AI disrupts the advertising profit pool.

AI is reshaping the advertising industry at an unprecedented pace and can be regarded as the engine of a new round of revolution in the advertising field.

According to the latest in - depth research by Goldman Sachs, AI will shake up the global advertising profit pool of approximately $470 billion in the next few years.

This transformation encompasses all aspects, including advertising placement methods, content creation processes, audience targeting, and creative production.

This transformation is not a distant future but an ongoing reality. #Google and #Meta are reaping the first - wave dividends with AI automation tools.

The AI tool matrix of Chinese technology giants such as #Tencent, #ByteDance, and #Alibaba is also changing the way of advertising creation, placement, and measurement.

01 How AI Reconstructs the $470 - Billion Advertising Profit Pool

Goldman Sachs' research lists digital advertising as the second - most mature track for AI applications after cloud computing and predicts that AI will reshape its profit landscape from four aspects:

Channel Shift ($170 billion opportunity): AI will accelerate the migration of traditional advertising budgets to more efficient and measurable digital channels.

The Goldman Sachs report points out that the penetration rate of digital advertising in the global total advertising has increased from 40.8% in 2017 to approximately 69% in 2024, with an average annual increase of about 4 percentage points.

However, the report also warns that if traditional media can deploy AI on their own to improve placement efficiency, the reconstruction of the profit pool may not be entirely captured by digital platforms.

Creative Production ($114 billion savings): Generative AI will significantly reduce the production cost of advertising creativity. The previously expensive and time - consuming creative development (accounting for 5% - 30% of total expenditure) will be replaced by AI - driven large - scale production.

A vivid example is that Kalshi used Google's text - to - video model Veo to produce an advertisement played during the NBA Finals at a very low cost.

Agency Reconstruction ($161 billion transformation): Automation platforms enable advertisers to directly reach the audience, challenging the core value of traditional advertising agencies. Meta has clearly stated that its goal is to fully automate the end - to - end advertising process, which will directly impact the annual revenue of global advertising agencies up to $161 billion.

Intermediary Compression ($25 billion squeeze): AI - driven one - stop advertising platforms are "abstracting" the technology value chain, reducing the demand for third - party advertising technology intermediaries and potentially squeezing their profit margin of approximately $25 billion.

In short, this is not a zero - sum game but a profound value reconstruction. The dividends brought by AI will be redistributed among platforms, advertisers, and some service providers, while traditional players unable to adapt to the changes will face elimination.

02 Pioneers of AI Advertising Products: Google Performance Max and Meta Advantage+

Goldman Sachs' research report praises Google's Performance Max (PMax) and Meta's Advantage+ as "the most successful comprehensive AI advertising products at present." These two giants have taken the lead in AI - empowered advertising.

They have a common feature: advertisers only need to provide marketing goals, budgets, creative materials, and audience clues, and the system can automatically complete cross - channel placement decisions and optimization using AI, achieving "one - click" advertising placement.

As an innovative advertising product launched by Google at the end of 2021, PMax does not correspond to a single advertising resource but spans across platforms such as Search, Display, YouTube, and Gmail.

After advertisers choose PMax, they only need to provide conversion goals (such as maximizing conversions, target CPA, or ROI), budgets, advertising materials (pictures, videos, titles, descriptions), and first - party data (referred to as "audience signals" by Google).

Subsequently, PMax uses AI to automatically formulate placement strategies, allocate budgets, select the best placement positions, and dynamically adjust bids according to the set goals until the advertisers' goals are achieved.

No manual intervention is required throughout the process - no need to manually set positions, bids, or repeatedly adjust material formats. The AI engine will handle all these.

PMax is very popular among advertisers: in the United States, the proportion of advertisers using PMax has soared from 2% in Q4 2021 to 59% in Q4 2024, and its placement expenditure accounts for 46% of the total Google advertising in the United States.

Meta Advantage+ operates on a similar concept to PMax. Although it was launched slightly later than PMax, its growth rate is also astonishing: the adoption rate of Advantage+ by US advertisers has climbed from 2% in Q1 2023 to 36% in Q4 2024.

This indicates that within just two years, more than one - third of Meta's advertising customers have started to rely on AI - driven automatic placement.

Of course, challengers in niche fields have also emerged. For example, AppLovin's Axon 2.0 focuses on mobile app advertising, and Pinterest's Performance+ specializes in visual search advertising. They are rapidly seizing market shares in their respective fields with the help of AI.

03 Chinese Players' Layout: Seize the AI Wave

In the global advertising landscape, Chinese advertising technology giants (such as ByteDance, Tencent, and Alibaba) are also heavily investing in AI to lead the next - generation advertising paradigm.

Based on the huge domestic market and with deep accumulation in algorithms and data, they are expected to consolidate their advantages and widen the gap with international giants in some fields with the help of the AI wave.

ByteDance: Recreate the Second Growth Curve with AI

As a leader in domestic mobile Internet advertising, ByteDance has risen with its personalized recommendation algorithm (the content distribution engine of Toutiao and Douyin), which itself reflects the power of AI in content recommendation and audience matching in advertising. In terms of AI creativity and placement, ByteDance is fully ramping up:

In early 2025, Douyin launched the "Jichuang AI" platform, marking the intelligent upgrade of its advertising creative production process. Jichuang AI integrates generative AI technology and provides a one - stop tool from script/copywriting generation, image generation to video production.

It is reported that on average, it takes less than 5 minutes to generate a video material, only 33 seconds for graphic materials, and even advertising scripts can be automatically generated within 10 seconds.

The platform has built - in more than 3,500 digital human images covering all industries. Advertisers can freely choose virtual spokespersons that match their target audience, eliminating the expensive cost of real - person shooting. At the same time, it also provides tools such as one - click generation of product cards and rapid generation of graphic materials to help advertisers efficiently create creative materials that meet market hotspots.

More importantly, Jichuang AI connects pre - placement insights and post - placement analysis: before creative output, the system provides industry data and inspiration suggestions to help find potential creative directions; after placement, it tracks material performance through intelligent data diagnosis and has a built - in "blockbuster fission" function that can automatically conduct secondary creation and derivation of high - efficiency materials to continuously optimize placement effects.

It can be said that ByteDance integrates creative production and placement optimization into a closed - loop through Jichuang AI, significantly reducing advertisers' creative bottlenecks and enabling them to respond to the market at a lower cost and faster speed. This is regarded as the key for ByteDance to explore the "second growth curve" of its advertising business, helping advertisers unleash creative potential and improve marketing ROI.

On the advertising placement side, ByteDance's strategy is exactly the same as Google's. UBMax launched by OceanEngine is called the "ByteDance version of PMax". UBMax covers almost all advertising resources of ByteDance (such as Douyin, Toutiao, and Xigua Video), achieving fully automated placement: advertisers only need to provide budgets, placement goals, audience signals, and creative materials, and the system will automatically complete strategy formulation, resource allocation, and optimization according to the goals.

UBMax is not only similar in name to PMax, but their cores are almost the same. In March 2025, ByteDance even switched almost all manual placement methods to UBMax, fully demonstrating its emphasis on and confidence in the AI - driven placement model.

In addition, OceanEngine also launched products such as "OceanEngine Quick Edition" earlier to serve e - commerce advertisers, also emphasizing that the system can intelligently find and expand the audience without manual selection. These measures indicate that ByteDance is fully embracing the AI - driven advertising placement paradigm and hopes to consolidate its leading position in the performance advertising field through technological upgrading.

In the international market, ByteDance has rapidly risen relying on TikTok and has become a new force challenging Google and Meta. One of the important reasons why TikTok can catch up and overtake in the advertising business is its excellent AI - based content recommendation and user insight capabilities, which bring extremely high user stickiness and advertising participation rates.

In 2024, ByteDance further deployed AI advertising tools globally: TikTok launched the Symphony Intelligent Creative Suite, including modules such as Symphony Creative Studio, AI assistants, and digital avatars, providing functions such as generative AI video production for advertisers.

Among them, Symphony Creative Studio can generate short videos with one click from text descriptions, automatically provide editable preview clips, and support powerful functions such as digital virtual humans, content remixing, and automatic translation.

In November 2024, TikTok announced that it would open this generative AI video creation platform to global advertisers. This means that TikTok's advertising customers can quickly customize high - quality advertising content with the help of AI without a complex production team, fully leveraging the marketing power of TikTok's short - video social platform.

ByteDance regards AI as the key to winning in its future advertising business and has built a complete AI tool chain from content production to placement and distribution. This will help it subvert the traditional advertising paradigm and launch a powerful challenge for the global advertising hegemony.

Tencent: Explore the Unique Value of "Social + Content"

Tencent is also sparing no effort in its AI advertising layout. Tencent's advantage lies in its social and content ecosystem (such as WeChat, QQ, Video Accounts, and Tencent Video), with a large amount of first - party data and advertising scenarios.

Facing the AI wave, Tencent chose to launch an advertising solution driven by a large - scale model based on its own technological accumulation.

Tencent Advertising's one - stop AI advertising creative platform, "Tencent Advertising Miaosi", is an AIGC creative suite built on Tencent's self - developed Hunyuan AI large - scale model.

The Miaosi platform provides multi - scenario creative generation tools for advertisers, including text - to - image generation (text - to - image), image - to - image generation (image - to - image), automatic synthesis of product backgrounds, and fine - tuning of specific styles with LoRA. In other words, whether advertisers need new posters, refined product images, or material iterations in different styles, Miaosi can generate them with one click.

Especially in character images and video materials commonly used in marketing, Tencent Miaosi's digital human function performs outstandingly: it has built - in dozens of digital human models of different ages, genders, and clothing styles. Advertisers can choose digital humans to showcase products just like selecting models.

According to Tencent, it used to take 2 - 3 days to produce a set of usable materials with real - person models and post - processing. However, with the Miaosi platform, the same amount of materials can be generated in 10 minutes, with efficiency increased by hundreds of times. This saves a lot of time and cost for advertisers in industries such as gaming, fashion, and education that require a large amount of character - based creativity.

Similarly, Tencent also opens up the multi - modal generation ability of the Hunyuan model to the advertising placement platform, allowing each advertiser to directly use the model to generate text, images, and even video content. This inclusive approach to AI creative ability helps improve the overall quality and richness of advertising materials. Tencent Advertising hopes to "unleash creative production capacity" through the Hunyuan large - scale model and connect the entire process of creative generation and placement optimization.

It can be seen that "Miaosi" is Tencent's strategic foothold for AI - empowered advertising, aiming to consolidate its service advantage in the brand advertising market and catch up with ByteDance in the performance advertising field.

In terms of advertising placement, Tencent is also accelerating its AI - driven transformation. Its proposed "Tencent Advertising 3.0 System" is essentially a highly automated and intelligent decision - making placement platform.

This system eliminates the steps of manually setting bids and defining target audiences. AI automatically processes based on the business information and goals submitted by advertisers. When advertisers use Tencent's advertising placement platform to promote WeChat ecosystem or Tencent Alliance advertisements, they can choose the intelligent placement mode: they only need to enter product selling points, target audience descriptions, conversion goals (such as form collection, order placement), and budgets, and the system will intelligently match suitable advertising positions (such as Moments, official account articles, QQ Space) and audiences.

Tencent also uses AI to improve advertising effect prediction and conversion modeling to make placement optimization more accurate. With the intervention of AI, Tencent Advertising begins to weaken the importance of fine - manual operations and emphasizes more on the information communication between advertisers and the system.

For example, placement specialists need to "tell" the system about the business in detail so that AI can understand and execute the correct strategy. If in the past, placement optimization was a game between people (advertising optimization specialists dealing with media traffic managers), then in Tencent's AI advertising era, it has become collaboration between humans and AI: humans provide high - quality input, and AI outputs optimized results. For Tencent, embracing AI - driven placement can not only improve the ROI of existing advertising customers but also attract a large number of small and medium - sized enterprises (especially those without professional marketing teams) because AI greatly reduces the skill threshold for them to use Tencent Advertising.

This actually expands Tencent Advertising's potential customer base and enlarges the market. Compared with international competitors, Tencent's scenarios (social + content) and data dimensions are relatively unique. AI may play a different role in understanding social relationships and interest maps than in search/e - commerce scenarios.

Overall, Tencent is creating its own AI marketing closed - loop through large - scale model - empowered creativity and multi - channel intelligent placement to comprehensively improve the quality and conversion of advertisers' placements and maintain competitiveness in the new technology track.

Alibaba: AI Paradigm Transformation in E - commerce Marketing

Alibaba's advertising business is highly integrated with its e - commerce platform, featuring transaction - oriented marketing scenarios (such as in - station promotion on Taobao/Tmall). In the face of the AI wave, Alibaba has built an AI tool matrix covering content production to placement around merchants' needs, striving to achieve a new realm of "making marketing easy for everyone in the world":

Alimama (Alibaba's marketing platform) has launched the "Wanxiang Laboratory" generative AI product in recent years, providing merchants with automatic material generation services. Merchants only need to input basic product information and simple instructions to generate high - quality materials such as product short videos and promotional graphics with one click. Take clothing merchants as an example. Alimama uses AI to provide nearly a hundred digital human models to achieve try - on demonstrations in different scenarios, generating high - quality videos in 30 seconds without real - person shooting, significantly reducing material production costs.

Statistics show that the production efficiency of materials intelligently generated through the Wanxiang Laboratory has increased by 5 times compared with traditional methods, and the click - through rate of product advertisements has increased by about 45%. In addition to visual content, Alimama also started early in copywriting generation - the "AI Intelligent Copywriting" system can generate thousands of product copywriting per second based on the massive high - quality content corpus on Taobao and Tmall. After multiple generations of evolution, this system can provide diverse wordings according to different product attributes, achieving personalized customization of product copywriting.

In 2023, Alimama further announced the upgrade of its AIGC creative ability, covering the entire - link intelligent creative support for graphics, videos, and landing pages through four products and six function points. That is to say, from product main images, short videos to detail pages and event pages, merchants can use AI to generate and optimize with one click. The launch of these tools has greatly relieved merchants' pressure on content output during major promotions, enabling small and medium - sized merchants to easily achieve batch content updates during both normal sales periods and major promotion periods.

If the Wanxiang Laboratory solves the problem of "creating materials from scratch", then Alimama's "Wanxiangtai Unbounded Edition" aims to solve the problem of "where to place the materials and who to show them to". The Wanxiangtai Unbounded Edition is a one - stop intelligent placement system that can intelligently place all kinds of graphic and video creat