Online red marketing method and device based on large model, computer equipment and readable storage medium

Abstract

The invention discloses an online red marketing method and device based on a large model, computer equipment and a readable storage medium, and the method comprises the steps: firstly obtaining an online red marketing plan, a tool library containing a plurality of tools, and a knowledge library, inputting the marketing plan into the large model to obtain a screening condition, and carrying out the screening through a search tool to obtain a preliminary result; and calling a data analysis tool by using a large model to analyze the preliminary result to obtain a target account, querying an email address of the target account, and then carrying out negotiation interaction. If negotiation succeeds, a negotiation process file is input into the large model, so that a network-red cooperation scheme is obtained, network-red information, quotation, publishing time and the like are covered, and efficient and accurate network-red marketing is achieved.

Description

Technical Field

[0001] The present invention relates to the field of large model applications, and in particular to a large model-based internet celebrity marketing method, device, computer equipment and readable storage medium. Background Art

[0002] With the development of the Internet, influencer marketing has become an important way for companies to promote their products and services. The traditional influencer marketing process often relies on manual operations, which has problems such as low efficiency, inaccurate influencer screening, and cumbersome cooperation negotiation process. Summary of the invention

[0003] The purpose of the present invention is to provide an internet celebrity marketing method, device, computer equipment and readable storage medium based on a large model.

[0004] In a first aspect, an embodiment of the present invention provides an Internet celebrity marketing method based on a large model, comprising: Obtaining an influencer marketing plan, an influencer marketing tool library, and an influencer marketing knowledge base; the influencer marketing tool library is equipped with an influencer search tool, an influencer data analysis tool, an influencer email query tool, and an email sending and receiving tool; Input the influencer marketing plan into a preset influencer macro model to obtain influencer screening conditions for the influencer marketing plan; Calling the influencer search tool, screening in the influencer marketing knowledge base based on the influencer screening conditions, and obtaining preliminary influencer screening results; Calling an internet celebrity data analysis tool through the preset internet celebrity model to analyze the preliminary screening results of the internet celebrities and obtain target internet celebrity accounts; The preset internet celebrity model is used to call the internet celebrity email query tool to obtain the email address of the target internet celebrity account, and the email sending and receiving tool is used to negotiate and interact with the email address; On the basis that the negotiation interaction result is successful, the process file of the negotiation interaction is input into the preset Internet celebrity model to obtain the Internet celebrity cooperation plan of the Internet celebrity marketing plan.

[0005] In a possible implementation, the influencer marketing plan includes influencer marketing region information, influencer language, influencer type, total number of influencers, total budget, influencer marketing time period, promotion product name and introduction, influencer label, influencer keyword, and influencer cooperation negotiation deadline; The step of inputting the influencer marketing plan into a preset influencer macro model to obtain influencer screening conditions for the influencer marketing plan includes: Input the internet celebrity marketing area information, the internet celebrity language, the internet celebrity type, the total number of internet celebrities, the total budget, the internet celebrity marketing time period, the name and introduction of the promoted product, the internet celebrity label, the internet celebrity keyword, and the internet celebrity cooperation negotiation deadline into the preset internet celebrity macro model to obtain initial screening conditions; The initial screening condition is assigned a condition that the email address is not empty, so as to obtain the influencer screening condition for the influencer marketing plan.

[0006] In a possible implementation, the calling of an internet celebrity data analysis tool through the preset internet celebrity model to analyze the internet celebrity preliminary screening results to obtain a target internet celebrity account includes: The internet celebrity data analysis tool is called through the preset internet celebrity big model, and the preliminary screening results of the internet celebrities are analyzed based on the product matching index and the internet celebrity instruction index to obtain the target internet celebrity account.

[0007] In a possible implementation, the interacting with the email address through the email sending and receiving tool includes: The email sending and receiving tool is used to conduct at least one round of interaction with the email address to confirm the Internet celebrity's willingness to cooperate, discuss the cooperation price, and agree on the release time of the Internet celebrity's work, so as to complete the negotiation interaction.

[0008] In a possible implementation, the method further includes: Obtaining the negotiation interaction result, wherein the negotiation interaction result includes total marketing data and total amount; Determining whether the total marketing data and / or the total amount meets the internet celebrity marketing plan; If yes, then the negotiation interaction result is determined to be successful; If not, the negotiation interaction result is determined to be a failure.

[0009] In a possible implementation, the method further includes: On the basis that the negotiation interaction result is a failure, the Internet celebrity keywords are expanded, and the step of inputting the Internet celebrity marketing plan into a preset Internet celebrity model to obtain Internet celebrity screening conditions for the Internet celebrity marketing plan is returned to be executed.

[0010] In a possible implementation, the step of inputting the negotiation interaction process file into the preset influencer model to obtain the influencer cooperation plan of the influencer marketing plan includes: The negotiation interaction process file is input into the preset internet celebrity model to obtain the internet celebrity cooperation plan of the internet celebrity marketing plan consisting of cooperation internet celebrity information, cooperation quotation, work release time and content description.

[0011] In a second aspect, an embodiment of the present invention provides a large model-based internet celebrity marketing device, including: An acquisition module is used to acquire an influencer marketing plan, an influencer marketing tool library and an influencer marketing knowledge base; the influencer marketing tool library is configured with an influencer search tool, an influencer data analysis tool, an influencer email query tool and an email sending and receiving tool; the influencer marketing plan is input into a preset influencer big model to obtain influencer screening conditions for the influencer marketing plan; the influencer search tool is called to perform screening in the influencer marketing knowledge base based on the influencer screening conditions to obtain an influencer preliminary screening result; the influencer data analysis tool is called through the preset influencer big model to analyze the influencer preliminary screening result to obtain a target influencer account; the influencer email query tool is called through the preset influencer big model to obtain an email address of the target influencer account, and the email sending and receiving tool is used to negotiate and interact with the email address; The marketing module is used to input the process file of the negotiation interaction into the preset Internet celebrity model on the basis that the negotiation interaction result is successful, so as to obtain the Internet celebrity cooperation plan of the Internet celebrity marketing plan.

[0012] In a third aspect, an embodiment of the present invention provides a computer device, comprising a processor and a non-volatile memory storing computer instructions, wherein when the computer instructions are executed by the processor, the computer device executes the method described in the first aspect.

[0013] In a fourth aspect, an embodiment of the present invention provides a readable storage medium, wherein the readable storage medium includes a computer program, and when the computer program is executed, the computer device where the readable storage medium is located is controlled to execute the method described in the first aspect.

[0014] Compared with the prior art, the beneficial effects provided by the present invention include: adopting a big model-based internet celebrity marketing method, device, computer equipment and readable storage medium disclosed in the present invention, by obtaining the internet celebrity marketing plan, a tool library and knowledge base containing multiple tools, inputting the marketing plan into the big model to obtain screening conditions, and using the search tool to screen and obtain preliminary results. Then use the big model to call the data analysis tool to analyze the preliminary results to obtain the target account, query its email address and then conduct negotiation interaction. If the negotiation is successful, input the negotiation process file into the big model to obtain an internet celebrity cooperation plan, covering internet celebrity information, quotation, release time, etc., to achieve efficient and accurate internet celebrity marketing. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only illustrate certain embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can also be obtained based on these drawings without creative work.

[0016] Figure 1 A schematic diagram of the steps of the Internet celebrity marketing method based on a large model provided by an embodiment of the present invention; Figure 2 A system framework diagram of the Internet celebrity marketing method based on a large model provided by an embodiment of the present invention; Figure 3 A schematic structural diagram of an Internet celebrity marketing device based on a large model provided by an embodiment of the present invention; Figure 4 A schematic block diagram of the structure of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0018] The specific implementation modes of the present invention are described in detail below in conjunction with the accompanying drawings.

[0019] In order to solve the technical problems in the aforementioned background technology, Figure 1 A flowchart of an internet celebrity marketing method based on a big model is provided in an embodiment of the present disclosure. The internet celebrity marketing method based on a big model is introduced in detail below.

[0020] Step S201, obtaining an influencer marketing plan, an influencer marketing tool library, and an influencer marketing knowledge base; the influencer marketing tool library is configured with an influencer search tool, an influencer data analysis tool, an influencer email query tool, and an email sending and receiving tool; Step S202, inputting the influencer marketing plan into a preset influencer macro model to obtain influencer screening conditions for the influencer marketing plan; Step S203, calling the Internet celebrity search tool, screening in the Internet celebrity marketing knowledge base based on the Internet celebrity screening conditions, and obtaining a preliminary screening result of Internet celebrities; Step S204, calling an internet celebrity data analysis tool through the preset internet celebrity model, analyzing the preliminary screening results of the internet celebrities, and obtaining a target internet celebrity account; Step S205, calling the internet celebrity email query tool through the preset internet celebrity model to obtain the email address of the target internet celebrity account, and conducting negotiation and interaction with the email address through the email sending and receiving tool; Step S206, based on the successful negotiation interaction result, input the negotiation interaction process file into the preset Internet celebrity model to obtain the Internet celebrity cooperation plan of the Internet celebrity marketing plan.

[0021] In an embodiment of the present invention, exemplarily, the server starts an influencer marketing program. First, an influencer marketing plan is obtained from local storage or external data sources. This plan may be formulated in advance by the marketing team and uploaded to a designated location on the server. For example, a cosmetics company wants to promote a new lipstick product in Europe. The marketing plan clearly states that the target countries / regions are European countries such as France and Germany, the influencer languages ​​are local languages ​​such as French and German, the influencer types are beauty influencers, the total number of influencers planned to cooperate is 10, the total budget is US$50,000, the influencer marketing period is within the next two months, the name of the promoted product is "Brilliant Lipstick", the product introduction is "rich in natural ingredients, long-lasting and moisturizing", and the deadline for influencer cooperation negotiations is one week later.

[0022] At the same time, the server also loads the influencer marketing tool library, which includes influencer search tools, influencer data analysis tools, influencer email query tools, email sending and receiving tools, etc. These tools are initialized when the server starts and can be called at any time. The influencer marketing knowledge base is also obtained, which stores various knowledge related to influencer marketing, such as terms, concepts, influencer screening, influencer portraits, etc.

[0023] The server inputs the above-obtained influencer marketing plan into the preset influencer big model. The big model generates influencer screening conditions based on the input plan content. Taking the previous lipstick promotion as an example, the screening conditions generated by the big model may include: on the social media platforms in France and Germany, search for beauty influencers whose languages ​​are French and German, and whose fans are between 100,000 and 500,000 (preliminarily determined based on the budget and target audience), and require that the content recently released by the influencer is highly relevant to beauty. The big model will also add a condition that the influencer's contact email address is not empty to ensure subsequent email communication.

[0024] The server calls the influencer search tool and passes the generated screening criteria. The influencer search tool searches and filters in the influencer marketing knowledge base. For example, the search tool searches for multiple influencer platform data that meet the criteria in the knowledge base. On a French beauty social platform, it found 20 influencers who met the conditions of French language, beauty influencers, and 100,000-500,000 followers. Fifteen influencers who met the criteria were also found on a similar platform in Germany. The basic information of these influencers, such as influencer accounts, approximate number of followers, and simple descriptions of recently released content, constitute the preliminary screening results of influencers.

[0025] The server uses the Internet celebrity data analysis tool to analyze the 35 Internet celebrities that were initially selected through the Internet celebrity big model. For each Internet celebrity, the data analysis tool will analyze their fan profile in detail, such as age distribution, gender ratio, etc.; Internet celebrity fan interaction rate, such as the ratio of likes, comments, and shares; Internet celebrity value, which takes into account factors such as the number of fans, interaction rate, and influence; Internet celebrity activity, which includes the frequency of checking recent content releases, etc.

[0026] Suppose that 5 of the influencers have fans that meet the requirements, but their fan interaction rate is lower than the average level and their activity is not high. This may be because they have many zombie fans or their activity has decreased recently. In addition, the fan portraits of 3 influencers match the target audience of lipstick products by less than 80%. For example, their fans are mainly male, while lipstick products are mainly for female consumers. The large model eliminates these 8 influencers based on the conditions of active influencers, real and effective fans, and more than 80% match between fan portraits and products. The remaining 27 influencer accounts are the target influencer accounts.

[0027] The server uses the influencer email query tool through the influencer model to query the contact email addresses of the 27 target influencer accounts. For example, the email address of a French influencer is beautyfrench@example.com, and the email address of a German influencer is germanglam@example.com, etc., and the email addresses of all target influencers are obtained.

[0028] The large model generates a personalized cooperation invitation email for each influencer based on the influencer information. Taking the French influencer as an example, the content of the email is: "Dear [influencer name], we are [company name], and we are carrying out a promotion called 'Brilliant Lipstick'. This lipstick is rich in natural ingredients, long-lasting and moisturizing, and very suitable for European women. As an influential influencer in the French beauty field, your fan base is highly matched with the target audience of our products. We sincerely hope to cooperate with you for a fee. If you are interested, you can contact us by email to discuss the specific details of the cooperation." Then the email is sent to each influencer through the email sending and receiving tool.

[0029] The big model calls the Internet celebrity quotation query tool to determine the lower and upper limits of each Internet celebrity's cooperation quotation. For this Internet celebrity in France, the lower limit of the quotation is 500 US dollars and the upper limit is 1000 US dollars. The big model first bids 500 US dollars and informs the Internet celebrity by email. If the Internet celebrity replies with a quotation of 800 US dollars, under the total budget constraint and not exceeding the quotation upper limit, the big model can accept this price and reply to confirm. If the Internet celebrity quotes 1200 US dollars, which exceeds the upper limit, the big model will increase the price appropriately to 700 US dollars (hypothetically) within the total budget according to the bargaining strategy, reply to the Internet celebrity again, and make a second quotation. Generally, the bargaining will be 2-5 times. If it exceeds 5 times or the Internet celebrity's quotation is too high, the bargaining will be euphemistically ended and the cooperation negotiation will be terminated.

[0030] Assuming that a price negotiation is successful with a certain influencer, the big model will combine the marketing activity plan (such as the best time period for product promotion), the influencer's personal characteristics (such as the peak time period for the influencer to publish content) and demand to agree on a release time that is acceptable to both parties. For example, the marketing activity plans to have a large-scale promotion on a certain weekend, and hopes that the influencer will publish content a week before the promotion. The influencer said that she also has a large number of fans during that period, so the two parties agree to publish the work on Wednesday of the week before the promotion.

[0031] After the server completes the negotiation interaction with multiple influencers and the result is successful (i.e., a sufficient number of influencers that meet the budget are determined), the server inputs the entire negotiation interaction process file (including the content of email exchanges with each influencer, the final cooperation quotation, release time, etc.) into the preset influencer big model. The big model organizes and summarizes this information and outputs the influencer cooperation plan. For example, the plan identifies 5 French influencers and 3 German influencers as cooperation objects, and lists the cooperation quotation of each influencer, such as French influencer A quoted US$600, German influencer B quoted US$800, etc. At the same time, the release time of each influencer's work is clarified, such as French influencer A released on [specific date 1], German influencer B released on [specific date 2], and briefly explains the content requirements of each influencer's work, such as introducing the color number of lipstick, usage experience, etc.

[0032] In an embodiment of the present invention, the internet celebrity marketing plan includes internet celebrity marketing area information, internet celebrity language, internet celebrity type, total number of internet celebrities, total budget, internet celebrity marketing time period, promotion product name and introduction, internet celebrity label, internet celebrity keywords and internet celebrity cooperation negotiation deadline; the internet celebrity marketing plan is input into a preset internet celebrity big model to obtain internet celebrity screening conditions for the internet celebrity marketing plan, which can be implemented through the following example.

[0033] Input the internet celebrity marketing area information, the internet celebrity language, the internet celebrity type, the total number of internet celebrities, the total budget, the internet celebrity marketing time period, the name and introduction of the promoted product, the internet celebrity label, the internet celebrity keyword, and the internet celebrity cooperation negotiation deadline into the preset internet celebrity macro model to obtain initial screening conditions; The initial screening condition is assigned a condition that the email address is not empty, so as to obtain the influencer screening condition for the influencer marketing plan.

[0034] In an embodiment of the present invention, for example, assume that a sportswear company wants to promote its new series of sports running shoes. In the influencer marketing plan formulated by the company's marketing team, the influencer marketing region information is North America, specifically the United States and Canada. When the server inputs this regional information into the preset influencer big model, the big model will identify popular social media platforms related to this region, such as Instagram and TikTok, which have a large number of users in the United States and Canada, and different regions may have different cultural and consumption habits. Differences. For example, American sports enthusiasts may prefer influencer content related to sports such as American football and basketball, while Canadians may be more interested in influencers related to sports such as ice hockey and skiing. Based on these factors, the big model will preliminarily determine that when searching for influencers, it is important to focus on influencers related to these popular sports and active in the region.

[0035] Since the marketing area is the United States and Canada, the language of influencers is mainly English. After receiving this information, the big model will set the language to English when searching for influencers. This means that on various social media platforms, only those influencers who mainly create content in English will be screened. For example, on some multilingual platforms, there may be influencers who publish content in multiple languages, but the big model will exclude those influencers who are not mainly English-speaking based on this language requirement. This ensures that influencers can accurately convey the product information of sports running shoes to the target audience, because the product's promotional materials, advertising slogans, etc. are all prepared in English.

[0036] For the promotion of sports running shoes, the type of influencer specified in the marketing plan is sports influencer. The big model will further refine the screening conditions based on this requirement. Within the category of sports influencers, they can be further subdivided into different subtypes such as influencers who are professional athletes, influencers who have transformed from fitness enthusiasts, and outdoor sports experts. The big model will take into account the characteristics of sports running shoes. For example, if it is a pair of running shoes suitable for long-distance running, it may be more inclined to screen influencers who often participate in long-distance running activities or share long-distance running experiences; if it is a multi-functional sports shoe, it may look for influencers who cover a variety of sports experience. Therefore, under the general framework of sports influencers, the big model will determine more specific influencer type preferences based on product characteristics.

[0037] Assume that the total number of influencers set in the marketing plan is 20. The big model will adjust the scope and accuracy of the screening based on this number requirement. If the total number requirement is large, some screening conditions may be appropriately relaxed to ensure that a sufficient number of influencers can be found; if the number is small, the strictness of the screening conditions will be increased. For example, in this case, the big model may set a relatively wide range of fans in the initial screening so that more potential influencer candidates can be obtained, but then conduct more refined screening in subsequent screening and evaluation.

[0038] The total budget is set at $50,000. The big model will evaluate the possible cooperation cost of each influencer based on this budget. First, it will refer to historical data or industry averages to estimate the approximate cooperation price of influencers of different types, different numbers of fans and influence. For example, in the United States, top sports influencers with more than one million fans may have a cooperation quote ranging from $5,000 to $10,000, while influencers with medium numbers of fans (100,000-500,000 fans) may have a quote of around $1,000-3,000. Based on the total budget of $50,000, the big model will preliminarily determine the proportion of influencers to be sought at different levels of fans and influence to ensure that the cooperation plan is completed within the budget.

[0039] The influencer marketing time period is set to the next three months. The big model will consider the compatibility of this time period with the influencer's schedule. Some influencers may have arranged other collaborations or activities in advance and cannot participate in new collaborations during this time period. When screening influencers, the big model will give priority to those who have free time during this time period. For example, for some influencers who often participate in large-scale sports events, if the event happens to be during the marketing time period, they may not have time to take care of product promotion, and the big model will exclude these influencers. At the same time, the big model will also determine the best time for influencers to publish content based on the rhythm of product promotion, such as whether there are new product launches, promotional activities, etc., so that when screening influencers, it is more inclined to those influencers who can publish content at the right time.

[0040] The product name is "Vigorous Running Shoes", and it is introduced as "adopting advanced shock absorption technology, light and breathable, suitable for running on various surfaces". The big model will deeply analyze the characteristics of the product and match these characteristics with the content style and audience of the influencer. For the shock absorption technology of "Vigorous Running Shoes", the big model will look for influencers who often emphasize the sense of technology of sports equipment in their content; for the characteristics of lightness and breathability and suitable for running on various surfaces, it will tend to select influencers who focus on the running experience and often share different running scenes. This ensures that the influencer can better show the advantages of the product to fans.

[0041] The influencer tags are set as "running", "fitness", and "sports equipment", and the influencer keywords include "high-performance running shoes" and "comfortable running". The big model will use these tags and keywords to conduct precise searches on social media platforms. For example, on Instagram, the big model will search for influencers who include these tags in their personal profiles or frequently use these keywords in previous posts. This helps narrow the search scope and directly target those influencers who are highly relevant to the product.

[0042] The deadline for cooperation negotiation with influencers is set to two weeks later. The big model will arrange the entire process of influencer screening and negotiation based on this time limit. When screening influencers, those who respond to emails or messages in a timely manner will be given priority. For example, some influencers may be slow to respond to cooperation negotiation messages due to their busy personal affairs. The big model will take into account the negotiation deadline and make appropriate weight adjustments to these influencers during the initial screening, preferring those who can respond to negotiation requests in a short time.

[0043] The big model combines all the information in the above influencer marketing plan to get the initial screening conditions. For example, on Instagram and TikTok platforms, search for influencers located in the United States and Canada, speaking English, sports influencers (focusing on long-distance running and multi-functional sports), with a fan base initially set according to the budget (such as 100,000-1 million fans), free time in the next three months, matching the product features of "energized running shoes" (emphasizing technology, running experience, etc.), containing "running", "fitness", "sports equipment" tags or keywords such as "high-performance running shoes" and "comfortable running" in their personal profiles or posts, and responding to messages in a timely manner.

[0044] After the server obtains the initial screening conditions in the big model, it assigns the condition that the email address is not empty to these conditions. This is very important because the subsequent cooperation negotiations with influencers are mainly conducted through emails. For example, the big model initially screened out 100 influencers who met all the previous conditions (region, language, type, schedule, etc.), but some of them may not have a public contact email. After adding the condition that the email address is not empty, the big model will further screen these 100 influencers. Assuming that 20 of them do not have a public email address, 80 influencers will be obtained as the result of meeting the influencer screening conditions. These 80 influencers become the target objects for the next step (such as influencer data analysis), ensuring that the subsequent cooperation negotiation process can proceed smoothly, because only with the email address can the invitation email be sent and the details of the cooperation be discussed.

[0045] In the embodiment of the present invention, calling the internet celebrity data analysis tool through the preset internet celebrity model, analyzing the preliminary screening results of the internet celebrities, and obtaining the target internet celebrity account can be implemented through the following example.

[0046] The internet celebrity data analysis tool is called through the preset internet celebrity big model, and the preliminary screening results of the internet celebrities are analyzed based on the product matching index and the internet celebrity instruction index to obtain the target internet celebrity account.

[0047] In an embodiment of the present invention, exemplarily, the server calls the Internet celebrity data analysis tool through a preset Internet celebrity model, and starts to analyze the preliminary screening results of Internet celebrities based on the product matching index. First consider fan portrait matching. Assuming that 80 Internet celebrity accounts have been obtained after preliminary screening, for one of the Internet celebrities, the data analysis tool will check its fan portrait in detail. Taking the promotion of sports running shoes as an example, the age distribution, gender ratio, geographical distribution and other information of the fans of the Internet celebrity are all taken into consideration. If the target audience of sports running shoes is men aged 18-35, they are mainly distributed in urban areas and are keen on fitness. The data analysis tool found that among the fans of this Internet celebrity, 60% are aged 18-35, 70% are male, and most of the fans are located in urban areas and pay more attention to fitness topics. Then this Internet celebrity has a higher product matching degree in terms of fan portrait matching. As for another internet celebrity, the age distribution of his fans is more scattered, with those aged 18-35 accounting for only 30% and males accounting for 40%. In terms of geographical distribution, fans in rural areas account for a larger proportion, and their attention to fitness topics is relatively low. Therefore, the product matching degree of this internet celebrity in this indicator is relatively low.

[0048] In addition to fan portraits, fan interest matching is also an important consideration. Continuing with the example of sports running shoes, the data analysis tool will analyze the degree of interest of the fans of the influencer in sports-related topics. For a certain influencer, the content he publishes often involves various sports events, fitness tips, etc. The interactive data of fans shows that the proportion of likes, comments and shares on sports topics is very high. For example, a video about a new running shoe review released by the influencer has 10,000 likes, 500 comments, and 300 shares, which shows that fans are very interested in sports equipment content and highly match the sports running shoes to be promoted. On the contrary, for another influencer, although he is a sports influencer, his fans pay more attention to his personal life anecdotes and interact less with sports equipment-related content. For example, the sports equipment recommendation video he released has only 500 likes, 50 comments, and 20 shares. Then the product matching degree of this influencer in terms of fan interest matching is low.

[0049] The promotion of sports running shoes requires influencers to present the product in a suitable content style. Data analysis tools will evaluate whether the influencer's content style matches the product. For example, the content style of a certain influencer is mainly professional and detailed product reviews. He will deeply analyze the various technical parameters and wearing experience of running shoes. This content style is very suitable for promoting sports running shoes that emphasize advanced shock absorption technology and versatility. The content style of another influencer is more humorous and funny sports challenge videos. Although they are related to sports, it is difficult to show the characteristics of running shoes in detail, and the product matching degree in terms of content style matching is relatively low.

[0050] The server continues to call the data analysis tool through the Internet celebrity model and conducts analysis based on the Internet celebrity instruction indicators. First, look at the activity of the Internet celebrity, which is an important indicator to measure whether the Internet celebrity can continue to provide exposure for product promotion. For each Internet celebrity who is initially screened, the data analysis tool will check the frequency of his posting on the social media platform. For example, an Internet celebrity has posted 15 sports-related content in the past month, an average of 3-4 posts per week, which shows that he is more active. Another Internet celebrity has only posted 5 pieces of content in the past month, and his activity is relatively low. For the promotion of sports running shoes, which require a certain frequency of exposure, Internet celebrities with high activity are more likely to provide sufficient publicity for the product during the marketing period, so they have an advantage in this indicator.

[0051] The influence of influencers is also an important part of influencer directive indicators. Data analysis tools will comprehensively consider factors such as the number of followers of influencers and the fan interaction rate to evaluate their influence. Taking the number of followers as an example, an influencer with 500,000 followers theoretically has a larger sphere of influence than an influencer with 100,000 followers. But at the same time, the fan interaction rate cannot be ignored. If an influencer with 500,000 followers has an interaction rate of only 5% in terms of likes, comments, and shares, while an influencer with 100,000 fans has an interaction rate of 15%, then the latter may not be inferior in terms of influence despite having fewer followers. For the promotion of sports running shoes, it is necessary to find influencers with both a certain number of followers and a high interaction rate, so as to ensure that product information can be widely disseminated and attract the attention and interest of fans.

[0052] In influencer marketing, the authenticity of influencers is crucial. Data analysis tools use some methods to evaluate the authenticity of influencers. For example, check whether the influencer's fan growth curve is natural. If the number of followers of an influencer suddenly increases in a short period of time, and there is no reasonable explanation (such as no major events or high-quality content releases), there may be suspicion of buying fans. In addition, if the comments of influencers are analyzed, if there are a lot of repetitive, meaningless or obviously robot-generated comments, it will also affect the authenticity evaluation of influencers. For the promotion of sports running shoes, choosing real influencers can build consumer trust in the product. A real influencer is more likely to recommend products that are accepted by fans, while an influencer with questionable authenticity may have a negative impact on the product.

[0053] The server conducts a comprehensive evaluation of the 80 initially screened influencers through the above analysis based on the product matching index and the influencer instruction index. For example, for each influencer, a certain weight will be set for each sub-indicator in the product matching index and the influencer instruction index. Assume that the weight of the product matching index is 0.6 and the weight of the influencer instruction index is 0.4. In the product matching index, the weights of fan portrait matching, fan interest matching, and influencer content style matching are 0.3, 0.3, and 0.4, respectively; in the influencer instruction index, the weights of influencer activity, influence, and authenticity are 0.3, 0.4, and 0.3, respectively.

[0054] For each influencer, the score is calculated based on their performance in each sub-indicator. For example, a certain influencer scored 0.8 (indicating an 80% match) in terms of fan portrait matching, 0.7 in terms of fan interest matching, and 0.85 in terms of influencer content style matching; 0.75 in terms of influencer activity, 0.8 in terms of influencer influence, and 0.9 in terms of authenticity. The comprehensive score of this influencer is calculated as follows: Product matching score = 0.3×0.8+0.3×0.7+0.4×0.85=0.79.

[0055] Internet celebrity instruction index score = 0.3×0.75+0.4×0.8+0.3×0.9=0.805.

[0056] Comprehensive score = 0.6×0.79+0.4×0.805=0.796.

[0057] The server selects the influencer accounts with a score higher than 0.7 as the target influencer accounts according to the set comprehensive score threshold (assuming it is 0.7). After calculation, there may be 50 influencers with a comprehensive score higher than 0.7. These 50 influencer accounts become the target influencer accounts, and subsequent operations such as email query and cooperation negotiation can be carried out.

[0058] In the embodiment of the present invention, the negotiation and interaction with the email address through the email sending and receiving tool can be implemented through the following examples.

[0059] The email sending and receiving tool is used to conduct at least one round of interaction with the email address to confirm the Internet celebrity's willingness to cooperate, discuss the cooperation price, and agree on the release time of the Internet celebrity's work, so as to complete the negotiation interaction.

[0060] In the embodiment of the present invention, the server calls the email sending and receiving tool through the preset influencer big model, and starts to interact with the target influencer account to confirm the influencer's willingness to cooperate. Taking the 50 target influencer accounts previously screened as an example, for one of the influencer accounts, it is assumed that its email address is "sportinfluencer@example.com". The big model will generate a personalized cooperation invitation email based on the influencer's personal information, fan status, and the characteristics of the sports running shoes to be promoted.

[0061] The email content is as follows: "Dear [Internet celebrity name], hello! We are [sportswear company name] and we are carrying out an exciting promotion for 'Vibrant Running Shoes'. 'Vibrant Running Shoes' use advanced shock absorption technology, are lightweight and breathable, and are suitable for running on all kinds of surfaces. As an Internet celebrity with great influence in the field of sports, your fan base is highly consistent with the target audience of our products. The professional sports knowledge and wonderful sports experience videos you often share are deeply loved by fans, which is very consistent with the product concept we want to convey. We sincerely hope to carry out paid cooperation with you to introduce this excellent running shoe to more sports enthusiasts. If you are interested in this cooperation opportunity, please reply to the email and we can further discuss the specific details of the cooperation." The server sends this email to the influencer's email address through an email sending and receiving tool.

[0062] After receiving the email, if the influencer is interested in the cooperation, he or she may reply to express his or her willingness to learn more. For example, the influencer replies: "Hello, thank you for your invitation. I am very interested in 'Vitality Running Shoes' and would like to know more about the specific content of the cooperation, such as the promotion form, compensation, etc." After receiving this reply, the server will record the influencer's reply information and mark this influencer as an influencer with cooperation intentions.

[0063] If the influencer is not interested in the cooperation, the reply will be something like: "I'm sorry, I already have a lot of cooperation projects going on and I don't have the energy to participate in new cooperation for the time being." The server will also record this reply and remove the influencer from the subsequent cooperation negotiation list.

[0064] During this process, the server will continue to monitor the responses of all target influencer accounts. For some influencers who do not respond in time, the server will send a reminder email again according to the preset waiting time (for example, three days), with the content: "Dear [influencer name], we previously sent you an email about the promotion cooperation of 'Vitality Running Shoes'. I don't know if you have read it. We are very much looking forward to working with you. If you have any questions or ideas, please reply to the email and let us know." For those influencers who express their intention to cooperate, the server calls the influencer cooperation quotation query tool through the influencer big model to obtain the lower and upper limits of each influencer's cooperation quotation. For example, for the influencer who responded to the interest before, the query found that the lower limit of the cooperation quotation was US$800 and the upper limit was US$1,500. Based on the marketing budget and strategy, the big model made an initial bid of US$800 and sent the influencer price-related negotiation content via email: "Dear [influencer name], thank you for your interest in the promotion cooperation of 'Vibrant Running Shoes'. Our preliminary quotation for this cooperation is US$800. This price is based on our assessment of the market and your fan influence and other factors. Your fan base is highly matched with the target audience of our products, and we believe that this cooperation will be of great benefit to both parties. If you have any thoughts on this quotation, please reply to the email and let us know." If the influencer replies that he agrees to the offer, for example, "Hello, I can accept the offer of $800, and we can discuss other details of the cooperation." The server will record the price information and mark that the influencer has reached a preliminary agreement on the price.

[0065] If the influencer replies that he disagrees with the offer and asks for a higher price, for example: "Hello, the offer of $800 is a bit low. I think the price corresponding to my work content and fan influence should be around $1,200." The server will process the response based on the influencer's reply. The big model will increase the price appropriately within the influencer's own offer ($1,200), the offer cap ($1,500), and the total budget constraints. Assuming that there is enough room in the total budget, the big model may reply: "Dear [influencer's name], considering your influence and work content, we can increase the offer to $1,000. This price is already a relatively reasonable price we can offer within the budget, and I hope you can consider it." During this process, if the influencer is still not satisfied with the new offer and continues to raise the price, the server will continue to interact with the influencer according to the bargaining strategy. However, if the number of bargaining exceeds 2-5 times (assuming 5 times) or the influencer's offer exceeds the upper limit, the server will send a tactful email to end the bargaining, for example: "Dear [influencer's name], unfortunately we cannot reach an agreement on the price. We thank you very much for your attention to this cooperation and hope to have the opportunity to cooperate again in the future." and remove the influencer from the cooperation negotiation list.

[0066] For influencers who have reached an agreement on cooperation willingness and price, the server begins to interact with the influencer to negotiate the release time of the influencer's works. The big model will make preliminary recommendations on the release time of the works based on the influencer marketing time period in the influencer marketing plan (for example, within the next three months), the product promotion rhythm (such as new product launches, promotional activities, etc.), and the personal characteristics of the influencer (such as the peak time for the influencer to release content). For example, the big model sends an email to the influencer: "Dear [influencer name], we have reached an agreement on cooperation willingness and price. According to our marketing plan, we hope that you can release promotional content about "energizing running shoes" on [specific date 1]. This date takes into account that we will be launching a large-scale promotional event a week later, and according to your fan interaction data, your fans are more active during this period. What do you think of this release time?" If the influencer replies that he agrees with the release time, for example: "Hi, [specific date 1] is fine for me, I can release the content at this time." The server will record this information and mark that the influencer has reached an agreement on the release time.

[0067] If the influencer replies that he / she disagrees with the release time and proposes other time requirements, such as: "Hello, I already have other arrangements for [specific date 1]. I think [specific date 2] will be more suitable." The server will process according to the influencer's reply. The big model will evaluate whether the [specific date 2] proposed by the influencer meets the requirements of the marketing plan, such as whether it is within the marketing time period, whether it will affect the overall rhythm of product promotion, etc. If [specific date 2] meets the requirements, the big model will reply: "Dear [name of influencer], [specific date 2] is OK, then we will confirm that you will release the promotional content about 'energetic running shoes' on [specific date 2]." If [specific date 2] does not meet the requirements, the big model will further negotiate with the influencer and propose other feasible time options, such as: "Dear [name of influencer], [specific date 2] does not quite fit our marketing plan. We can consider [specific date 3]. This time is also within our marketing time period and is close to the time of our promotional activities. Do you think it is okay?" In this process, in order to ensure the efficiency of the entire negotiation interaction, the total number of communications (including the total number of communications to confirm the willingness to cooperate, discuss the cooperation price, and agree on the release time) shall not exceed 10. If no agreement is reached after more than 10 times, the server will decide whether to continue negotiations with the influencer or terminate the cooperation negotiations.

[0068] The server continues to conduct at least one round of the above interactions for each target influencer account, including confirming the influencer's willingness to cooperate, negotiating the cooperation price, and agreeing on the release time of the influencer's work, until all interactions are completed. For example, after a period of negotiation, 30 influencers may have reached an agreement on the willingness to cooperate, price, and release time, and these 30 influencers will become the finalized cooperative influencers. The server will organize and record all email exchanges, finalized cooperation conditions, and other information during the entire negotiation interaction process as the result of completing the negotiation interaction, in preparation for the subsequent generation of the influencer cooperation plan.

[0069] In the embodiments of the present invention, the following implementation modes are also provided.

[0070] Obtaining the negotiation interaction result, wherein the negotiation interaction result includes total marketing data and total amount; Determining whether the total marketing data and / or the total amount meets the internet celebrity marketing plan; If yes, then the negotiation interaction result is determined to be successful; If not, the negotiation interaction result is determined to be a failure.

[0071] In an embodiment of the present invention, illustratively, after completing the negotiation interaction with each Internet celebrity, the server begins to obtain the total marketing data in the negotiation interaction results. The total marketing data includes multiple aspects. The first is to determine the number of Internet celebrities to cooperate with. Assume that after a series of previous negotiation interactions, it is finally determined to cooperate with 30 Internet celebrities. For each Internet celebrity, the server will record the number of fans, for example, the number of fans of Internet celebrity 1 is 500,000, the number of fans of Internet celebrity 2 is 300,000, and so on. The number of fans of all cooperating Internet celebrities is summarized, which is part of the total marketing data, which reflects the potential audience size that can be reached through these Internet celebrities.

[0072] In addition to the number of fans, the regional distribution data of the influencers’ fans is also included. For example, 10 of the influencers’ fans are mainly concentrated in the eastern United States, 15 of the influencers’ fans are in major cities in Canada, and the other 5 influencers’ fans are widely distributed, covering multiple regions in North America. This regional distribution data helps to evaluate the potential exposure of products in different regions.

[0073] At the same time, the fan interaction rate of influencers is also an important part of the total marketing data. The server will count the fan interaction rate data of each cooperating influencer, such as the interaction rate of influencer 1's likes, comments, and shares is 10%, and the interaction rate of influencer 2 is 8%, etc. By combining these interaction rate data, it is possible to estimate the possible participation of fans when the influencer promotes the product.

[0074] The server also obtains the total amount in the negotiation interaction results. During the negotiation and interaction process with influencers, the cooperation quotation of each influencer has been determined. For example, the cooperation quotation of influencer 1 is $1,000, the quotation of influencer 2 is $800, and so on. The server adds up the quotations of all cooperating influencers to get the total amount. Assume that the total quotation of 30 influencers is $25,000. This total amount is the cost expenditure part of this influencer marketing activity, which needs to be compared and evaluated with the budget in the influencer marketing plan.

[0075] First, let’s look at the number of influencers required in the influencer marketing plan. Assuming that the plan originally set a target of 20-30 influencers to work with, and the 30 influencers actually confirmed to work with are just within this range, from the perspective of the number of influencers, the total marketing data meets the plan requirements in this regard.

[0076] Regarding the number of fans, if the marketing plan expects to reach at least 8 million fans through cooperation with influencers, the server will calculate the total number of fans of all cooperating influencers. For example, for the 30 influencers mentioned above, add up their number of fans (500,000 + 300,000 + ...), assuming the total is 9 million fans, then the number of fans also meets the requirements of the influencer marketing plan.

[0077] Regarding the geographical distribution of fans, if the goal of the marketing plan is to focus on covering major urban areas in the United States and Canada, based on the previously obtained data on the geographical distribution of fans, there are 25 Internet celebrities whose fans are concentrated in these key areas, which is basically in line with the marketing plan's expectations for geographical coverage.

[0078] In terms of fan interaction rate, if the marketing plan wants the average fan interaction rate of the cooperating influencers to reach more than 8%, the server will calculate the average interaction rate of all cooperating influencers. Assuming the calculated average interaction rate is 9%, this means that the fan interaction rate also meets the requirements of the influencer marketing plan.

[0079] The total budget set in the marketing plan is $50,000. The actual total amount calculated is $25,000, which is less than $50,000 and meets the requirements of the influencer marketing plan.

[0080] Since the total marketing data (number of influencers, fans, geographical distribution, interaction rate) and total amount meet the requirements of the influencer marketing plan, the server determines that the negotiation and interaction result is successful. This means that the preliminary negotiation stage of the entire influencer marketing campaign has been successfully completed and can enter the next stage, such as further determining the cooperation details with the influencer and preparing promotional content.

[0081] The server will record the successful judgment result and may trigger some subsequent operations, such as sending a notification to the marketing team to inform them that the Internet celebrity negotiation was successful and the cooperation can proceed as planned. At the same time, the server may start to organize the relevant information of the cooperation with the Internet celebrity, such as integrating the cooperation terms, release time and other information of each Internet celebrity, so as to facilitate subsequent execution and monitoring.

[0082] Suppose a situation where the total marketing data or total amount fails to meet the influencer marketing plan. For example, although the number of fans and geographical distribution meet the requirements, the average fan interaction rate is only 6%, which is lower than the requirement of more than 8% in the marketing plan, or the total amount is calculated to be US$60,000, which exceeds the budget of US$50,000. In this case, the server determines that the negotiation interaction result is a failure.

[0083] When it is judged as a failure, the server will take corresponding measures. It may generate a detailed report indicating in which aspects the marketing plan is not met, such as "the fan interaction rate is not up to standard, the average interaction rate is 6%, which is lower than the planned requirement of 8%" or "the total amount exceeds the budget, the actual total amount is US$60,000, and the budget is US$50,000". This report can be provided to the marketing team so that they can analyze the problem and decide on the subsequent action plan. At the same time, the server may perform some remedial operations on the failure according to the pre-set strategy, such as re-evaluating the cost-effectiveness of the influencer in the case of total amount overspending, considering whether some influencers with higher quotations can be replaced, or renegotiating the price with the determined influencers; when the fan interaction rate is not up to standard, re-screening some influencers with higher fan interaction rates for negotiation, etc.

[0084] In the embodiments of the present invention, the following implementation modes are also provided.

[0085] On the basis that the negotiation interaction result is a failure, the Internet celebrity keywords are expanded, and the step of inputting the Internet celebrity marketing plan into a preset Internet celebrity model to obtain Internet celebrity screening conditions for the Internet celebrity marketing plan is returned to be executed.

[0086] In the embodiment of the present invention, for example, when the server determines that the negotiation interaction result is a failure, it will first conduct a comprehensive and in-depth analysis of the reasons for the failure. It is assumed that the reason for the failure is that the fan interaction rate in the total marketing data does not meet the requirements of the Internet celebrity marketing plan, and the total amount exceeds the budget.

[0087] For fans interaction rate not meeting the standard, the server will check the fan interaction data of each influencer in detail. For example, it is found that although some influencers have a large number of fans, their interaction rate is low. This may be because these influencers have a relatively simple content style or their interaction with fans is not active enough. In terms of total amount overspending, the server checks the cooperation quotation of each influencer and finds that some influencers' quotations are too high, exceeding the reasonable range based on their fan influence and marketing effect estimates.

[0088] Based on the above analysis, the server determines that it needs to make adjustments in finding more cost-effective influencers with higher fan interaction rates. This leads to the necessity of expanding the influencer keywords, by expanding the search scope and finding more potential influencers who meet the requirements.

[0089] Since the fan interaction rate did not meet the standard, the server began to consider keywords related to high fan interaction rates. For example, the keywords in the previous influencer marketing plan may be "sports running shoes" and "fitness" and other relatively broad words related to products and fields. Now, in order to increase the fan interaction rate, the server will add some keywords related to the interaction method, such as "question-and-answer sports sharing" and "fan interaction challenge". These keywords can help screen out influencers who are good at promoting products by interacting with fans.

[0090] At the same time, the server will also consider keywords related to the content style of the influencer to increase the fan interaction rate. For example, keywords such as "humorous sports commentary" and "real experience sharing" are added. Because influencers with a humorous style or focusing on sharing real experiences are often more likely to attract fans' attention and interaction.

[0091] In the case of total amount overspending, in addition to paying attention to the cost-effectiveness of the influencer, the server will also expand some keywords related to the influencer cooperation model. For example, keywords such as "low-cost cooperation method" and "product replacement cooperation" are added. These keywords help find influencers who are willing to accept other cooperation models besides cash payment, thereby reducing costs without reducing marketing effectiveness.

[0092] In addition, the server may add some keywords related to emerging influencers, such as "potential influencers", "newly emerging sports influencers", etc. In order to increase their popularity, emerging influencers may accept relatively low bids. At the same time, they may have a unique fan base, which may have a higher interaction rate because they have a closer relationship with the influencer and are in the early stages of the influencer's growth.

[0093] The server re-enters the influencer marketing plan containing the expanded influencer keywords into the preset influencer model. For example, the influencer keywords in the original influencer marketing plan are "sports running shoes" and "fitness", which are now changed to "sports running shoes", "fitness", "question-and-answer sports sharing", "fans interactive challenge", "humorous sports commentary", "real experience sharing", "low-cost cooperation method", "product replacement cooperation", "potential influencer", "newly emerging sports influencer", etc.

[0094] After receiving the updated marketing plan, the big model starts to regenerate the influencer screening criteria. The big model will adjust the search strategy on different social media platforms based on the new keywords. For example, for the keyword "question-and-answer style sports sharing", the big model will search for influencers with similar question-and-answer content in their historical posts; for the keyword "product replacement cooperation", the big model will search for influencers who have had product replacement cooperation models in their profiles or previous cooperation experiences.

[0095] When regenerating influencer screening conditions, the big model will comprehensively consider all marketing plan information, including influencer marketing area information, influencer language, influencer type, total number of influencers, total budget, influencer marketing time period, promoted product name and introduction, influencer tags, newly expanded influencer keywords, and influencer cooperation negotiation deadline, etc.

[0096] Taking the regional information as an example, the model will still search for influencers in North America (the United States and Canada) based on the new keywords in this region. If the influencer's language is English, the model will filter influencers with English content. If the influencer type is a sports influencer, the model will pay more attention to those sports influencers that meet the new keywords, such as sports influencers who are good at "humorous sports commentary".

[0097] In terms of the total number of influencers, due to the previous failed negotiations, this number may be adjusted according to the specific circumstances of the failure. If the previous failure was due to too many influencers being found but of varying quality, the total number requirement may be appropriately reduced; if it is due to insufficient numbers to meet marketing needs, this number may be maintained or slightly increased.

[0098] The total budget limitation still exists, and the big model will adjust the screening conditions based on the total budget and newly added keywords related to cost control (such as "low-cost cooperation mode" and "product replacement cooperation"). For example, those influencers who offer high prices and do not meet the new cooperation model will be excluded from the initial screening.

[0099] The influencer marketing time period (e.g., within the next three months), the name of the promoted product ("energizing running shoes"), and the introduction ("adopting advanced shock absorption technology, light and breathable, suitable for running on all kinds of surfaces") remain unchanged. The big model will continue to ensure that the selected influencers can effectively promote the product within this time period. Influencer tags may be expanded or adjusted based on new keywords. For example, the original "sports equipment" tag may be refined into "cost-effective sports equipment" and other tags related to new keywords.

[0100] Finally, the deadline for negotiation with influencers (e.g. two weeks later) will also be taken into account by the big model. When screening influencers, priority will be given to those who can complete the negotiation within this time. Just like the previous screening process, the filtering condition will also be given the condition that the email address is not empty to ensure that subsequent negotiations can proceed smoothly.

[0101] Once the big model regenerates the influencer screening conditions, the server will start a new round of influencer marketing process according to the new screening conditions. This includes calling the influencer search tool, screening in the influencer marketing knowledge base based on the new screening conditions, and obtaining new preliminary screening results for influencers; then calling the influencer data analysis tool through the preset influencer big model, analyzing the new preliminary screening results for influencers, and obtaining new target influencer accounts; then calling the influencer email query tool through the preset influencer big model to obtain the email addresses of new target influencer accounts, and conducting new negotiation interactions with these email addresses through the email sending and receiving tool, repeating the previous steps of confirming the influencer's willingness to cooperate, negotiating the cooperation price, and agreeing on the release time of the influencer's works, and finally obtaining new negotiation interaction results again to determine whether they meet the influencer marketing plan, until a successful negotiation interaction result is obtained or a failure occurs again and further adjustments are required.

[0102] In an embodiment of the present invention, the process file of the negotiation interaction is input into the preset Internet celebrity model to obtain the Internet celebrity cooperation plan of the Internet celebrity marketing plan, which can be implemented through the following example.

[0103] The negotiation interaction process file is input into the preset internet celebrity model to obtain the internet celebrity cooperation plan of the internet celebrity marketing plan consisting of cooperation internet celebrity information, cooperation quotation, work release time and content description.

[0104] In the embodiment of the present invention, the following is an example of a detailed scenario of each step of this part of the technical content, taking the server as the execution subject: After the server completes the negotiation interaction with the influencer, the first thing it needs to do is to collect all the email exchanges related to the negotiation interaction. For example, for each cooperating influencer, the server will extract the initial cooperation invitation email, the influencer's reply expressing the willingness to cooperate email, multiple rounds of emails discussing the cooperation price, and the emails agreeing on the release time of the work from the storage of the email sending and receiving tool. These emails contain rich information. For example, the cooperation invitation email mentions the detailed information of the promotion product, the advantages of the influencer, and the general intention of the cooperation; the price negotiation email records the quotation, price increase, and bargaining process of both parties; the email agreeing on the release time clarifies the final release date of the work.

[0105] In addition to email information, the server will also organize other data related to negotiation interactions. This includes data such as the number of fans, fan interaction rate, and activity of the influencer obtained from the influencer data analysis tool. These data are used to evaluate the value of the influencer and the match with the product during the negotiation process. For example, an influencer has 500,000 fans, a fan interaction rate of 10%, a high level of activity, and publishes 3-4 sports-related content per week on average. The server will integrate this data with the email information to form a complete negotiation interaction process file.

[0106] The server inputs the organized negotiation interaction process file into the preset influencer big model. After receiving the file, the big model begins to analyze and process the information in it. The big model has predefined algorithms and logic, which can identify various information elements in the file and build influencer cooperation plans based on these elements. For example, the big model will first identify the identity information of each influencer, which is obtained from the recipient address of the email, the signature in the influencer's reply, etc. Then, the big model will parse the email content and extract key information such as the cooperation quotation and the release time of the work.

[0107] For information about cooperating influencers, the big model will combine the email content with other data. It will obtain basic information about the influencer, such as the name and contact information, from the email, and obtain information about the influencer's influence from data analysis. For example, if an influencer states in an email that he has many years of experience in the sports field and has professional certification, and his number of fans and interaction rate data also show that he has a high influence, the big model will integrate this information to form a detailed section about the cooperating influencer information about this influencer.

[0108] In terms of cooperation quotations, the big model will carefully sort out the emails discussing the cooperation price. From the initial quotation, the influencer's feedback, each price increase or price reduction, to the final quotation. For example, for a certain influencer, the initial quotation was $800. After the influencer raised objections and two rounds of price increases, the final cooperation quotation was $1,000. The big model will accurately record this final quotation and include it in the cooperation quotation section of the influencer cooperation plan.

[0109] Regarding the release time of the work, the big model will extract the accurate date and time information from the email that agreed on the release time. For example, if the email clearly states that the influencer will release promotional content about "energizing running shoes" on [specific date], such as October 15, 2023, the big model will record this date as the release time part of the influencer cooperation plan.

[0110] In the influencer cooperation plan, the information section of the cooperating influencer will be very detailed. For each cooperating influencer, in addition to the basic information and influence-related information mentioned above, it may also include the geographical distribution of the influencer's fans. For example, the fans of a certain influencer are mainly concentrated in California, New York, and Ontario, Canada, which is of great significance for the promotion of products in specific regions. The big model will also summarize the content style characteristics of the influencer based on the content of the email and data analysis, such as this influencer is good at introducing sports equipment in a humorous way, and often combines his own sports experience to share. This information will be recorded in detail in the cooperating influencer information section.

[0111] The final quote for each influencer will be clearly listed in the quote section. For example, influencer 1's quote is $1,000, influencer 2's quote is $900, etc. This helps the marketing team with cost accounting and budget control. At the same time, if there are any special quote terms during the negotiation process, such as additional bonuses or partial compensation based on the promotion effect, they will also be explained in this section. For example, influencer 3's quote is $800, and if the number of likes for the promotion content exceeds 10,000, he will receive an additional sports product worth $200.

[0112] The release time section will be recorded according to the exact date and time extracted from the email. In addition to the date, it may also include the specific time period of the release, such as morning, afternoon or evening. Because different time periods may affect the exposure of the work and the interaction of the audience. For example, influencer 1 will release promotional content at 10 am on October 15, 2023, and influencer 2 will release it at 8 pm on October 20, 2023. These precise release schedules help the marketing team to plan and coordinate the overall promotion activities.

[0113] The big model will generate the content description part according to the content of the email, the characteristics of the influencer, and the promotion needs of the product. For each influencer, the content description will guide the influencer to create promotional content. For example, for influencer 1, the content description may be: "Please focus on the shock absorption technology and lightweight and breathable characteristics of the 'Vibrant Running Shoes' in the promotional content, and share it in combination with your real experience during long-distance running. At the same time, you can show the appearance design of the running shoes and mention our upcoming promotional activities." For influencer 2, the content description may be: "You can make a comparison video to compare the 'Vibrant Running Shoes' with other similar products, highlighting the advantages of our products in being suitable for running on various roads. At the end of the video, guide fans to pay attention to our official website for more product information." These content descriptions ensure that the influencer can produce promotional content according to the requirements of the marketing plan, thereby improving the effectiveness of product promotion.

[0114] Finally, the big model presents the constructed Internet celebrity cooperation plan as a whole. This plan will be stored on the server in a format that is easy to understand and operate, such as in the form of a table or document. The marketing team can easily view this plan to understand the detailed information, cooperation quotation, work release time and content requirements of each cooperative Internet celebrity. According to this Internet celebrity cooperation plan, the marketing team can further arrange related work for product promotion, such as preparing product samples to send to Internet celebrities, coordinating internal marketing resources with Internet celebrities' promotion activities, etc. At the same time, the server can also monitor and evaluate the promotion activities of Internet celebrities based on this plan, such as reminding Internet celebrities when the work release time is approaching, and analyzing the content description and actual promotion effect after the promotion activity ends.

[0115] In order to more clearly describe the solution provided by the embodiment of the present invention, a relatively complete implementation method is provided below. Figure 2 .

[0116] The present invention consists of an internet celebrity marketing knowledge base, a large model, and an internet celebrity marketing tool library.

[0117] 1) Influencer Marketing Knowledge Base: includes basic knowledge such as terms and concepts related to influencer marketing, as well as advanced knowledge such as influencer screening, influencer portraits, influencer communication skills, influencer invitation email skills, influencer quotation estimation methods, influencer bargaining skills, and influencer marketing tool instructions. Using RAG technology and mounting the influencer marketing knowledge base can enhance the professional capabilities of the big model in the field of influencer marketing.

[0118] 2) Big model: It is the brain of the present invention. It can call relevant influencer marketing tools, use influencer marketing knowledge base and its own wisdom according to the influencer marketing plan input by the user, and independently complete the cooperation negotiation such as docking and bargaining with influencers. Finally, it outputs an influencer cooperation plan that is agreed upon with influencers, including a list of influencers to be cooperated with after communication, the influencer's quotation, the content of the influencer's work and the release time, etc.

[0119] 3) Influencer Marketing Tool Library: A set of tools specifically used in the field of influencer marketing that can improve the efficiency and quality of influencer marketing, including influencer search tools, influencer data analysis tools, influencer email query tools, influencer cooperation quotation query tools, and email sending and receiving tools. The functions of each tool are introduced as follows: Internet celebrity search tool: You can search for Internet celebrities on various platforms based on keywords, country / region, language, number of fans, cooperation quotes and other screening conditions; Internet celebrity data analysis tool: used to analyze Internet celebrity fan portraits, Internet celebrity fan interaction rate, Internet celebrity value, Internet celebrity activity and other detailed Internet celebrity data; Internet celebrity email query tool: used to query the contact email of Internet celebrities; Internet celebrity cooperation quotation query tool: used to query the estimated lower and upper limits of Internet celebrity cooperation quotation; Email sending and receiving tool: used to negotiate cooperation with influencers. Bind an email account to the tool, and the big model can negotiate cooperation with influencers through email.

[0120] According to the functional modules, the present invention is composed of an Internet celebrity search agent and an Internet celebrity communication agent. The former is responsible for the preliminary screening and secondary screening of Internet celebrities, while the latter is responsible for the cooperation negotiation of Internet celebrities.

[0121] The processing flow is briefly described as follows: 1) Input influencer marketing plan: The user first inputs the influencer marketing plan, including the target country / region for influencer marketing, influencer language, influencer type, total number of influencers, total budget, influencer marketing time period, promotional product name and introduction, influencer label, influencer keyword, influencer cooperation negotiation deadline, etc. The influencer marketing time period refers to the time period for the release of influencer works. Here, a time period is provided to facilitate negotiation with the influencer on the release time point suitable for the influencer.

[0122] 2) Preliminary screening of influencers: The big model automatically generates influencer screening conditions based on the influencer marketing plan input by the user, and adds a condition that the influencer’s contact email address is not empty. It then calls the influencer search tool, passes in the influencer screening conditions, and obtains preliminary screening results for influencers.

[0123] 3) Secondary screening of influencers: The big model uses the influencer data analysis tool for the influencers initially screened above to further analyze the matching degree between influencers and products, and the quality of influencers, and select high-quality influencers.

[0124] 4) Internet celebrity cooperation negotiation: The big model uses the Internet celebrity email query tool to obtain the contact email of the Internet celebrity from the secondary screening. The email sending and receiving tools are used to complete multiple rounds of Internet celebrity cooperation negotiation, which generally includes the following contents: a) Confirm the influencer’s willingness to cooperate: Generate a personalized cooperation invitation email based on the influencer’s information. The email content includes an introduction to the marketing activities, the promotional products, the influencer’s advantages, the matching degree between the product and the influencer, and the willingness to cooperate for a fee. Finally, if the influencer is interested, he or she can contact us by email to discuss the specific details of the cooperation.

[0125] b) Discuss the cooperation price: The big model calls the Internet celebrity quotation query tool to obtain the lower and upper limits of each Internet celebrity's cooperation quotation. Then contact the Internet celebrity by email to inform us of our cooperation quotation (the first communication is the lower limit price of the Internet celebrity's cooperation quotation), and ask the Internet celebrity to confirm. If the Internet celebrity has any objections, the big model can appropriately increase the price based on the first quotation under the conditions of the Internet celebrity's own quotation, within the Internet celebrity's quotation upper limit, and under the total budget constraint, automatically make a second quotation, and then contact the Internet celebrity by email to confirm. Until the price is negotiated with the Internet celebrity. In order to ensure the Internet celebrity experience, the number of bargaining is generally 2 to 5 times. If it exceeds 5 times or the Internet celebrity's quotation is too high, the bargaining will be euphemistically ended, and the current cooperation negotiation will be terminated. There will be opportunities for cooperation later.

[0126] c) Agree on the release time of the influencer’s work: Based on the marketing activity plan, the influencer’s personal characteristics and needs, agree on a release time that is acceptable to both parties.

[0127] In order to control the budget and avoid contacting too many influencers, it is necessary to control the total number of influencers contacted in the first batch. For example, the number of top, middle and tail influencers should be 10 times the number of target cooperation influencers respectively. Subsequently, the influencers will be contacted in batches according to their response rate and intention.

[0128] 5) Evaluation and adjustment: After the above steps are basically completed, an overall evaluation should be conducted to determine whether the total data and total amount of the cooperating influencers meet the expectations of the marketing plan. If they exceed, they can be deleted; if they are insufficient, you can start again from the second step, expand keywords, screen more influencers, and complete cooperation negotiations.

[0129] 6) Output the Internet celebrity cooperation plan: After the above steps are completed, or after the timeout, the big model enters the result output stage, sorts out and summarizes the Internet celebrity negotiation situation, and outputs the confirmed Internet celebrity cooperation plan, including the cooperating Internet celebrity, cooperation quotation, work release time, and content description.

[0130] Here is a design example of a large model prompt for an influencer marketing assistant / agent: 1) Role definition: Role: You are an influencer marketing assistant who is good at influencer marketing and is committed to helping clients increase brand awareness and sales.

[0131] 2) Skill definition: Skill: Skill 1: Get a customer marketing plan 1. Ask the client for key information about the influencer marketing plan, including target countries / regions, influencer languages, influencer types, total number of influencers, total budget, marketing time period, promotional product name and description, and influencer cooperation negotiation deadline.

[0132] 2. If the customer fails to provide complete information, promptly remind them to supplement it.

[0133] Skill 2: Initial screening of influencers 1. Develop internet celebrity screening criteria based on the marketing plan provided by the client.

[0134] 2. Call the Internet celebrity search interface, pass in the filtering conditions, and obtain the preliminary filtered Internet celebrity results. If paginated returns are returned, obtain all paginated data.

[0135] Skill 3: Secondary screening of influencers 1. For the initially screened influencers, call the influencer data analysis interface to query the influencer analysis data.

[0136] 2. Conduct secondary screening based on conditions such as active influencers, real and valid fans, and more than 80% match between fan portraits and products, and eliminate influencers who do not meet the requirements.

[0137] Skill 4: Negotiating with influencers 1. Use the influencer email query tool to obtain the influencer’s contact email.

[0138] 2. Call the Internet celebrity cooperation quotation query tool to determine the lower and upper limits of the Internet celebrity cooperation quotation.

[0139] 3. Communicate through email sending and receiving tools: Confirm cooperation intention: Generate a personalized invitation email, which includes an introduction to the marketing campaign, product introduction, description of the influencer's advantages, description of the matching degree, and willingness for paid cooperation. If the influencer is interested, invite them to discuss the details of the cooperation.

[0140] Bargaining strategy: Offer a lower limit to the influencer's price. If the influencer has any objections, increase the price appropriately based on the influencer's reply, upper limit and total budget constraints. Generally, bargain 2-5 times. If it exceeds 5 times or the price is too high, end the negotiation in a tactful manner.

[0141] Agree on the release time: Determine a release time acceptable to both parties based on the marketing activity plan, the characteristics and needs of the influencer. The total number of communications shall not exceed 10 times.

[0142] Skill 5: Evaluate and adjust 1. After completing the above steps, conduct an overall assessment to determine whether the total data and total amount of the cooperating influencer meet the expectations of the marketing plan.

[0143] 2. If the number exceeds the limit, it can be deleted. If it is less than the limit, start from the second step to screen more influencers and negotiate cooperation.

[0144] Skill 6: Output cooperation plan 1. Organize and summarize the negotiation status with internet celebrities, and output the confirmed internet celebrity cooperation plan, including the cooperating internet celebrities, cooperation quotation, work release time and content description.

[0145] 3) Restriction requirements: limit: Only focus on matters related to influencer marketing and refuse to handle tasks that are not related to influencer marketing.

[0146] Follow the given steps and format strictly and do not deviate from it.

[0147] Ensure output is accurate, clear and compliant.

[0148] Please refer to Figure 3 , Figure 3 An embodiment of the present invention provides a large model-based Internet celebrity marketing device 110, including: The acquisition module 1101 is used to acquire an influencer marketing plan, an influencer marketing tool library and an influencer marketing knowledge base; the influencer marketing tool library is configured with an influencer search tool, an influencer data analysis tool, an influencer email query tool and an email sending and receiving tool; the influencer marketing plan is input into a preset influencer big model to obtain influencer screening conditions for the influencer marketing plan; the influencer search tool is called to perform screening in the influencer marketing knowledge base based on the influencer screening conditions to obtain an influencer preliminary screening result; the influencer data analysis tool is called through the preset influencer big model to analyze the influencer preliminary screening result to obtain a target influencer account; the influencer email query tool is called through the preset influencer big model to obtain an email address of the target influencer account, and the email sending and receiving tool is used to negotiate and interact with the email address; The marketing module 1102 is used to input the process file of the negotiation interaction into the preset Internet celebrity model based on the successful negotiation interaction result, so as to obtain the Internet celebrity cooperation plan of the Internet celebrity marketing plan.

[0149] It should be noted that the implementation principle of the aforementioned Internet celebrity marketing device 110 based on a large model can refer to the implementation principle of the aforementioned Internet celebrity marketing method based on a large model, and will not be repeated here. It should be understood that the division of the various modules of the above device is only a division of logical functions. In actual implementation, they can be fully or partially integrated into a physical entity, or they can be physically separated. And these modules can all be implemented in the form of software called by processing elements; they can also be all implemented in the form of hardware; some modules can also be implemented in the form of software called by processing elements, and some modules can be implemented in the form of hardware. For example, the Internet celebrity marketing device 110 based on a large model can be a separately established processing element, or it can be integrated in a chip of the above device. In addition, it can also be stored in the memory of the above device in the form of program code, and called and executed by a processing element of the above device. The function of the Internet celebrity marketing device 110 based on a large model. The implementation of other modules is similar. In addition, all or part of these modules can be integrated together, or they can be implemented independently. The processing element described here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each module above can be completed by an integrated logic circuit of hardware in a processor element or an instruction in the form of software.

[0150] For example, the above modules may be one or more integrated circuits configured to implement the above methods, such as one or more application specific integrated circuits (ASIC), or one or more microprocessors (digital signal processors, DSP), or one or more field programmable gate arrays (FPGA), etc. For another example, when a module above is implemented in the form of a processing element scheduling program code, the processing element may be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0151] The embodiment of the present invention provides a computer device 100, which includes a processor and a non-volatile memory storing computer instructions. When the computer instructions are executed by the processor, the computer device 100 executes the aforementioned Internet celebrity marketing device 110 based on a large model. Figure 4 As shown, Figure 4The computer device 100 is a block diagram of a structure of a computer device 100 provided in an embodiment of the present invention. The computer device 100 includes a big model-based online celebrity marketing device 110, a memory 111, a processor 112 and a communication unit 113.

[0152] To achieve data transmission or interaction, the memory 111, the processor 112 and the communication unit 113 are electrically connected to each other directly or indirectly. For example, the electrical connection between these elements can be achieved through one or more communication buses or signal lines. The big model-based Internet celebrity marketing device 110 includes at least one software function module that can be stored in the memory 111 in the form of software or firmware or solidified in the operating system (OS) of the computer device 100. The processor 112 is used to execute the big model-based Internet celebrity marketing device 110 stored in the memory 111, such as the software function modules and computer programs included in the big model-based Internet celebrity marketing device 110.

[0153] An embodiment of the present invention provides a readable storage medium, which includes a computer program. When the computer program is running, it controls the computer device where the readable storage medium is located to execute the aforementioned big model-based Internet celebrity marketing device 110.

[0154] For illustrative purposes, the foregoing description is made with reference to specific embodiments. However, the above illustrative discussion is not intended to be exhaustive or to limit the present disclosure to the precise form disclosed. Numerous modifications and variations are possible in accordance with the above teachings. These embodiments are selected and described in order to best illustrate the principles of the present disclosure and its practical application, so that those skilled in the art can best utilize the present disclosure and utilize various embodiments with different modifications to suit the intended specific application.

Claims

1. A big model-based online celebrity marketing method, characterized in that: include: Get the influencer marketing plan, influencer marketing tool library, and influencer marketing knowledge base; The influencer marketing tool library is equipped with influencer search tools, influencer data analysis tools, influencer email query tools, and email sending and receiving tools; Input the influencer marketing plan into a preset influencer macro model to obtain influencer screening conditions for the influencer marketing plan; Calling the influencer search tool, screening in the influencer marketing knowledge base based on the influencer screening conditions, and obtaining preliminary influencer screening results; Calling an internet celebrity data analysis tool through the preset internet celebrity model to analyze the preliminary screening results of the internet celebrities and obtain target internet celebrity accounts; The preset internet celebrity model is used to call the internet celebrity email query tool to obtain the email address of the target internet celebrity account, and the email sending and receiving tool is used to negotiate and interact with the email address; On the basis that the negotiation interaction result is successful, the process file of the negotiation interaction is input into the preset Internet celebrity model to obtain the Internet celebrity cooperation plan of the Internet celebrity marketing plan.

2. The method according to claim 1, characterized in that The influencer marketing plan includes information on influencer marketing regions, influencer language, influencer type, total number of influencers, total budget, influencer marketing time period, name and introduction of promoted products, influencer tags, influencer keywords, and deadline for influencer cooperation negotiations; The step of inputting the influencer marketing plan into a preset influencer macro model to obtain influencer screening conditions for the influencer marketing plan includes: Input the internet celebrity marketing area information, the internet celebrity language, the internet celebrity type, the total number of internet celebrities, the total budget, the internet celebrity marketing time period, the name and introduction of the promoted product, the internet celebrity label, the internet celebrity keyword, and the internet celebrity cooperation negotiation deadline into the preset internet celebrity macro model to obtain initial screening conditions; The initial screening condition is assigned a condition that the email address is not empty, so as to obtain the influencer screening condition for the influencer marketing plan.

3. The method according to claim 1, characterized in that The method of calling the internet celebrity data analysis tool through the preset internet celebrity model to analyze the preliminary screening results of the internet celebrities and obtain the target internet celebrity account includes: The internet celebrity data analysis tool is called through the preset internet celebrity big model, and the preliminary screening results of the internet celebrities are analyzed based on the product matching index and the internet celebrity instruction index to obtain the target internet celebrity account.

4. The method according to claim 1, characterized in that: The negotiation and interaction with the email address through the email sending and receiving tool includes: The email sending and receiving tool is used to conduct at least one round of interaction with the email address to confirm the Internet celebrity's willingness to cooperate, discuss the cooperation price, and agree on the release time of the Internet celebrity's work, so as to complete the negotiation interaction.

5. The method according to claim 1, characterized in that The method further comprises: Obtaining the negotiation interaction result, wherein the negotiation interaction result includes total marketing data and total amount; Determining whether the total marketing data and / or the total amount meets the internet celebrity marketing plan; If yes, then the negotiation interaction result is determined to be successful; If not, the negotiation interaction result is determined to be a failure.

6. The method according to claim 2, characterized in that The method further comprises: On the basis that the negotiation interaction result is a failure, the Internet celebrity keywords are expanded, and the step of inputting the Internet celebrity marketing plan into a preset Internet celebrity model to obtain Internet celebrity screening conditions for the Internet celebrity marketing plan is returned to be executed.

7. The method according to claim 1, characterized in that The step of inputting the negotiation interaction process file into the preset influencer model to obtain the influencer cooperation plan of the influencer marketing plan includes: The negotiation interaction process file is input into the preset internet celebrity model to obtain the internet celebrity cooperation plan of the internet celebrity marketing plan consisting of cooperation internet celebrity information, cooperation quotation, work release time and content description.

8. A network celebrity marketing device based on a large model, characterized in that: include: The acquisition module is used to obtain the influencer marketing plan, influencer marketing tool library and influencer marketing knowledge base; The influencer marketing tool library is configured with influencer search tools, influencer data analysis tools, influencer mailbox query tools, and mail sending and receiving tools; the influencer marketing plan is input into a preset influencer macro model to obtain influencer screening conditions for the influencer marketing plan; the influencer search tool is called to screen the influencer marketing knowledge base based on the influencer screening conditions to obtain preliminary influencer screening results; The internet celebrity data analysis tool is called through the preset internet celebrity model to analyze the preliminary screening results of the internet celebrities to obtain the target internet celebrity account; the internet celebrity email query tool is called through the preset internet celebrity model to obtain the email address of the target internet celebrity account, and the email sending and receiving tool is used to negotiate and interact with the email address; The marketing module is used to input the process file of the negotiation interaction into the preset Internet celebrity model on the basis that the negotiation interaction result is successful, so as to obtain the Internet celebrity cooperation plan of the Internet celebrity marketing plan.

9. A computer device, characterized in that: The computer device comprises a processor and a non-volatile memory storing computer instructions. When the computer instructions are executed by the processor, the computer device executes the method according to any one of claims 1 to 7.

10. A readable storage medium, characterized in that: The readable storage medium includes a computer program, and when the computer program is executed, the computer device where the readable storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

Images