Wang sir, 11 years of experience in Internet operation, CDA data analyst license holder, has successfully operated 2 e-commerce stores with an annual turnover of 50 million yuan, and has done bidding ranking and search engine optimization for 2 years.

Data analysis is the most basic skill in e-commerce operations. To put it simply, data analysis is to find problems and provide data reference for solving problems. Experienced data analysts know that it is better to trust data than to have no data. The data lies where it lies. The key is that before analyzing, you should have a clear thinking logic: why do you analyze the data? What do you hope to get through data analysis?

The general analysis data logic is as follows: determine the purpose of analysis -> collect data -> sort out data -> analyze data -> get some analytical ideas

Today, based on the actual e-commerce operation cases, I will tell you these three practical data analysis methods:

1. Funnel analysis model

2. Trend analysis method

3. TOPN analysis method/two-eight principles

01. Funnel analysis model

The most commonly used data analysis model in operation work is the funnel data model. I believe everyone is familiar with this model. The reason why I mainly use the funnel model in my work, or all the market operation and promotion business of promotion and transformation are inseparable from the funnel model, because conversion is the goal of our work. And motivation.

For example, an e-commerce platform set up a four-step conversion funnel according to the path of "enter the registration page - start registration - submit verification code - registration successful". Through data analysis, it was found that the conversion rate of the second to third steps was low, and many users lost in this link, resulting in a significant reduction in the number of users who successfully registered in the end. The problem link was located in the "start registration" - "submit verification code" link.

But the current situation of the problem is like this. What is the reason for the large number of users in this link? We made some assumptions:

1. Is it related to the platform used by the user? Is there any difference in product function design between PC and mobile?

2. Is it related to the mobile phone platform? Is there any difference between Android and iOS users in this link?

3. Is it related to the browser? Are there any bugs in verification in different browsers?

4 Other kinds.

The above assumption is to split this problem from different dimensions, and then look at the conversion funnel of users in each dimension?

The analysis found that the number of user registrations and the registration conversion rate of Chrome browser are much lower than those of other browsers. Compared with each step of conversion, it is found that the conversion rate from the first step to the second step is not significantly different from others, and the conversion rate from the second step to the third step is very low. Most users do not submit the verification code, but directly Left the page.

This strange conversion leak immediately attracted attention. The test found that there was indeed a bug in the Chrome browser in obtaining the verification code, which affected the user's registration. After the research and development solved this problem, the registration conversion rate under the browser was significantly improved.

As an operator, you must have the ability to analyze data, otherwise you only have "feelings" and "ideas" when looking at the data in the background, and you are not clear about how to do it. The reason why I took the CDA data analyst exam is that I have encountered too much data in the operation work, but I can't let the data guide decision-making, and the analysis is not in place. It is really recommended that all e-commerce operations and products have taken the CDA data analyst exam. In the preparation process, you should learn the business data analysis model, Excel, SQL and PowerBI data visualization. After learning, it will be of great help to work. A person who understands the operation of data analysis tools can really kill most colleagues in seconds. The advantages are too obvious.

02. Trend analysis

The trend analysis method, also known as the comparative analysis method, is the horizontal analysis method, mainly based comparison or ring comparison through the same indicators or ratios of continuous data to derive their change direction, amount and magnitude to perceive the overall trend. The microtrend may last for a week, while the macro trend may last for a quarter.

The best way to find data trends is to use time series. In the time series, the x axis is always time, and the y axis is any variable (total revenue, sales, etc.) to be measured. There are mainly analyzing latitudes: period-time trend, day-by-day trend, week-by-week trend, month-by-month trend, season-by-season trend.

Trend analysis focuses on the changes in sales over time. Time series enables us to easily find patterns in data:

Will the sales peak or decline on the weekend?

Are there any abnormal peaks in the data?

As time goes by, will the sales increase or decrease?

How does the market change over time?

Case:Why did the sales volume show an upward trend from July 2021 to December 2023? Analyze the overall trend of monthly sales in detail, as shown in the figure below.

Then make a horizontal comparison of the data of three years, such as the figure

Through trend analysis, it is found that there are different degrees of peaks in May, September and November every year. November is the peak, and the overall trend is upward. Next, why did the sales volume rise?

Reviewing the marketing activities of previous years, the Double Eleven promotion in November every year is the reason for the increase in sales. However, for product promotion, in actual work, the cost issues need to be further considered. For the control of ROI, blind promotion is not advisable.

03. TOPN analysis method

The TOP-N analysis method is usually used to analyze the contribution of customers, stores or products to the whole. This method helps enterprises focus resources and attention on the most valuable parts by selecting the top N data for in-depth analysis, so as to improve efficiency and profitability.

Case:

This is the data analysis of the background of a cosmetics e-commerce. Combined with the business background, sales amount and sales volume, we can find:

1. Because the customer unit price of whitening products is the highest, although the sales volume is not the largest, the sales and profit are the highest in the product line.

Two. Acne products are the best-selling and are the most popular series of hot products.

Next, we need to analyze the specific products, and we can analyze the main types.

By listing the sales details of the three categories, we find that there are 1 to 2 core products in each category. The final sales amount structure of different series of goods will be different due to the factor of customer unit price. According to this analysis result, the product structure can be optimized.