaction and win-win between brands and users.
As for the level of "Tao", I always believe that the test of "simplicity" is never how well the hand is done, but whether the brand owner really understands marketing and how to integrate resources to achieve the best marketing effect.
If you want to really find valuable ideas, it's not enough to rely on advertising agencies to compete with each other.
You need to really connect this circle and community, and encourage more AI designers, creatives, and super individuals to participate in co-creation through bonus incentives.
In this way, you can not only collect the best quality works, but also establish your own private domain user pool, and stimulate the communication potential of the majority of netizens, so that your brand can get more exposure to natural topics.
I think it's a win-win situation for both "Lizi" and "Face".
This article is from the WeChat public account: A Symbol Studio (ID: One-Symbol), author: Jin Xin YOYO
This content is the author's independent view and does not represent the position of Tiger Sniff. May not be reproduced without permission, please contact [email protected] for authorization
People who are changing and want to change the world are all in the Tiger Sniff APP