Research on the enterprise online marketing mode based on data mining

Abstract

With the development and popularity of mobile networks, online shopping has gradually become a trend. For enterprises, the traditional marketing mode has been difficult to play an effective role when facing the emerging online shopping mode. This study aims to improve the revenue benefits of online shopping. This paper first introduces the traditional marketing mode and then selects the data mining model used for consumption preference segmentation to build an online marketing mode. An example analysis was conducted on a book sales company and a real estate company. The results showed that more users in this community preferred five types of books, and the percentages from high to low were teaching and learning materials, modern novels, popular science books, historical literature, and classical novels; more customers preferred online platforms among the channels for collecting information on home purchase. No matter it was the book sales company or the real estate company, compared with no fluctuation in the company’s turnover under the traditional marketing mode, the turnover of the company increased month by month after adopting the online marketing mode.

Keywords

Internet marketing community network data mining

1. Introduction

The emergence of the Internet and smart mobile terminals has led to the development of the online shopping industry. With the advantages of the Internet and smart mobile terminals, the online marketing mode is able to push marketing information to potential consumers more widely and precisely than the traditional marketing mode. The traditional marketing mode promotes the awareness of their products mainly by means of advertising. Despite the fact that the online marketing mode will also use advertising, the traditional marketing mode does not filter the type of customer but uses a large amount of advertisements to conduct a wide range of publicity. Although awareness can be improved with widespread publicity, the cost of publicity is high, and the number of customers that can be attracted is limited. The online marketing model uses the Internet to push ads and will target customers according to the type of online community the customer is in, thus improving push efficiency. With the rapid development of economy, people’s living standards have been gradually improved, and the most intuitive embodiment is that wages earned by legitimate labor have been generally improved. The general rise of the wage level makes people choose more diversified when shopping, which in turn stimulates enterprises to provide more commodity choices [1]. In order to attract more customers, in addition to improving their service quality (such as cheap and beautiful goods), enterprises will also expand their product popularity through marketing, so that customers are more inclined to products of enterprises [2]. However, with the popularity of the Internet and mobile terminals, e-commerce has gradually increased, and more and more people tend to more convenient online shopping [3]. On the Internet, as users can access to more information, traditional marketing methods have been difficult to meet the business needs of enterprises. In order to improve the profit of products, enterprises have gradually developed an online marketing mode suitable for the Internet environment. The online marketing mode is based on the Internet. Compared with the market demand orientation in the traditional marketing mode, the online marketing mode pays more attention to the customer demand orientation, and the product information in the marketing mode is not only released by the enterprise, and customers will also play a role in the publicity of the product information. Riasi and Pourmiri [4] evaluated the impact of online marketing strategy on the ecotourism industry of Iran. They found that the expansion of online marketing increased the number of foreign tourists visiting Iranian natural tourist attractions, but had no significant impact on the number of domestic tourists visiting these tourist attractions. Thaichon and Quach [5] studied the influence of online advertising on Australian children’ consumption intention of fast food and found that the fast food advertising on social networking sites could manipulate the purchase possibility, views on fast food and eating habits of young audience. Watfa et al. [6] implemented more effective target marketing activities using the group member information available on social networking sites and then carried out an experiment by sending a spam to more than 1000 Facebook users and associating it with some product pages. They found that the success rate of the method was 82%. Menelec and Jones [7] investigated the link between networking and marketing and discussed the extent to which small professional services firms use their networks for marketing activities and found that “core groups” was very important in their network structure. Qing et al. [8] used big data as a background to study the innovation strategy of online marketing. They found through investigation that big data could help marketers to analyze consumer behavior preferences and market trends. All of the above-mentioned studies evaluated Internet-based online marketing modes and all found that online marketing modes can enhance marketing outcomes. However, only the effectiveness of using the Internet for promotion was described in the above studies. In this study, in addition to using the Internet for wide-scale publicity, we also use data mining algorithms to mine the user profile on the Internet, so as to classify users and achieve targeted publicity, ensuring that the online marketing mode not only covers a wide range of target customers but also allows targeted publicity to potential customers. This paper briefly introduces the traditional marketing mode and the Internet-based online marketing mode. In the online marketing mode, data mining algorithms were used to classify Internet users so as to achieve precise pushing of promotional strategies. Then, a book sales company and a real estate company were taken for instance analysis. The final results showed that the data mining algorithm could effectively classify customer types and thus make online marketing more targeted, and the turnover of both companies has increased significantly after adopting the online marketing mode. The contribution of this paper is to provide an effective reference for the online marketing mode to improve business turnover.

2. Traditional marketing model

Marketing activities of an enterprise refer to the process that an enterprise obtains customers’ needs in the target market through market research and enables customers to understand and purchase the commodities provided by the enterprise. In the period of underdeveloped Internet communication, consumers’ consumption activities are often carried out in offline stores; the way that consumers get commodity information is mainly daily browsing and advertising. The traditional marketing model is shown in Fig. 1, which is generally composed of three steps. The first step is market research [9], including questionnaire survey, telephone consultation, email consultation, investigation of the sales situation of representative merchants in the local market, etc., and the consumption tendency of customers in the local market is analyzed by relevant data obtained by market research. The next step is strategy formulation. According to the market situation that is obtained through market research and analysis, the corresponding marketing strategy mode is formulated. The commonly used strategy modes are advertising marketing mode [10], text message marketing mode, telephone marketing mode, etc. The final step is strategy implementation. Product marketing activities are carried out in accordance with the marketing strategy formulated in the previous step. Market research is performed again after a period of time to analyze the effect of the marketing strategy and correct it. The above three basic steps are repeated. The traditional marketing mode is market-oriented and mainly focuses on consumers. When it is applied to online marketing, it is equivalent to giving up some consumers and paying too much attention to the whole, so that the publicity among most consumers is too average, which eventually leads to the reduction of attraction and the low conversion rate of customers, but the marketing cost will not reduce [11].

Figure 1.

The basic process of the traditional marketing mode.

3. Data mining-based marketing mode

E-commerce that is booming rapidly with the popularity of Internet and shopping app makes the traditional marketing mode gradually expose its defects, as mentioned above. In order to ensure the online sales performance of enterprises and broaden the sales channels of products, enterprises began to adjust the traditional marketing mode based on the Internet, so as to change it into online marketing mode. The online marketing mode is based on various communication platforms on the Internet. During the operation of these platforms, platform participants will gradually gather to form a community network [12]. Participants in the same community usually have a single common characteristic or multiple common characteristics. The community network based online marketing mode is to carry out targeted marketing activities through the Internet communication platforms such as e-mail, forum, and post bar by analyzing community network.

Figure 2.

The data mining-based online marketing mode.

Different from the traditional marketing mode in which media such as advertising, telephone, text message are used for marketing content publishing, the data mining-based online marketing mode uses the community network platform in Internet marketing content publishing. The marketing content will permeate with different degrees in the communication process of a community network platform along with the attributes of different communities. Users in the same community often know as long as one user knows. If the marketing content meets the wishes of users, users will spontaneously transfer the marketing content to another community in which they participate to form a chain communication [13]. In this process, enterprises hardly need to bear the marketing cost. The basic flow of the data mining-based online marketing mode is shown in Fig. 2. Firstly, enterprises will collect relevant data information from the community network platform and analyze the distribution and demand preferences of users. Community network platforms include social networking platforms (WeChat, QQ, Baidu Tieba, etc.), video platforms (Blibili, iQIYI, Tik Tok, etc.), portals (Jinri Toutiao, Sohu, etc.), etc. The above network platforms have a large number of users and clearly divided sectors, such as WeChat Moments, different theme of video websites and section forum of portals, various themes of video websites, and the forum of portal websites [14]. Users of these platforms are potential consumers for enterprises. These users get information from these platforms and spread it in their communities.

Enterprises not only collect data from the community-based network platform to analyze the overall preference of consumers, but also collect the feedback of consumers in the platform and the feedback of enterprise products affected by marketing information. Enterprises will mine feedback data [15] and analyze different needs of consumer groups, so as to implement more targeted marketing activities. There are many kinds of data mining technologies. This paper mainly introduces a data mining model for consumption preference division. The basic principle of this data mining technology is to solve the group preference of various product information under the premise of considering the cost of information acquisition, so as to minimize the difference of preference difference between group and individuals of any two kinds of information among current product information. The calculation model is as follows:

objective function:

$\displaystyle\left\{\begin{array}[]{l}\min J\\ J=\sum\limits_{k=1}^{M}{\sum\limits_{i=1}^{N}{\sum\limits_{i<j}^{N}{((X_{ik}-X% _{jk})-(X_{ig}-X_{jg}))^{2}\omega_{k}\cdot\textit{GP}_{i}}}}\\ \textit{GP}_{i}=\frac{\sum{{v_{j}}\mathord{\left/{\vphantom{{v_{j}}{v_{i}}}}% \right.\kern-1.2pt}{v_{i}}}}{N-1}\\ v_{i}=\frac{c_{1}o_{i}+c_{2}j_{i}}{\sum{c_{1}o_{j}+c_{2}j_{j}}}\\ o_{i}=\frac{P_{i}}{\sum{P_{j}}}\\ j_{i}=\frac{M_{i}}{\sum{M_{j}}}\\ \end{array}\right.,$ (1)

where $J$ is the objective function, which represents the consistency of group preference, $M$ is the number of consumers, $N$ is the quantity of product information, $X_{ik}$ and $X_{jk}$ represent the degree of preference of consumer group to product $i$ and $j$ respectively, $\omega_{k}$ represents the weight of consumer $k$ , $\textit{GP}_{i}$ represents the unit information acquisition cost of product $i$ , $P_{i}$ represents the cost of obtaining information of product $i$ , $M_{i}$ represents the storage capacity of production $i$ information, $o_{i}$ and $j_{i}$ are the normalization value of $P_{i}$ and $M_{i}$ respectively, $v_{i}$ is the value of common normalization of $o_{i}$ and $j_{i}$ , and $c_{1}$ and $c_{2}$ are the weight of $o_{i}$ and $j_{i}$ respectively.

In the data mining-based online marketing mode, after mining the data collected from the platform and consumers with the above data mining technology, the marketing department of the enterprise analyzes the mining results, so as to formulate or adjust the marketing content and form for different consumer groups in the analysis results. After that, the adjusted marketing content will be released to the community network platform, and the data feedback of the platform and consumers continues to be collected and mined. The above steps are repeated. This is the online marketing mode.

4. Case study

4.1 Case overview

Company X is a company that mainly sells books, and the books cover many types, including popular science books, historical documents, classical novels, modern novels, etc., and can meet various requirements of buyers. The company has more than one thousand employees and also has book storage points all over the country, so that the goods can be sent to the buyer quickly. In addition to the complete range of books and rapid book logistics, the company also provides perfect after-sales services. If the sold books are damaged due to non-human factors, customers can apply to the company for a replacement at any time. In addition, the company provides pre-purchase consulting services and supplements the supply at any time.

Company Y is a real estate company that develops and sells residential properties. The company has 200 employees and includes human resources, finance, sales, procurement, and planning departments. The company has established large commercial complexes such as home building materials in the Fujian to attract other brand merchants and drive employment by setting up a variety of construction projects.

4.2 Marketing means

4.2.1 Traditional marketing model

Before online shopping was popularized, the offline marketing mode was the main mode. The two companies analyzed in this paper sell different products, books and houses, but the two companies have something in common in their traditional offline marketing modes. The first form is sitting marketing [16], i.e., a kind of marketing mode waiting for customers to consult and buy. Under this marketing mode, the companies do not actively sell products, but the customer actively initiates the purchase intention, and the consumer is a group usually. Although the companies are in a passive position under this sales mode, the transactions are often completed quickly. The second form is channel sales. Under this marketing mode, the companies do not directly sell books, but entrust the sales task to other organizations. The third form is travel marketing, which is corresponding to sitting marketing. The companies actively market their products door-to-door to broaden the sales scope, and moreover they also carry out advertising promotion.

4.2.2 Online marketing mode

After the popularization of online shopping, with the entry of enterprises providing online shopping services, the share of company X in the book sales market has gradually decreased. In order to adapt to the online shopping mode and increase its share in the book sales market, the online marketing mode which was targeted on the community network was adopted.

Users in the community network platform were analyzed using the data mining technology described above to divide needs of users for different types of books, subdivide the levels of potential consumers, and adopt different marketing strategies for different levels of consumers.

In the community-based network platform, such as WeChat Moments [17] and Weibo, for the different levels of consumers analyzed before, attractive topics were developed, and diversion links were added. Users who were attracted by topics in the network platform were absorbed into the community network that the company could radiate.

In the community network that the company could radiate, the company released the concerned information to the attracted groups, forwarded the evaluation of other customers, and answered the questions of consumers in time.

The soft advertisement with attractive titles was released in the community network platform of the company, so that consumers who were attracted to the network platform spontaneously delivered the soft advertisement to their respective network communities to expand the attractiveness of the company.

The above is the main means for Company X to use the Internet for online marketing. Company Y mainly deals with residential sales, so the process of using the Internet for online marketing will be different. Its basic process after using the Internet can be divided into land auction stage, design and construction stage, marketing stage and property management.

In the land auction stage, for a real estate company, land is important. The Internet is used to achieve customer demand-oriented land selection. The means include cooperation with big data companies to obtain information on customer demand and use of C2B platforms to interact with customers to understand site selection intentions.

In the design and construction stage, the C2B platform of the Internet is used to fully interact with customers, collect their needs and conduct targeted online and offline promotion, while attracting customers to participate in the design of the building to enhance the attractiveness of the product; at the same time, according to the results of the survey of customer needs, a standard library of price and products is established to reduce the trial-and-error cost of repetitive design.

In the pre-marketing stage, the core users and opinion leaders in the online community enhance are taken advantage of to improve the speed of communication; in the customer introduction period, an O2O marketing platform is created on the Internet, so that the physical merchants and the online platform can be combined to enable customers to use both online payment and offline services.

In the property management stage, the Internet of Things is used for intelligent management of residential communities.

From the above residential sales process, we can see the role of the Internet. The application of the network data mining algorithm to the network marketing mode studied in this paper is mainly reflected in the land auction stage and the design and construction stage. In these two stages, data mining algorithms are used to mine and classify customer needs, so that targeted publicity and sales can be conducted.

4.3 Marketing results

In order to make the online marketing mode of the company more targeted, it mined the user data on different community network platform using data mining technology, and Fig. 3 shows the book type purchase tendency of users obtained after mining users of the network platform. It was seen from Fig. 3 that users in the network community mainly tended to purchase five types of books, and the proportions were 43.1% of popular science books, 20.3% of historical documents, 11.4% of classical novels, 59.6% of modern novels and 70.2% of teaching auxiliary materials.

Figure 3.

The book type tendency of users in the network platform obtained by data mining technology.

As shown in Fig. 4, the book sales turnover of the company in the year before applying the online marketing mode fluctuated from 250000 yuan to 400000 yuan/month, and the turnover of the company had obvious fluctuation when using the traditional marketing mode. It was seen from Fig. 4 that the turnover of March $\sim$ May and September $\sim$ October was high and the turnover in the rest months was low. After the online marketing mode was adopted, the turnover of the company showed a steady upward trend compared with the traditional marketing mode. In the early stage of implementing the online marketing mode, the turnover was lower than the traditional marketing mode. In the early stage of online marketing, the company needed to carry out data mining for users in the community-based network to design targeted strategies and then mine the feedback data to adjust the strategies, and the above process took some time and took up some resources of the company; therefore the turnover was lower compared to the traditional marketing mode in the early stage.

Figure 4.

Changes of book turnover of the company before and after the adoption of online marketing mode.

In order to be more targeted in the promotion of residential sales, the data mining algorithm was used to investigate the purchase intention channels of customers, and the results are shown in Fig. 5. As can be seen from Fig. 5, 7.7% preferred TV advertising channels, 4.1% preferred radio channels, 7.2% preferred outdoor advertising channels, 2.1% preferred newspaper channels, 44.8% preferred online platform channels, 6.5% preferred market auction channels, 5.7% preferred real estate trade fair channels, 16.6% preferred intermediary channels, and 5.3% preferred other channels. It was concluded that the buyers tended to obtain information from online platforms more often. Therefore, when placing advertisements, the company can focus on the community on the online platform, especially the core users and opinion leaders in the community.

Figure 5.

The purchase intention channels of buyers obtained by data minging technology.

As shown in Fig. 6, the company’s housing sales fluctuated between 500,000 RMB and 550,000 RMB in the year before adopting the online marketing mode and tended to be stable; after one year, the company’s housing sales showed a stable upward trend, but slightly lower than the sales without the online marketing mode at the beginning. The reason for the lower sales was that mining the data of the users in the community-based network and mining the feedback data to adjust the strategy took some time and occupied some company resources at the beginning of the online marketing.

Figure 6.

Changes in the housing turnover one year before and after adopting the online marketing mode.

5. Discussion

The emergence and popularization of the Internet, mobile terminals and online shopping apps have made e-commerce, a new business mode, develop rapidly. However, the big data and high degree of communization on the Internet make it difficult for traditional marketing methods to adapt to and have an effective effect [18]. In order to improve the performance of enterprises, it is necessary to adjust the traditional marketing mode for the emerging e-commerce. The new online marketing mode makes use of characteristics of community network platform to achieve the low-cost and high-income marketing effect. The online marketing mode points the marketing subject to users on community network platforms. Users on the community network platform will spontaneously form groups with different themes according to their interests and hobbies, and each user will often participate in more than one community. Therefore, enterprises can analyze the network platform through data mining technology to obtain the community division of users and use marketing strategies according to different themes of communities, and users in the community will spontaneously transfer marketing information to other communities in which they participate to realize the diffusion of information; in this diffusion process, enterprises almost do not need to invest in costs [19].

In order to verify the effectiveness of the data mining-based online marketing mode, this study analyzed changes of a company that mainly sells books and a real estate company before and after the adoption of online marketing. The book sales company analyzed the book tendency of users on various network platforms through data mining technology. The real estate company analyzed the housing purchase channels selected by clients with the data mining algorithm. The result have shown above. In the analysis of the book sales company, the number of customers who preferred teaching auxiliary materials accounted for the highest proportion, which was because most people attach importance to education and all kinds of teaching auxiliary materials are nearly rigid demands in the implementation of education; the number of customer who preferred classical novels accounted for the lowest proportion, which was because the interesting nature of classical novels is difficult to match with most modern young people. In the analysis of the real estate company, most clients preferred to use online platforms when collecting information on housing purchases because of the convenience of the Internet and mobile smart terminals.

Then it is the comparison of turnover. The turnover of the two companies on the year before the application of the online marketing mode fluctuated obviously with month, while the turnover on the year after the application grew steadily with month. The traditional marketing mode is too passive. Although channel sales has the initiative to choose channel providers, the sales depend on channel providers, resulting in a low stability. Although the travel marketing mode has high initiative, it highly depends on the business level of the salesmen; therefore it has fluctuations. The more targeted marketing strategy in the online marketing mode attracts more users, who spontaneously spread the product in their respective community networks, not only making the companies more attractive in scope and more sticky for users, but also making the cost of publicity lower and ultimately leading to a steady increase in turnover.

The limitation of this paper is that only the data mining algorithm was used to simply classify customers, and few subjects were analyzed; therefore, the future research direction is to deepen the data mining algorithms for customer information mining, and increase the objects for example analysis.

6. Conclusion

This paper briefly introduced the traditional marketing mode and the data mining-based online marketing mode and then analyzed company X which mainly sells books and company Y which sells properties. The results are as follows. It was found from data mining that the users in the network community mainly tended to buy five types of books, and the proportion was 43.1% for popular science books, 20.3% for historical documents, 11.4% for classical novels, 59.6% for modern novels and 70.2% for teaching materials. It was found after using the data mining algorithm that 7.7% of the home purchase customers prefer TV advertising channels, 4.1% preferred radio channels, 7.2% preferred outdoor advertising channels, 2.1% preferred newspaper channels, 44.8% preferred online platform channels, 6.5% preferred market auction channels, 5.7% preferred real estate trade fairs channels, 16.6% preferred intermediary channels, and 5.3% preferred other channels; 5.7%, 16.6% preferred intermediary channels, and 5.3% preferred other channels. The comparison of the turnover of the companies before and after applying the online marketing mode showed that the turnover of the companies under the traditional marketing mode fluctuated up and down with month and the turnover in the year after applying the online marketing mode rose steadily with month.

References

Schwarzl

Grabowska

. Online marketing strategies: the future is here. Journal of International Studies. 2015; 8(2): 187-196.

Wen

Dou

Chen

. Exploring the effectiveness of consumer creativity in online marketing communications. European Journal of Marketing. 2015; 49(1-2): 262-276.

Kannan

. Attributing conversions in a multichannel online marketing environment: An empirical model and a field experiment. Journal of Marketing Research. 2014; 51(1): 40-56.

Riasi

Pourmiri

. Effects of online marketing on Iranian ecotourism industry: Economic, sociological, and cultural aspects. Management Science Letters. 2015: 915-926.

Thaichon

Quach

. Online marketing communications and childhood’s intention to consume unhealthy food. Australasian Marketing Journal. 2016; 24(1): 79-86.

Watfa

Najafi

Bakkar

. Targeted online marketing using social networking. International Journal of Online Marketing. 2015; 3(3): 37-50.

Menelec

Jones

. Networks and marketing in small professional service businesses. Journal of Research in Marketing and Entrepreneurship. 2015; 17(2): 193-211.

Qing

Deng

. Research on network marketing reform under the background of big data. Journal of Physics: Conference Series. 2021; 1744(3): 032006 (6pp).

Stoica

Orzan

Dobrescu

Radu

. Online marketing communication using websites. A case study of website utility in accessing european funds in the tourism field regarding northeastern romania. Procedia Economics and Finance. 2015; 23: 926-935.

10.

Antonopoulos

Veglis

Emmanouloudis

. Online marketing for media: The case of greek news websites. International Journal of Marketing Studies. 2017; 9(2): 104-112.

11.

Popa

. Understanding students needs for a more effective online marketing in the higher education system. Annals of the University of Oradea, Economic Science Series. 2015; 24: 139.

12.

Howison

Finger

Hauschka

. Insights into the Web presence, online marketing, and the use of social media by tourism operators in Dunedin, New Zealand. Anatolia. 2015; 26(2): 269-283.

13.

Opreana

Vinerean

. A new development in online marketing: Introducing digital inbound marketing. Expert Journal of Marketing. 2015; 3(1): 29-34.

14.

Ding

. Analysis on the construction of online marketing platform for performing arts ticketing under the background of big data. IOP Conference Series Materials Science and Engineering. 2020; 750: 012014.

15.

Micu

Geru

Radu

. Analyzing user sentiment in social media: Implications for online marketing strategy. Psychology and Marketing. 2017; 34(12): 1094-1100.

16.

Joemsittiprasert

. Moderating effect of creditability on the relationship between online marketing communication, web quality and culture experience. Polish Journal of Management Studies. 2019; 19(2): 231-239.

17.

Falaix

Terfous

Corneloup

. Online marketing and erotic power: the case of Roxy – Marketing en ligne du surf et pouvoir érotique: Le cas Roxy. Recherches Féministes. 2019; 32(1): 127-146.

18.

Pavithra

Nayak

. Challenges in online marketing of dry chilli in selected APMC of Karnataka. International Journal of Current Microbiology & Applied Sciences. 2019; 8(1): 456-464.

19.

Khataibeh

. A study approach for online marketing over social media. International Journal of Advanced Trends in Computer Science and Engineering. 2019; 8(1.5): 96-107.