This AIGC can not only write Xiaohongshu copywriting, but also do the development of new catering pr

The technical discussion and communication planning of AIGC and large models have been "noisy" for nearly a year.

But in the "Thousand Model War", how many players can make AIGC technology really land, so that AIGC can empower the marketing of small and medium-sized enterprises to truly enter the "capillaries" of the real economy?

Uncle's feeling is that there are very few!

It is precisely because there are few that we must talk about this case I heard yesterday at the "Seeing the Truth" conference hosted by my predecessor Xu Zhibin.

In the past, the consumption scene of crayfish was often the same as the depth of smoke and fire in the market, and in 2015, the brand called Pirate Shrimp Rice was born, subverting the dining scene of crayfish, with more than 20 Q-bomb crayfish meat, so that the working meal can also realize "crayfish freedom".

The consumption scene has been innovated, and the "signature crayfish rice" has become a popular item on the Internet, and more and more stores of pirate shrimp rice have been opened.

But in fact, in order to enrich SKUs and meet more choices for customers, Pirate Shrimp Rice has also developed many other flavors of rice bowls, such as Tom Yum Gong Shrimp Pot, Braised Pork Rice, and Sauerkraut Fish Rice, but the sales of crayfish rice have always occupied an absolute advantage.

Founder Liu Qinggang wanted to find a new growth curve.

But here comes the problem, before deciding what to cook, the information to assist Boss Liu's decision-making is very limited, and there are 3 core problems:

First, "Pirate Shrimp Rice" does not have a special research and analyst, and the boss can only pat his head. In fact, I have filmed it before, and it has been successful;

Second, in the new product development process of the traditional catering industry, the whole new product development cycle is long, the cost is high, and the fault tolerance rate is also low

Third, I finally made a new product and tested it in a single store, only the simple and crude data dimension of sales, and the consumer contact point was not only single, but also could not get real valuable feedback, such as why this new product did not sell well, Boss Liu only accepted the results, but could not find the key reason, and the business was not closed-loop.

Since the beginning of this year, Lao Liu has also heard a lot about how AI large model technology can help all walks of life reduce costs and increase efficiency, but he muttered in his heart: I am a small and medium-sized catering enterprise, although AIGC is popular, but for the traditional catering industry with a low degree of digitalization, is it reliable for AIGC to empower this?

With a skeptical attitude, Boss Liu opened the first step of this exploration and threw out the soul torture -

AI, AI, tell me, pork, beef, chicken, mutton, which one should I make?

At the "Seeing the Truth" scene, Li Jiao, general manager of the commercial marketing department of 360 Group, deeply restored the "problem-solving ideas" of 360ai digital employees, which was divided into three steps

Li Jiao, General Manager of Commercial Marketing Department of 360 Group.

Step 1: Establish an enterprise knowledge base, create "industry experts", and assist in category selection.

In addition to the core operating data of the enterprise, the third-party reports and external authoritative data that Boss Liu pays attention to on a daily basis are also uploaded to the enterprise knowledge base of 360ai digital employees, which is the so-called "feeding data".

Based on the in-depth Xi of these data, 360ai digital employees quickly grew into "AI product managers in the catering industry", and Boss Liu asked the above soul and sent it to TA.

The workflow for a F&B AI product manager is as follows:Analyze brand consumption data, analyze the characteristics of core consumer groups, analyze the consumption behavior of core groups, search for current consumption trends and hot words of Chinese fast food, and finally, output a new product development suggestion report, and the conclusion is - beef rice.

Step 2: A b test to assist in taste decision-making.

A new question is here again - Boss Liu's pirate shrimp rice has more than 100 stores across the country, distributed in different regions, for users with different taste preferences, what flavor of beef rice can gain the favor of the most taste buds?How to reach the widest range of target groups in the most efficient way and get real feedback?

In the past, Boss Liu needed to make every new product, and then take it to the store to sell, and then look at the data.

In the case of pirate shrimp rice,Lao Liu omitted the process of "cooking" and real shooting, and with the help of AI drawing tools - 360 intelligent drawing, two flavors of dishes were generated**, using massive user resources on the 360PC sideAt the node near lunch, the 360 intelligent painting generated by the excellent ** through the 360pc lock screen pictorialPush to white-collar workers in the four cities of Beijing, Shanghai, Shenzhen and HangzhouAsk them to choose directly.

Eventually, 480,000 clicks, 736% of people chose tomato fat beef rice!

Not only is the process lighter, but the larger sample size also provides assurance for test resultsThe time and cost of the entire test were reduced by 80% and 90%, respectively.

Step 3: 360AI digital human meets the needs of single-store testing.

Is the new product popular?The sales data of the single-store test is one thing, if you want to collect more real feedback, you need the store staff to conduct on-site communication or return visits, and the results obtained are not of high quality, and may also affect the customer's dining experience.

Through the method of 360AI digital human + large model, Pirate Shrimp Rice based on the AI digital human created by the brand IP Captain Uncle appeared in offline stores, and established "** communication" with customers through online user research questionnaires to gain valuable user feedback.

Furthermore, by enriching the knowledge base and replicating the experience and knowledge of the brand founder, "Boss Liu" can enter each offline store and assume functions such as customer service, product shopping guide, and nutrition consulting.

I never expected that AIGC was already so down-to-earth, and when I heard this, the uncle began to reflect: Why are my articles still codeword, why can't I use ChatGPT well?

Unexpectedly, there was something that made the uncle sweat even more.

Step 4: Xiaohongshu grass copywriting is generated, and the efficiency of social media operation is improved.

It turned out that the pirate shrimp rice Xiaohongshu official account that we saw also came from 360ai digital employee - Xiaohongshu copywriting expert.

How does TA do it?

First of all, you need to explain exactly what you need AI to help you solve, such as:

I want to generate a copy of the pirate shrimp rice new tomato fat beef rice Xiaohongshu grass copy, which requires highlighting the mellow taste, nutritious and delicious, about 500 words.

360AI digital employees operating in AI agent mode are not only "tools" to perform tasks, but also "agents" that can call external tools - through intent recognition, search for relevant information on the whole network, grasp the rules of popular models, write Xiaohongshu copywriting, and let the results be based on factual content, which solves the "illusion" problem of large models. By dismantling the complex experience into a structured process, the experience of copywriters is empowered to users, and copywriters can easily become Xiaohongshu operation experts.

The result?In the past, Boss Liu's social ** operation was only a little girl, and she wrote 3 Xiaohongshu articles a day at most, but now, she can produce 50 copywriting content for you to choose from at will.

It's no wonder that the first to be laid off were graphic designers and content operations ......

Not only that, as AI digital humans are becoming more and more landed, customer service is also about to be laid off.

Step 5: AI digital humans help acquire and retain customers.

As mentioned above, Boss Liu made n digital clones for himself, and used the "Uncle Captain" IP image created by 360 Zhihua, and uploaded different knowledge bases, so that he could be competent for different roles, such as user research, intelligent customer service, etc., which greatly reduced costs and improved efficiency.

Li Jiao told everyone that in fact, making a digital clone for the boss has one of the biggest benefits for brand public relations people, that is, the boss records ** for some public relations activities, and will never "talk nonsense" again.

As a long-term observer of brand public relations and marketing, Uncle still feels quite deeply. How can we really seize the wave of AIGC and use it for enterprises to reduce costs and increase efficiency?Uncle talks about 3 points of experience:

Only those who can "land" large models are valuable to "business".

This is the most impressive point.

Is it really able to help enterprises reduce costs and increase efficiency?This is a problem that all platforms participating in the "Thousand Model War" need to face and tell the story of "circle", especially in the current environment, enterprises are reducing employees, how to increase efficiency at the same time?

Uncle has seen a lot of AIGC application cases, and to be honest, they are either in the concept stage or are similar on a large scale. You get an AI drawing, and I'll follow up right away;He made an AI clip**, and I followed up immediately, and the volume was worthless.

Uncle believes that the large model empowers enterprise marketing, and it must help enterprises reduce costs and increase efficiency in business, so as to be valuable. It can be widely used in all walks of life, especially in small and medium-sized enterprises, and lower the threshold for the application of large models.

"Generative big models" are not a "panacea".

Uncle paid special attention to itLi Jiao said that 99% of enterprises do not have the ability to directly use large models to solve productivity. What do you mean?

No matter how good ChatGPT is, most companies, especially non-Internet companies that do not have the ability to develop large models, don't need to know ChatGPT at all. No matter how efficient the AIGC tool is, can the employees of the enterprise control it?

One of my senior brothers at Peking University HSBC was working on ChatGPT-related entrepreneurship projects, and he gave me a 3Version 5 of the account, let me try it, I used it a few times, and then gave up, because I couldn't even "ask questions".

The "360 intelligent painting" mentioned above is an example, in fact, it is difficult for non-professional designers or AI enthusiasts to generate ** that meets expectations through "simple questions", and360 Smart Drawing reduces the difficulty of AI mapping by providing massive models, and users only need to modify prompts and adjust parameters to generate **,This greatly improves the accuracy and controllability of AI mapping.

Uncle believes that lowering the threshold for the use of AIGC tools is the key to AIGC empowering more industries. It's like asking the leader to tell you how to write a good article, which is difficult, but it's much easier and more efficient if you write an outline and let the leader make suggestions for revision.

AI is "subverting" the traditional marketing chain.

From strategic insight, to marketing promotion, to sales conversion, and user operation, there are opportunities to combine with AIGC in the whole marketing chain, I hope you can accept the word "subversion" as soon as possible after reading this article, because only if you accept it, will you make changes!

Why did AIGC change the field of marketing in the first place?

First of all, the content forms generated by AIGC, such as text and audio, are needed by the marketing industry. The massively generated content also happens to be the key to the real realization of "thousands of people" in marketing under the superposition of complex media environment and increasingly segmented user preferences.

Secondly, it is becauseThe workflow of marketing content is relatively clear, and the task dismantling is smootherTo put it simply, it can be abstracted into three important steps: "search, read, and write", and the agent mode mentioned above also officially follows this workflow, searching for information on the whole network, reading relevant content, and finally writing the popular Xiaohongshu copy.

Take Task Teardown as an example.

An outstanding classmate of my uncle at Peking University's HSBC Business School, named Pan Lihua, is the head of marketing at Swire Coca-Cola, and he told me that when Coca-Cola was managing front-line marketing personnel, what these people were going to do when they arrived at the store were all disassembled into modules, all of which were digital operations, and the machine automatically sent out the work data of these people by email every day.

He also gave an example of Liu Xiang, Liu Xiang broke the world record, on the one hand, because of his good genes, on the other hand, his coaching team dismantled the whole process of hurdle running, and what Liu Xiang had to do was to practice Xi one by one.

Finally, the uncle made a brief summary. Let's go back to the core case of this article "Pirate Shrimp Rice", large model + enterprise knowledge base + agent + AI digital human, this set of "combination punches", AIGC is the super employee who understands customers best, and even the super boss, which may be the prototype of intelligent marketing in the future market!

If your perception is not so strong now, Uncle thinks that the only reason may be a matter of time, but you have seen it and used it, and the bonus is gone.