Managers at all levels need more manual tabulation in order to make decisions, and the final decisions can not flow back into the system, and it is difficult to quantify and track decisions.
The retail center is the front line of the retail enterprise and is also the hub of retail decision making. In order to realize the vision of completing data-driven decision making, I plan to empower enterprise data decision making by designing a data analysis and decision collaborative system that serves the retail center to help better integrate data analysis and business strategy.
The significance of big data analytics is not to present cool reports, but to clearly organize the layers to open the true face of the business and discover the problems in the business.
Before the design, we can take a look at how data decision-making is built in existing data decision-making platforms.
What are the core needs being met?
The company purchased the Zero Hopper Big Commodity system is built on the basis of processing commodity management business, to build the business control model of each work node.
Although it is a system that fully serves the merchandise, ultimately the good or bad of the merchandise is still reflected in the retail side, and all the merchandise strategy is also based on the retail results.
In terms of retail data analysis, this system consists of four modules: district monitoring, district store monitoring, quality control monitoring, and band monitoring; the tracking of retail results can be drilled down from district to store, and ultimately, the channels can be matched with the right commodities through the redeployment.
The analysis of commodities is holistic, with sell-out and discount control as the hero.
When the retail department looks at the data, small changes in individual stores generally don't affect the big picture, but changes in individual stores can be fatal to the store itself, and failure to make timely adjustments can lead to the store's demise.
The management system suitable for merchandise, which is conducive to the overall adjustment of merchandise, is slightly less supportive of retail data analysis.
Based on the TOC theory of commodity flow full life cycle operation management system, by the life cycle system, burst peak flat lag, dynamic pricing, forecasting system of four core engines to support the sales control, library control two business modules.
This system is a typical report-based system, which is the four core engines:
The life cycle also shows the retail performance of a commodity in the five stages of the pre-pre-pre-season;
The bursting of prosperity and lagging is a data indicator that is processed and then categorized and graded with the label of bursting of prosperity and lagging, so that the various dimensions of the commodity can be evaluated in terms of bursting of prosperity and lagging;
Then if it is found that the merchandise retail results did not go as expected, the business people can intervene through dynamic pricing to ensure that the final goal is accomplished;
The prediction system has not been much practical effect.
The combination of merchandising and retailing is considered in this system, and based on data analysis, it is not purely a matter of transferring goods; rather, it is a matter of adjusting the retail operation strategy to achieve the planning goals, and rolling the plan through data feedback to achieve a controllable retail result.
I agree with the design of this system, but I think the system is a bit too big a step. In the report presentation as the core function of the system, has begun to study how to predict the business through Ai, automated adjustment of business strategy, which seems to be a little too radical.
At present, the company in the understanding of the business situation for data analysis, the main source of data is the company's ERP report center.
The report center provides: retail summary report, retail summary (headquarter cost) report, store business daily report, store operation analysis, retail payment method report, retail price range analysis report, store time period performance analysis, store monthly business index analysis, store daily business index analysis, store daily comparison report, store weekly comparison report, store monthly comparison report, product mix and match sales analysis, weather and temperature retail sales analysis, store daily comparison report, store weekly comparison report, store monthly comparison report, product mix and match sales analysis, store monthly analysis report, store monthly analysis report. Matching Sales Analysis, Weather & Temperature Retail Enquiry Report, Merchandise Retail Ranking, Store Good & Late Sales (by Item) Analysis, Store Good & Late Sales (by Barcode) Analysis, Store Warehouse Retail Ranking, Salesperson Ranking, Employee Performance Analysis, Retail Analysis Report, Store Order Analysis Report, Retail Tracking Report, Store Order Analysis Report (Excluding Returns & Exchanges), Owned & Affiliated Retail Analysis Report, Store Store Warehouse Retail Merchandise seasonal percentage reports, retail category percentage reports, and so on.
These reports from multiple dimensions and statistical caliber reflect the specific product sales, sold pieces, style, year-on-year, chain, completion rate, ranking and other data indicators. Based on these indicators combined with labeling data, you can analyze the three dimensions of people, goods, and field.
Often used based on DuPont analysis for business dismantling, based on the data on the judgment of the development of retail strategy and commodity strategy, the formation of the corresponding solution to the problem.
The above three cases, respectively, from the pure commodity management, commodity + retail, retail three types of system data analysis system discussion. We can clearly see: merchandise and retail intermingled influence, but also need to peel off layers to find the root cause.
In order to increase retail performance, drive business growth can be done through standardized synergistic level-by-level data analysis, all data from the system, all decision-making return to the system. The strategy of the data decision-making system is to build a decision-aid platform developed by BI>BA>AI step by step, is the relevant strategy in the above case suitable for the data decision-making system?
In the following article, I will discuss how to build a data analytics system as an example of a retail data analytics system.
"All business data, all data business", the process of enterprise development are experiencing business standardization and business data.
The next "data business" is to integrate the data that has become an asset into the process of creating business value as a means of production, so that it continues to generate value, and in the process of combining data and business also encountered a great challenge:
1) Huge demand for data analytics tabulation
The conventional data analysis system in a retail enterprise is a very important part of the business process. Regular data analytics in a retail enterprise can provide standardized reports for initial judgment of business conditions.
When you need to dig deeper into the data issues, you need to combine more data indicators, and by determining the causal relationship of multiple variables, you can find the influencing factors that trigger changes in performance, and then formulate activity strategies.
But the combination of multiple data indicators produces a staggering amount of reports, and two indicators under two dimensions can draw N reports, which is why cousins work so hard.
2) Analysis and business execution synergy difficulties
The fundamental data source for data analysis is the terminal, and the final execution of business decisions is the terminal. Business decision makers need a perfect feedback path and strategy tracking path to observe and listen to the frontline voice and implement the operational work.
The retail center, as the hub of enterprise retail decision-making, needs to discover problems through data analysis to make scientific decisions and promote the complete implementation of strategies.
And through the systematic module to serve this business chain, so that data and decision-making can come from the terminal, to the terminal to form a complete closed loop of data and decision-making, to help data-driven business.
Help the business to understand the current situation of operation, track retail dynamics, efficiently analyze data, and form store diagnosis;
Support the data analysis work from terminal store staff to regional managers to headquarters management, and be able to pass the problem level by level;
Feed back to the system the marketing plans and response programs formulated by each level, to form a closed-loop of data and decision-making;
Cultivate employees' data management awareness, and drive the whole staff to be data-driven. Employee data management awareness, driving the whole staff to think based on data.
In the enterprise, there are many kinds of sales channels, brand-based enterprises will have self-operated, affiliate, hosting, distribution and so on. In this article, take self-owned retail as an example, retail analysis according to the enterprise management structure, by the terminal clerk, regional manager, regional head, headquarters level by level analysis report.
This model divides the roles through the organizational structure hierarchy, relying on the hierarchical relationship of the organizational structure, configure the account data permissions, can meet the business needs.
As shown in the figure, in a retail analytics system, there are four levels of organizational partitioning from top to bottom:
Account 1 is the -Headquarter management role, in the position of the root node, the scope of data permissions is "all" departments;
Account 2 is the -Region management role, data permissions range is "all" departments;
Account 2 is the -Region management role, data permissions range is "all" departments;
Account 2 is
Account 4 is the -Store Employee role with data permissions for "this store and comparison stores".
Different roles have different data permissions, and the data may be presented differently due to different concerns under each role.
1) Terminal store-level data analysis needs
The store manager's study of the data to be committed to how to improve the performance of each week, there are the following tasks need to be carried out:
Check the completion of the performance of the store;
Check the differences between the store's TOP models and the region's TOP, to find out if the store can also make a breakthrough;
Check the success rate of the trial sale of each style;
Check the success rate of the trial sale of each style;
Check the success rate of the trial sale of each style. The success rate of various styles of trial sales, and put forward specific replenishment proposals;
For slow-selling models to analyze the reasons, develop further programs, if necessary, to the district and headquarters to apply for assistance.
2) Sub-district level data analysis needs
Partition in the work of the summary, mainly to find the reasons for the fluctuations in the performance of the district:
Analysis of the performance of the district (single week year-on-year ring, cumulative year-on-year ring) differences: If the report presents the results of the growth of the report to summarize the reasons for the performance of the good, and review the specific actions, the success of the experience along to the next week; If the figures are not optimistic enough to find the root cause of performance setbacks, targeted measures, efforts to enhance the week's retail;
In terms of style, to check the TOP models SKC list of the district with other divisions TOP there is no difference between the differences in the models to other divisions to learn from the experience and internal dissemination.
3) Region level data analysis needs
Region in the work of the summary, mainly to find the region's performance progress:
analysis of the region's performance (single week year-on-year comparison, cumulative year-on-year comparison) differences in the completion of the progress of the dynamic pricing, adjust the promotional strategy and frequency;
in terms of goods, to see the region's year seasonal categories and other attributes of sales proportion is appropriate. Attributes of the sales ratio is appropriate to determine whether the sales rhythm is correct, the implementation of reasonable replenishment and clear promotion plan.
4) Head office level data analysis needs
When the head office is doing a summary of the work, it is mainly to analyze the progress of the completion of the strategic objectives:
Analyze the progress of the performance of the region, and according to the progress of the completion of the activities to give support, discounts, or clear the promotion of activity authorization;
Confirmation of the financial cash flow situation, to protect the financial health;
In the case of the goods. Analyze the progress of inbound inventory and category structure distribution, and make adjustments by chasing orders appropriately;
Follow up on the progress of strategic projects, and assess the value and results of the projects.
5) Information feedback business needs
Management at all levels in the process of data analysis needs to know what is happening at the terminal, in addition to the terminal also has the demand for feedback needs, and finally in the management decision to reach the terminal it is also necessary to follow up on the implementation of the task:
Terminal self-assessment of the business information to report;
Terminal competitors to report on the business situation;
Management task execution tracking.
1) Terminal store user demand analysis
2) Sub-district supervisor user demand analysis
3) Regional manager user demand analysis
4) Head office decision-making demand analysis
The final presentation of the data analysis is the key indicators, and the business people are also based on these indicators to grasp the current status of the operation.
The most important thing to determine the core indicators is to find the right core indicators, a commodity in the retail industry in different life cycle stages of the core indicators of the operating results of the marketing weight is constantly changing, to reconsider the weight for the business needs.
The drill-down function allows business people to drill down to different granularity of any combination of indicators as needed.
Enhancing the data analytics capabilities of the entire workforce has a variety of value to the organization, not necessarily in terms of directly increasing revenue, but also in terms of lowering operational costs, and even more often in terms of lowering the risk of decision making and the failure rate.
But many employees, especially terminal employees may not have this awareness, in order to better grasp the terminal situation, through the points system to pull employees to participate in enterprise data management.
1) Store points issuance rules
In the daily use of the platform, many products have membership levels, but B-side products in order to improve the enthusiasm of employees to participate can also learn from this model.
In the data analysis system to follow up on the interaction to distribute points, points can be issued from the point of view of generating value for the enterprise, there are the following cases:
Self-analysis of the conclusions and competitor information to mention;
The conclusions of the analysis were adopted.
Points in this system in accordance with the following rules:
The number of points can be added to the project or modified in the future according to the actual situation
2) Store points entitlement
In order for the store to continue to participate in the analysis of the data, but also to allow the new stores to actively participate in the design of points based on points of the redemption rules to consume points. Points entitlement design is an important step in guiding employee participation and strengthening employee habits, and the points are summarized monthly as a reference value for judging the month's performance.
Points benefits include:
Bonus or honorary awards according to the monthly points ranking;
Redemption of exclusive coupons: one-time consumption, exclusive to members, non-refundable, with validity and consumption restrictions;
Participation in headquarters data analysis training.
Points clearing rules:
Rolling monthly clearing of points from 3 months ago.
Costing: The direct cost to the organization is the investment in points, which is ultimately settled financially in the form of coupons.
Employees get points mainly by submitting analysis reports and reporting on competitor dynamics, which can be normally constrained under the company's management structure to avoid malicious brushing of points. In order to prevent the behavior of improper access to points, you can bury the data, monitor the amount of information submitted and the amount of information adopted, when the adoption rate is too low to warn or punish the store.
1) End-store user data analysis
2) Sub-district supervisor user data analysis
3) Regional manager user data analysis
4) Head office decision-making data analysis
4.2 Decision-making Collaboration
Selected a few typical functions to do the prototyping demonstration, and more prototypes are still in the drawing process:
Retail The business is in a constant iterative process of summarizing the last week's work and strategizing about the next week's work. Interpreting retail reports is all about analyzing problems from the data, and retail data is very informative, with multiple metrics that are recursive and correlated.
The author from multiple roles, from the point - line - surface Multi-angle reaction to the results of the operation, to help analysts understand the status quo, to find business problems.
In the analysis at the same time through the decision synergy module presents the terminal judgment of the business, so that decision makers can see the insights of the terminal; and you can follow up on the implementation of the strategy to track the implementation of the business, so that the terminal to understand the decision maker's ideas, so that data-based decision-making into the business process, so that the iteration of the retail strategy can be traced.