Daily Interaction: How Big Model x Big Data Drives High quality Brand Growth

Mondo Technology Updated on 2024-01-30

At present, the "100 model war" has entered the second half of the spelling and application, which has brought huge opportunities to reduce costs and increase efficiency to thousands of industries. Focusing on the field of brand marketing, what changes has the large model brought to the industry?How can enterprises use large models to optimize digital marketing methods, improve digital marketing effects, and ultimately drive brands to achieve high-quality growth?

Liu Yu, Senior Vice President of Daily Interactive (Getui) and brand digital marketing expertStarting from the perspectives of industry trends, scenario applications, and product landing, it will bring the sharing of "AI is not only GC, but also TA" to brand advertisers and related practitioners, and explain the new ideas of brand digital marketing in the era of large models.

Expert presentation

Liu Yu has been with Daily Interactive for nearly ten years, fully responsible for the strategic development and continuous innovation of the company's products and business, and has rich experience in the fields of digital intelligence marketing, data intelligence business planning and implementation, and has served as the executive director of the School of Software of Zhejiang University, and a distinguished lecturer for graduate courses.

Key takeaways

In this wave of large models, AIGC has brought a huge boost in productivity and innovation to enterprises. For the field of marketing, AI can not only be GC, but also TA. The innovative combination of artificial intelligence and data can help brand owners find target groups more conveniently and efficiently, and improve the efficiency and effectiveness of corporate brand marketing

AITA's "Smart Crowd" tool is an innovative result of the large-scale model application of Daily Interactive in the field of brand marketing. In this product, we use a large model to give it an understanding of natural semantics, and use data weaving technology to realize the semantics of the data space, and realize innovative functions such as "seedless insights" and "conversational promoting".」

With the advantages of convenient operation, rapid feedback, and no need for seed people, AITA mainly brings two improvements to brand owners: one is to lower the operational threshold and improve marketing efficiency, and the other is to broaden the application scenarios of data in the field of brand marketing」

Front-line marketers only need to put forward marketing demands in the dialog box anytime and anywhere like daily conversations, and AITA can quickly give their own marketing strategy suggestions and crowd targeting results, especially suitable for some lightweight marketing scenarios that require real-time feedback, scenarios that lack seed crowds such as new product promotion, and some enterprises with insufficient data volume」

The long-term value of the large model will be realized through industry applications, and scenarios are the key. The company is also continuing to explore the application of large models in various industries and fields, and hopes to work with more industry partners to create and jointly build new large model application scenarios and basic capabilities, so that large models can shine brighter in industry applications, and data intelligence can benefit more industries and fields through the power of large models.

[The following is a transcript of the interview].

01. The application of large models is accelerating to various fields, and the marketing field is no exception. What kind of opportunities do you think big models can bring to the marketing field?

In this wave of large-scale models, AIGC has brought a huge improvement in productivity and innovation to enterprises, and has gained unprecedented market attention. Especially in the field of marketing, generative AI has shown great energy in the output of creative content such as text and multimodal (**), which has brought great convenience to creative planning, content production and other departments. With the blessing of the large model,Content that used to take weeks or months to produce can now be completed in days or even hours, allowing the amount of content that can be delivered per day to grow exponentially。The significant improvement in the efficiency of content output allows the marketing team to easily create tens of thousands of ideas in a short period of time, fully unleashing the imagination of creative personnel, and will further stimulate the innovation potential of enterprises, so that enterprises are more passionate and motivated in their imagination and planning for future business.

In addition,AI can not only be GC, but also TA。TA stands for Targeting Audience. In other words, the targeting of the target group, which is very critical in brand marketing, can also be "AI". This innovative way of combining artificial intelligence and data can help brand owners find target groups more conveniently and efficiently, and improve the efficiency and effectiveness of corporate brand marketing.

02. Daily Interactive released the "AITA Smart Crowd" tool at the 2023 World Internet Conference, which attracted market attention. Can you tell us more about this product, and what are the specific features?

AITA's "Smart Crowd" tool is the result of the large-scale model application innovation of Daily Interactive in the field of brand marketing, and we all affectionately call AITA "Love It" internally, which is the pinyin of AITA's 4 letters.

The first part of AITA is "AI" (Artificial Intelligence), which means that it uses advanced artificial intelligence algorithms represented by large models. "TA" is a common targeting audience in the marketing industry. In the AITA product, we:The large model is used to give it an understanding of natural semantics, and the data weaving technology is used to realize the semantics of the data space, so that the whimsical ideas of practitioners in marketing and other fields can be transformed into data entities and data labels in the data space through dialogue with large models, and then correspond to the target populationInnovative features such as "seedless insights" and "conversational promoting" have been implemented

The user experience of "AITA Smart Crowd" is similar to that of the AI model, which is mainly based on questions and answers, and also integrates more rigorous crowd definition and estimation functions, so that customers can use AITA with both a relaxed and lively dialogue process and a rigorous and efficient quantitative part, so that they can complete the process of market analysis, insight and docking more smoothly, and also allow the marketing field to enjoy the dividends brought by AI earlier.

03. Daily Interactive has an intelligent marketing SaaS product "Shupan", how to understand the relationship between it and the "AITA Smart Crowd" tool?

Shupan is an intelligent marketing SaaS product that Daily Interactive has been maintaining and iterating. Since its birth, Shupan has provided more than 100 brand owners with digital brand marketing services such as consumer insight, market analysis, crowd selection and ** docking, and has a very good reputation.

Figure: The product interface of the number plate.

In the past, Datapan mainly provided crowd insight and crowd targeting services through labeling and modeling, but "AITA Smart Crowd" adopts an innovative way of AI targeting crowds, which is also a new capability upgrade of Datapan.

With the help of large models and industry semantic weaving technology, AITA Smart Crowd allows customers to self-service industry insights and crowd selection on digital disks through the interaction mode of natural language dialogue, and obtain real-time feedback, so that customers can work more efficiently and have a lower threshold for operation. In the future, AITA will complement and coexist with traditional methods such as labeling and data modelingJointly serve the needs of brand owners in all scenarios and at all levels, so that the company's brand marketing services are more comprehensive and thoughtful.

04. In terms of experience and effect, what kind of improvement do you think the "AITA Smart Crowd" tool can bring to brand owners?

AITA "Crowd Selection" deeply integrates the capabilities of large models with the company's digital intelligence capabilities to achieve AI-targeting TA (AI-targeting audience). With the two unique advantages of "seedless insight" and "conversational promotion", AITA has mainly brought two improvements to brand owners: one is to lower the operational threshold and improve marketing efficiency, and the other is to broaden the application scenarios of data in the field of brand marketing.

The traditional "label targeting" method requires brand marketers to check and combine labels according to their own cognition of consumers and their understanding of daily interaction data, while another "model targeting" method requires brands to provide one side of seed data as a premise, and daily interaction is analyzed and expanded through machine learning. These two methods have certain requirements for the professionalism of marketers, and may also involve more manual communication with the consulting team and data analysis team.

In some lightweight marketing scenarios that require real-time feedbackWith "AITA Smart Crowd", front-line marketers do not need technical background, they only need to put forward marketing demands in the dialog box anytime and anywhere like daily conversations, and AITA can quickly give their own marketing strategy suggestions and crowd targeting results. It can be seen that there is not only a free and flexible demand collection process, but also an accurate and quantitative data feedback process, and the whole process can be said to be a combination of art and science, which greatly reduces the threshold of data application and improves the efficiency of business operation.

At the same time,In some specific marketing scenarios such as new product promotionBrand owners often lack seed users and are not clear about their target consumer groupsAnother example,Some companies are struggling with a lack of dataIt is difficult to analyze and gain insight into consumers' characteristics and preferences from multiple dimensions, which in turn affects the accuracy of target targeting and restricts the improvement of brand digital marketing effect.

In these scenarios, brand owners can use "AITA Smart Crowd" to quickly generate target audiences that meet the needs of brand advertisers without providing seed data, and only need to use simple natural language conversations to help brand advertisers efficiently connect with consumers, solving the dilemma of brand owners lacking seed groups.

05. What is the feedback after the release of "AITA Smart Crowd"?

There was a lot of feedback and everyone "loved it". Among them, there are well-known dairy customers, as well as mother and baby, beauty, FMCG and snack food customers, all of whom have given us very positive and positive feedback.

In the process of use, customers will communicate with AITA in a self-service manner for projects that usually take time and energy, and the data analysis and extraction results that may take days or even a week can be shortened to a few minutes or hours to get a quick feedback, and some can even reach the second level of extraction. With the blessing of AITA, customers can get better queue services for many heavyweight and lightweight needs.

In this process, AITA has achieved a good combination of man and machine, it is like opening a door, through this door, people can enjoy higher performance algorithms through more efficient machines, and enter the wonderful world of data. On both sides of this door, actuallyA groundbreaking connection between human experience and the world of dataOn top of this connection, the customer can truly and intuitively experience the charm of the world of data, and the experience of being able to get feedback and response to his ideas in a very short period of time is very beautiful.

Customer feedback is what drives us forward. We hope to continue to iterate and polish this product with customers, and welcome more data sources and more customer scenarios to join us, so that AITA can integrate more data portrayals in the physical world, connect with more scenario applications, and better serve the marketing practices of brand customers."

06. In the field of brand marketing, why should daily interaction choose to start from the scene of crowd insight and precise targeting?

Algorithms, computing power, and data are the three major elements of AI large models, and high-quality data and algorithms are the scarce elements in the development of AI large models, which greatly affect the effect of large models in the industry. Daily Interaction (Individual Push) started with developer services and has a massive amount of data accumulation. Massive amounts of data are essential for consumer insights, which makes daily interactions a clear natural advantage in the field of brand marketing.

Plus, daily interactionsIt also has a very significant advantage in terms of domain know-how。Since the launch of brand marketing services in 2015, Daily Interactive has served more than 100 customers in FMCG, beauty, mother and baby, automotive and other industries, and has very rich practical experience in brand digital marketing.

Based on these advantages, we can provide more comprehensive, direct and accurate market quantitative statistics and domain know-how (industry knowledge) for training for the large model, so that our large model can better understand marketers, understand consumer groups, understand the market situation and even future trends, greatly improve the effectiveness and stability of the large model in the marketing field, so that the "AITA Smart Crowd" tool can truly bring value to brand owners.

07. From the beginning of the "100 model war" to the present, the large model has entered a new stage of application landing. In the future, what kind of exploration will daily interaction have in the field of large models?

Large models can now be described as "popular fried chickens", and companies with strength, computing power, and data are developing their own large models, and there is a great momentum of "arms race". But in fact, for the natural language, images and other objects currently processed by large modelsThe industry is looking forward to further iterations on the basis of public resources in a planned and organized manner

We think,The long-term value of the large model will be realized through industry applications, and scenarios are the key. 。Therefore, the company is also continuing to explore and promote the application of large models in various industries and fields. For example, inDeveloper Services AreaThis is equivalent to the combination and mutual iteration of the two large models of AITA and AIGC, through the repeated matching and optimization of people and content, so that APP developers and operators can find the target group more efficiently, and achieve better matching between user groups and users, content and creativityInIntelligent transportation fieldWe have pre-trained the physical models of people and space-time, vehicles and space-time, equipment and space-time, and created a platform that can more quantitatively and intelligently characterize and optimize the efficiency of traffic operation to the traffic police department, so that the daily commuting and travel of citizens are more convenient and smooth.

Daily Interactive also hopes to work with more industry partners to create and build new large model application scenarios and basic capabilities, so that large models can shine more brightly in industry applications, and data intelligence can benefit more industries and fields through the power of large models

Related Pages