AI Marketing: Baidu Brings a New Sample

When major manufacturers are competing in the commercialization process of large-scale models, marketing has quietly become the fastest landing scenario for large-scale models.

For example, the strongest technology stock of the year in the AI marketing field has emerged. As of the end of November, the cumulative increase of AppLovin, the "AI application superstar" in the US stock market, has exceeded 700% within the year, far exceeding the increase of NVIDIA in the same period. The reason is that the company mentioned that the assistance of AI has continuously improved the AXON advertising engine, making it an important catalyst for business growth. In addition, the latest news shows that Open AI is striving to explore new business models. Their CEO, Altman, is considering adding advertisements to ChatGPT and has hired the former Google advertising executive to handle it personally. Previously, Google has also made commercial exploration in advertising marketing.

In China, Baidu is also exerting efforts in AI marketing and is even to some extent ahead. On December 5, the 2024 Baidu Hot AI Marketing Conference was held in Chengdu. At the conference, the business system was newly upgraded to "Baidu Companion Flight".

It is introduced that this system is based on the Wenxin Large Model. In the next two years, the business space will be expanded by 10 times, and the business operation time will be extended by 3 times. More than 70% of Baidu's customers are already using Baidu's generative AI marketing capabilities.

In recent years, generative AI is accelerating the reconstruction of business logic. In the marketing field, from meeting the needs of information and content to undertaking better business operations, large-scale models have correspondingly improved efficiency and effectiveness, and made the creation and placement of advertisements more intelligent.

01 AI Opens Business Space and Extends Business Operation Time

The emergence of large-scale models has made marketing more intelligent and precise. Through the analysis and mining of massive data, AI can gain insights into consumers' behavior patterns, preferences, and needs, thereby constructing detailed customer profiles for enterprises.

This enables marketers to no longer blindly place advertisements and promotions, but to precisely target and push products and services to the target customer groups who are most likely to be interested. This precise marketing method greatly improves the efficiency and effectiveness of marketing, reduces costs and resource waste.

For users, it can also greatly reduce the cost of finding core information from a massive amount of information.

In the "Baidu Companion Flight" business system, Baidu applies the capabilities of large-scale models to products. Currently, more than 20% of the search result pages contain AI-generated content, providing users with efficient content summaries.

Among these summaries, many have commercial value demands. For example, after users understand the advantages of an item and want to directly add it to the cart, or after users understand the professional services of an enterprise and want to further communicate specific needs online. In these scenarios, on the premise of meeting user needs, the large-scale model also further opens up business space for enterprises.

In addition, in the current era of product promotion, note content with the nature of "guides" is widely popular on various Internet platforms, and note content on Baidu is also increasingly loved by more users.

In this scenario, Baidu AI will determine the note and other content that best suits the needs of users. Through the rich components carried by the notes, while meeting the needs of users, enterprises can also obtain business space from it, and the content will also accumulate as the assets of enterprises, helping enterprises leverage more traffic opportunities.

Large-scale models also show amazing capabilities in content creation. It can quickly generate attractive marketing materials such as advertising copy, video scripts, images, notes, and highlights of short dramas. Moreover, the quality of these AI-generated contents is not inferior to that created by humans, and in some cases, it is even more in line with consumers' preferences and tastes.

In addition, AI can also automatically adjust the style and form of the content according to different marketing scenarios and target customers to achieve the best communication effect.

In Baidu's AI marketing system, the representative of content consumption marketing is short dramas. It is reported that on Baidu, high-quality short dramas can be distributed in a rich field, such as video information flow, short drama channels, search topics, search collections, etc., which not only brings users favorite content but also creates more business space for the copyright owners of short dramas, achieving the integration and win-win situation of usage and business.

What Robin Li previously expressed as "The intelligent entity is the website in the AI era" is also well reflected in the business field.

Obviously, the intelligent entity can provide services to customers 24 hours a day without interruption, quickly and accurately answer customers' questions, and greatly improve the efficiency and quality of customer service. Customers no longer need to wait for a long time for the response of manual customer service, and they can also get satisfactory answers and help at any time.

It is introduced that Baidu currently has more than 20,000 intelligent entities providing professional services every day, and a single user has a maximum of 288 rounds of conversations.

More importantly, in order to help enterprises develop their own intelligent entities, Baidu Marketing combines a large number of practical experiences of customers and summarizes a set of intelligent entity development methodology "C-A-R-E", helping merchants develop more excellent intelligent entities just like training their own employees. In addition, Baidu Marketing has introduced more than 50 industry experts to deeply optimize the conversations and impart industry-specific knowledge to the intelligent entities.

Of course, the intelligent entity can also be combined with the digital human to bring users a more immersive experience.

In terms of channels, Baidu's delivery platform is also continuously evolving and improving.

According to the latest data, with the support of the large model in the planning stage, the delivery platform reduces the account setup time by 50%. In the implementation stage, AI Max can help enterprises achieve the monitoring and optimization capabilities of industry experts, with an average conversion rate increase of 7.2%. In the feedback stage, the time required for analysis and optimization can be reduced from 2 hours to 5 minutes.

02 Baidu AI Has Reached the Moment of Reevaluation

At this year's Baidu World Conference, Robin Li once pointed out that the biggest change in this industry since the development of large-scale models for two years is that large-scale models have basically eliminated hallucinations, and the accuracy of answering questions has been greatly improved.

Baidu's exploration of large-scale models reflects Baidu's understanding of the development process of large-scale models.

For example, the recently launched iRAG (image based RAG) technology is very typical. This technology eliminates "hallucinations", that is, it removes the last stumbling block that hinders the commercialization of generative AI - now the technology has reached the critical point of being implemented.

From the perspective of Baidu's overall AI system at present, Baidu has constructed a complete technical link and is accelerating towards "realization".

At the 2024 World Conference, Baidu released products such as Free Canvas, more than 100 industrial applications, and Miao Da. At the 2024 Hot AI Conference, Baidu Marketing's business system was newly upgraded to "Baidu Companion Flight". Baidu's multi-scenario application product matrix with the "intelligent entity" as the core has initially taken shape.

As a giant from the Internet era, Baidu has been continuously exploring at the forefront of technological innovation. Especially, its investment and R & D achievements in the AI field have attracted much attention.

Generative AI is an important branch of AI technology. With the continuous development of AI technology and the in-depth application of commercialization, Baidu is standing at a new historical node. From the innovation of search engines to the expansion of cloud computing services, and then to the comprehensive integration of AI technology, Baidu has demonstrated sufficient technical reserves and broad growth prospects.

If compared with Google, which has a similar business model overseas, it can be seen that Baidu is also about to usher in its own reevaluation moment in the secondary market due to AI.

For a long time, Google has been the company with the lowest stock price among the seven major technology stocks in the US stock market. However, after the earnings report was released in April this year, Google's market value increased by $225 billion, becoming the third company with a daily market value increase of more than $200 billion.

The reason is that Google seized the opportunity of AI to upgrade its own search and advertising business. Google launched Performance Max in 2023, using AI generation technology to automatically generate the title, description, and related images of advertisements, as well as match the most suitable advertising form and placement channel based on the advertising goals, budgets, industries, audiences and other information provided by advertisers.

Google data shows that Performance Max helped thousands of advertisers increase the average conversion rate by 30% in the testing stage.

Similarly, with the intensification of market competition and the change of user needs, the traditional advertising model has faced certain challenges. The emergence of generative AI provides a new opportunity for the upgrade of Baidu's advertising business.

Advertising is Baidu's cash cow. After being transformed by generative AI, its ability to create cash flow will also be correspondingly improved. Google was recognized after being evaluated by the new valuation system of the capital market, and Baidu will also win the market with the same logic.