Research on Future-Oriented Manager Service Design Under the Background of New Retail

Abstract

At present, traditional management strategies in retail industry have gradually failed to meet the expectation of rapidly expanded consumers. Although some retail managers start to deploy new management strategies so as to establish better user connections with users, the management system is relatively isolated and not big enough to form big data. This paper begins with market research and user research, digs out the bilateral pain points of managers and consumers from the research, puts forward solutions and discusses the design and testing processes of establishing a manager service system taking consumers and managers as the main line under the background of new retail. The contributions of this paper to the research on new retail service design lie in a complete set of new retail service design process based on regional conditions and some discussions on how to serve bilateral audiences in service design.

1 First Section

1.1 A Subsection Sample

Previously, retail and e-commerce were two relative concepts, and until August 2010 when Alex Rampell, founder of TrialPay, a US payment company, introduced the concept of O2O (Online to Offline) [1], the online and offline united communication drew a new blueprint for a business model. In the area of e-commerce, Liu Xiaohong indicated that Groupon in the US, the earliest ancestor of the group buying model, is the business paradigm of the Dian Ping and Meituan in China [2]. Similarly, Lu Yiqing and Li Chen also claimed in the study of the O2O business model that the group buying model is still the prevalent O2O business model in China around 2013 [3]. What’s more, Sun Hua thought it has also profoundly affected the development of China’s mortar retail industry [4]. In October 2016, Jack Ma aired his new viewpoints on the Alibaba Computing Conference, in which he introduced the concept of new retail. New retail utilizes AI computing and IoT technology and requires that online enterprises divert warehouses via inventory retailization to arrange offline channels overall, while offline enterprises highlight the efficiency advantages of traditional retail by using intra-city logistics and delivery to make offline business online [5, 6]. Further, from the perspective of retail format, Zhou Yong believed that the unmanned shops like Amazon GO will increasingly cater to the preferences of consumers and are the future development direction of retail industry, but the completely unmanned shop model is not suitable for the development of China [7]. Despite a few managers plus smart services are the main development direction of future retail shops, there is still a lack of relevant research on a few-people smart shops.

There are some problems with the current consumer management methods as well. First of all, Liping deemed that apart from carrying out execution and management, obtaining the consumer data serves as the most important part in the process of transforming the new retail business [8]. Previous retail enterprises relied on print media, marketing and other traditional advertising media for publicity, but Ni Ning and Jin Shao pointed out in the research on communication strategies in the big data age that under the background of the current huge information flow, the one-way output of traditional media is difficult for consumers to notice and capture among quick and complicated information [9]. Second, when propaganda and planning come into play and after consumers make the purchasing behavior, retail enterprises can merely count commodity data but cannot grasp peer-to-peer consumer data [10]. Then, even though purchase items, price ranges, consumption frequencies and consumption time data can be tracked by using membership cards, too small sample size and low data value still exist [11].

In order to address this problem, now many offline retail enterprises conduct data mining plans and their strategies are focused on: serving members in the form of own APP to achieve the conversion of entity membership cards. However, the fragmentation of current mobile device APPs is severe, and the number of the APPs actively used by a user per day is only about 9, making it rather hard for managers to collect large amounts of user data thereby. In addition, all enterprises and systems are relatively independent, and if they make research and development separately, the development are both costly and in a long cycle, and more importantly, data cannot be shared, resulting in “big data” not that “big” [12]. Consequently, the disjunction and separation of marketing, sales and evaluation systems are common and pervasive in the transformation process of traditional retail industry.

Through literature review, we found that a few-people smart shop service system is still a blank area. Although managers and consumers directly related to a few-people smart shops are in a service provider-receiver relationship, under the consumer-centered service design, the problems of efficient management and employee-friendly management systems are ignored. Vice versa, under the manager-centered service design, there is a service gap between systems, and user experiences would be inconsistent and non-fluent. In summary, relying on the a few-people smart shop service system, this paper will tease out the problem in the bilateral service design between consumers and managers.

2 Survey and Analysis of Strategies and Scopes

2.1 Survey Method Selection

Via the qualitative analysis with logical thinking, we have a fuzzy preliminary estimate of the products that have been made - the a few-people smart shop service system, then the correctness of the preliminary estimate needs to be tested through market research, and we need to find the typical target audiences of the developed product, dig out the audiences’ pain points and potential needs through user research. Nevertheless, due to the particularity of the bilateral audiences in this study, except consumers, enterprises’ managers are not the target users in the general sense, it is thus necessary to discuss and determine the research methods in view of this situation.

The most commonly used online questionnaire method at present is excluded by us. Although online questionnaires have various forms, are low in cost and large in sample size, because of the particularity of the surveyed respondents, there is no guarantee that online surveyed people are the managers and participants of retail enterprises. Therefore, we predicted that the received questionnaires would almost be unconvincing ineffective questionnaires, which fail to guide our judgments positively.

Hence, after evaluation, we determined two survey methods: field research and focus group. We carried out a single 30-minute short-time and high-frequency field research respectively on five typical retail shops through the three dimensions of time, places and marketing activities, conducted a focus group interview of 5 people based on the survey data and conclusions of the field research, delved deeper into the problems found during the interview and discovered new problems during discussion.

2.2 Survey Data Analysis

Field Research Analysis

As shown in Table 1, it’s the problem observation and recording table during the field research on Lawson Convenience Store and preliminary analysis. As shown in Table 2, it’s the product characteristics recording table of five stores. Because there were too many different variable elements in different shops, the data of the five stores were integrated in the summarizing process. The discovered problems were discussed in depth in the focus group.

Table 1. Problem observation and recording on Lawson Convenience Stores

Table 2. Product characteristics of five Lawson Convenience Stores

Focus Group Analysis

Although the members in the focus group should be the testees with the same background and incomes, because the product requires a more macro management strategy and also the implementation details during the implementation process, the focus group consists of three direct salespeople, an assistant store manager and a store manager. For a clear expression, the needs and suggestions are summarized and listed in Table 3 in order of priority.

Table 3. Focus group summary

According thereto, we need to tap the immediate needs of audiences so as to find their real needs.

2.3 Retail Management Needs Tapping

During the focus group discussion process, it was found that some of the problems we predicted did exist, such as the slight differences in the thinking perspectives between employees and managers. From the managers’ point of view, the external target requires to increase the sales of high-margin products, and the internal target requires to reduce losses, cut down in-store expenses and reduce accidental goods damage and theft, thereby achieving the ultimate goal of maximizing profits. The thinking perspective of employees lies more in handling in-store issues more conveniently and efficiently and improving work efficiency, which directly and indirectly affect consumers’ shopping experience and return rate.

According to qualitative analysis, firstly, increasing sales needs to enhance consumers’ demand, so it is necessary to provide the goods more closely associated with consumers, which requires the sold goods match the consumers around the current store. Secondly, the reduction of losses needs to reduce expiry damage. As a consequence, not only the turnover rate of goods requires to be increased, but also products should be purchased in reasonable quantities. Both of these two issues need to be evaluated comprehensively based on the consumption needs, purchasing power, consumption desires and other relevant information of the consumers at specific locations so as to reduce the unsalable goods due to inaccurate judgment on surrounding consumers’ needs and incorrect goods ordering.

Besides, the discussion among the focus group led us to new discoveries. Both managers were preparing new stores, and they all claimed that the scientific assessment of new store site selection is extremely difficult. Generally, without the market research data used for quantitative analysis, retail managers can only draw conclusions by means of repeated field visits and extensive qualitative analysis, which may result in the imbalance of passenger flow and rent, seriously affecting enterprise managers’ return on investment. Therefore, the product we design should provide appropriate services ahead of audiences’ shop setting-up to enrich the entire service process and form complete and smooth new retail solutions.

As a result, to sum up, we found that managers’ most fundamental need is to have a tool that can help enterprises to understand consumers more accurately and locate potential consumption needs.

2.4 Retail Management Needs Subdivision

According to the problems collected during the process of designing survey and research as well as the potential needs, we tapped subdivision functions from the system service touchpoints, built a functional structure tree and wrote prioritizing needs documents, on which the iterative development orders in the future will be based as the priority thereof.

Table 4 describes that we divided the iterative development priorities according to users’ needs into four categories - necessary, optional, user-expected and future-feasible.

Table 4. Functional requirements priority

3 Analysis of the Service System Structure

At present, goods labels have been widely used in retail industry. As early as 1700, Europe printed the first batch of labels used on medicines and cloths for goods identification. From “Identification Label” to “Value Recommendation Label”, labels have been refined and personalized. Goods exist because of consumers and consumers have different attributes due to purchasing, and the two influence each other [13]. The labels of goods enable consumers to find the goods they want more quickly; whereas managers can efficiently recommend potentially demanding goods based on consumer labels. Therefore, in combination with the previous design research, the project will focus on consumer labels as the core to make a bilateral friendly service management application between managers and consumers – so that enterprise managers are provided with scientific management programs, and consumers can get high-quality service by using this system. Furthermore, data labeling not only can make data more intuitive, but build privacy protection policies to minimize the risks caused by the leaks of consumers’ privacy.

3.1 Service System Architecture Combing

According to the detailed user needs documents, the system architecture takes “searching for stores” - “operating stores” - “promotions” as user paths for information architecture combing.

According to the conclusions of the design research, this system takes the consumer label as the core to implement highly efficient and precise management service, so the above three journey phases are corresponding to three service tools: Consumers Community Analysis Tool, Consumers in-Store Service Tool and Community Marketing Decision Tool. The 3C tools include a number of sub-tools, and the main function hierarchy structural diagrams thereof are shown in Fig. 1.

Fig. 1.

Main function hierarchy structural diagrams.

3.2 Consumers Community Analysis Tool

As a service tool in the store searching stage, the consumers community analysis tool includes three sub-modules: respectively, (a) location-based community analysis, (b) consumer label smart learning and (c) consumer privacy protection tool:

• (a) This system has already been prepared for providing service for retail managers before their store setting-up. First, the insight into location-based consumers will predict the passenger flow of the site selected by managers and make scientific stocking recommendations based on community personality label attributes. Compared to the previous ways that managers relied on the observation at selected spots and in-person visits, etc., this system can help retail managers make site selections more easily and scientifically. Second, this system can effectively reduce unnecessary store losses brought about by inaccurate stocking during the trial run by analyzing the attributes of passenger flow around the store. Since the system needs to be based on a large amount of social label data, it should consider accessing the Tencent MIND with the most abundant community resource data in order to achieve the goal of precisely targeting communities.

• (b) Each purchase by a user is a piece of valuable data, so the purchase behaviors under this system will be transformed into label data to be fully prepared for the all-way marketing activities. The system needs to connect retailer’s own APP data to retain users’ habits; and multiple stores jointly manage and data are connected to one another… The data information of users gained through each channel is multi-dimensionally integrated to shape complete consumer labels.

• (c) Since the system involves multiple user privacy, privacy protection policies must be designed at the system hierarchy. Thus, in this system, the system unifiedly generates user identification codes for all users, and all user data exist in the form of labels. In this way, all user information will not be manually backtracked.

3.3 Consumers in-Store Service Tool

As a service tool in the store operating stage, the consumers in-store service tool includes four sub-modules: respectively (a) entering-store face recognition, (b) emotion recognition, (c) centrally controlled voice service and (d) electronic price tag service:

• (a) When a consumer enters the store, the cloud computing device inside the store starts to match and lock in the user, and retrieve the business interests under the user’s label library to see if there is a promotional product in line with the consumer’s interest label.

• (b) When a user views the products on the shelf, the user’s micro-expressions (positive emotion, hesitation or negative emotion) are identified at a certain frequency. When decision-making hesitation occurs, the voice assistant service is accessed at the electronic price tag to aid decision-making. Somber expression data would be recorded to reserve marketing analysis data (which will be used as the data source for goods personality analysis). To reduce development costs, the Azure cognitive service API is accessed in the storefront system.

• (c) Replacing traditional centrally controlled broadcasting, the voice assistant can play music based on in-store consumers’ music interests most of the time; and also broadcast targeted marketing contents to in-store consumers at a very low frequency. Broadcasted voice cannot be heard for being personal to someone, but the broadcasted marketing contents match the users’ interests.

• (d) Dynamic electronic price tags can conveniently adjust the displayed price and names of products in real time, and meanwhile can change interface display in some particular occasions, such as accessing the voice assistant image for users; the dynamic highlighting and flashing of promotional products and so on.

3.4 Community Marketing Decision Tool

As a service tool in the promotion stage, there are two sub-modules under the community marketing decision tool: respectively (a) online community marketing and promotion tool, (b) offline store stocking predicting tool. They are used to guide managers’ marketing activities and complete complex tasks via one click.

• (a) The online promotion module helps managers to reduce the barriers to Internet marketing technology, use the data of the consumers community analysis tool, provide the one-click production tool for promotion contents and directly access the landing pages on promotional platforms.

• (b) The offline decision tool can provide managers with optimized stocking suggestions. A comprehensive analysis is carried out according to store sites, community labels, in-store consumption situations, consumption decision-making hesitation data, expired goods destruction and other data, and the longer the system is used, the more complete and the cleverer the self-learning of the system is. In addition, the tool is also directly connected with the stocking warehouse, so that it can be adjusted according to the stocking suggestions and generate a stocking plan via one click. For long-term backlogged goods, it will suggest stopping/reducing stocking, provide promotional advice so as to reduce losses.

4 Service System Interaction and Visual Design

Interaction and visual design processes are designed in accordance with the high-level, multi-channel, interactive-card design strategies. During the prototype design phase, according to the testing results of the alpha version, we need to visualize a large amount of data, simplify the bottom button and summarize the functional structure tree as three sections of “Dynamic Billboard”, “Cloud Service” and “Me”. Moreover, the “one-click completion” button is upgraded to the card to reduce click paths and improve operational efficiency. The high-fidelity prototype of the second version of shopkeep APP is shown in Fig. 2, and its visual design is shown in Fig. 3.

Fig. 2.

Shopkeep APP high-fidelity prototype.

Fig. 3.

Shopkeep APP visual design.

The interface color of the design can be changed according to stores and personal preferences, and the use of linear design allows users to better focus on the changes in data to meet the users’ “exhilarating points”. Also, most of the three-level and four-level operations are integrated into the one-level and two-level pages, greatly reducing the user path and click depth plus improving the operating efficiency by 45% in the case of continuous operation.

5 Service System User Testing

5.1 Cognitive Walkthrough and Think-Aloud Protocols

The MOCKINGBOT platform is utilized in the process for prototype testing, allowing 12 investigators to traverse and speak out their ideas simultaneously to test the various functions of the user label management system and fill in the SUS and NPS scales after the testing is finished.

In the testing process of Cognitive Walkthrough and Think-Aloud Protocols, the entire course is videotaped, and recording focuses on the part contrary to testees’ expectations. After sorting it out, it was found that three testees did not know functions of the page data of the consumers community analysis tool; and one testee doubted the reliability of the expiration-reminding function.

5.2 SUS Testing

After traversing the system, the testees filled in the System Usability Scale, and the obtained SUS score was 82.5. The specific scale data are shown in Table 5.

Table 5. SUS scale results

What’s more, the score of learnability scale calculated based on the data was 83.75; and the score of usability scale was 82.19, which has increased by 10.6% compared with the testing results of the alpha version. The increase in SUS score shows that the service system provides a clearer interaction structure design after the second iteration, effectively lowering testees’ cognitive difficulty and learning threshold.

5.3 NPS Testing

After the SUS testing, we collected testees’ Net Promoter Score - 8.6, which has increased by 22.1% compared with that of the alpha version - 6.7. We interviewed three testees who participated in the testing twice. They held the application interfaces of the second version are more comfortable and modern, clearer in the differentiation of functions in the visual senses and easier to learn and use. Besides, we asked the testees who scored relatively high and low in terms of NPS about their scoring reasons, among which the testee scoring 6.7 said the application functions are not prominent enough to easily find the function he needs; and the testees scoring 9 to 10 said similar tools for guiding management are extremely scanty in the market, and the subdivided functions will be a tremendous help to them.

6 Conclusion and Prospect

After accomplishing the prototype design, multiple design testing is required to make the whole system more reasonable and easy to use. During the testing, it was found that the older testees with poorer learning ability have lower acceptance of the system, so the subsequent iterative design needs to focus on strengthening learnability.

This paper mainly discusses a manager service system and the entire service design process of its subsequent mobile application, starts with the social insight and demand survey to grasp the function needs of the system, establishes the system concepts and digs out the subdivided demand points in order to sort out the structure tree of the entire system, and finally in accordance with interaction strategies, conducts rolling design - testing - feedback – design on the system prototype to continuously improve services and application. Although the consumption characteristics and managers’ business vary from region to region, the overall research framework and design process are of sustainable use and reference.

Furthermore, the limitation of this study lies in this system is a future-oriented manager service system in retail industry. Therefore, some modules and functions of offline stores currently require a relatively large amount of investment to be executable, but it’s still feasible. The costs of the system will decline as technology costs in the future drop, enabling the business systems in smart cities to run more efficiently and provide better services by making good use of the big data at the retail level.