This article introduces several cross-border e-commerce models commonly used by sellers, such as shopping basket analysis, RFM model, funnel model, and Boston matrix model, to achieve high-quality sales growth and product layout optimization through accurate data analysis. Now you can find these templates by going to [Digital Cross-border BI], and click on the template link to reuse them with one click.
There are many seller operation partners who have communicated with Xiao Jiu, and usually in data analysis, in addition to the most basic gross profit, conversion rate, click-through rate and other addition, subtraction, multiplication and division calculations, other analyses do not know where to start. In fact, every data analyst not only needs to be proficient in analysis tools, but also needs to master some common and valuable data analysis methods and cross-border e-commerce models.
In order to help you better get started with cross-border e-commerce data analysis, Digital Cross-border BI has sorted out several commonly used data analysis methods and models, and attached reusable templates, so that you can easily master all kinds of high-level data analysis, let's learn together!
It can be used when analyzing the sales of cross-border ECShopping basket analysis model
The shopping basket analysis model is a common data model in retail and e-commerce, which treats all combinations of goods purchased in a shopping behavior of each consumer as a "shopping basket", and then analyzes the data of different product combinations in the shopping basket to reveal the correlation between products.
There are three key metrics for basket analysis:1) Support, 2) Confidence, 3) Lift
In order to give you a better understanding of shopping basket analysis, Xiao Jiu will give you an example: Suppose a supermarket sells 100 items in a month, of which there are 5 orders for product A, 10 orders for product B, and 3 orders for goods A and B at the same time.
Probability of buying product A: p(a)=(5 100)*100%=5%.
Probability of buying product B: p(b)=(10 100)*100%=10%.
Probability of buying both A and B products: p(A b)=(3 100)*100%=3%.
Through basket analysis, we can attract customers to switch from buying only one product to buying multiple products, thereby increasing the sales amount of the entire shopping basket, maximizing sales growth, and also discovering some new cross-selling and ** sales opportunities:
Consider whether there is a possibility of matching sales;
Set up a coupon for several products to guide each other and increase sales at the same time;
Do targeted advertising, consider contextual advertising;
Upstream and downstream product development.
When conducting shopping basket analysis, we only need to record the SKU and order status of the product, and through simple probability calculation, we can get the support, confidence and promotion of various products, so as to optimize the product layout and promotion strategy.
The RFM model is a commonly used cross-border e-commerce model to measure customer value and profitability.
The model can help enterprises segment customers and identify the most valuable customer groups through the combination of three indicators: the time of a customer's last purchase (recency, r-value), the purchase frequency (frequency, f-value) in a period of time, and the amount of consumption (monetary, m-value) over a period of time.
In cross-border sellers, repurchase rate analysis is often used to complete the basic RFM analysis. Through the calculation of repeat purchases, we can better understand the consumption behavior of customers and explore the needs and interests of users for goods or services. At the same time, it can also be judged according to the customer's repurchase situation, and the aspects that the store needs to focus on improving.
When conducting repurchase analysis, we only need to record the customer's purchase time, purchase frequency and consumption amount within a certain period of time, and make statistical calculations, so as to determine the type of customer according to the RFM score. Generate marketing strategies for different types of customers.
The funnel model is also one of the very commonly used cross-border e-commerce models.
Generally speaking, the funnel model is the core process step that the user behavior path has to go through, and the whole funnel model is to first split the entire purchase process into steps, then use the conversion rate to measure the performance of each step, and finally find out the problematic link through abnormal data indicators, so as to solve the problem, optimize the step, and finally achieve the purpose of improving the overall conversion rate.
Advertising insights often use funnel models: they reflect the number of customers in a campaign, from impressions, clicks, add-ons, to orders. From the largest impressions to the smallest orders, the funnel model visualizes the churn process.
When analyzing product quality, you can use the [Boston Matrix] to analyze products by combining advertising data and return data. The Boston matrix uses two-dimensional four-quadrants to divide products into four types: star, Taurus, skinny dog, and problem.
We can develop the platform's product matrix in the manner of the Boston matrix, and measure the advertised products when analyzing the productsThe cost of a single orderwithroiManifestation.
ROI: The return on investment, the ratio of ad sales to ad costs.
Cost per order: The total cost of products associated with the campaign divided by the total order volume.
According to the analysis of these two dimensions, the advertising products are divided into four quadrants, and the middle line of the division is the average of the individual order cost and ROI, and it can be found that the advertising products in the upper left corner have worse performance, because these products have high order costs and low ROI. The closer to the fourth quadrant in the lower right corner, the higher the ROI can be achieved through lower costs, proving that it has the highest value and can continue to increase investment. Find our very own celebrity and skinny dog products.
The above cross-border e-commerce models can help cross-border e-commerce sellers draw more accurate conclusions for different analysis goals, make a more structured and systematic data analysis process, and achieve high-quality analysis.