One formula and five indicators help you build data analysis thinking as a product manager!
|
Question: For improving the data analysis capabilities of products, are there any specific suggestions and good learning resources (for e-commerce)? I beg for some helpful advice! Answer 1: Mr. Gu, Taomi Product Manager 1. Understand the business and be familiar with the data framework and system Understand what your business does, what your business development plan is, what the core measurement indicators are, list the KPIs or core indicators. Generally, there are only a few key indicators. Then break down the core indicators. This also needs to be done based on the attributes of your business. What are the elements of your business that will affect this indicator? The advantage of splitting is that you can clearly understand the composition of a specific indicator, just like how a butcher cuts up an ox. Of course, this process can be broken down further by adding some common attributes, such as time, user gender, user age, user occupation and other common dimensions for further detailed segmentation. 2. Think about existing data indicators; analyze them in multiple dimensions to find patterns After becoming familiar with the indicators and frameworks that the product needs to pay attention to, understand the operating status of each existing indicator. It would be better if there are indicators for comparison with the same industry to see if there is room for improvement. Or, you may want to improve a certain indicator and how much it will be improved through a certain operational action; through a series of relatively accurate hits, you can estimate to what level the operation can improve the indicator. Another benefit of having a precise model is that after understanding your core users, you can conduct product research and demand exploration for this group of users separately, which will help you determine which indicators can be improved by what means; at the same time, find patterns and find ways to do some analysis on the broken down indicators. I personally feel that the analysis here does not necessarily require very complicated means. What is more important is a feeling and awareness. 3. Verify the rules and summarize the experience Once you have found the pattern and understood it in your heart, you will be more clear-headed next time you do things and have a deeper understanding of the product. For many things, you become familiar with them little by little, and gradually develop a sense of familiarity through in-depth studies. Data is just a topic that brings you and your product closer to each other. For more advanced and in-depth data communication, it is better to leave it to professional data processors, just like not everyone is a psychological counselor. In short, for PM, I personally feel that data is a kind of consciousness rather than technology, a summary of methods rather than theoretical science. Paying attention to data is an advantage. Every product manager needs to understand the simplest formula before designing a product: Product value = revenue from the product - design, R&D and operating costs > 0 For example, for points-based products, if the net increase in sales after using the points product * interest rate - the value of goods offset by points (operating costs) - design and development costs > 0, and if users will use the points for a long time, the design and development costs can be ignored, and other data can be obtained relatively easily. For example, for products with page redesigns, if the number of loyal users brought by the redesign * the value of each loyal user - the design, R&D and operating costs of the new page > 0, it means that the redesign is successful. As long as the product manager grasps this basic formula, other in-depth data analysis can be left to more professional personnel. The product manager's main focus should still be on user needs analysis. (The above answer is slightly abridged. If you want to see a detailed example version, please click to read the original text) Answer 2: mrjesse e-commerce product manager It has been explained in detail above, let me add: As an e-commerce product manager, after all, we are not a professional data product manager. We don’t need to understand too deeply, we just need to focus on the most core content. Before clarifying data analysis, we must first clarify several major business indicators of data analysis of e-commerce products: 1. User perspective 1. Total number of active users, number of newly registered users, and total PV; 2. The number of active users, number of newly registered users, and total PV in a certain period of time; 3. Conversion relationship within a certain period of time; 4. Registration conversion under different channels; 5. Subsequent retention rate of registered users under different registration channels; 6. Distribution of repeat purchases by users who have purchased within a certain period of time 2. Order perspective 1. Today’s total number of orders, sales, order unit price, number of users with orders, and subsidy ratio; 2. The total number of orders, sales, order unit price, number of users with orders, and subsidy ratio for each day in the past week; 3. Average delivery time of orders per day in the past week; 3. Product perspective 1. According to the product group, the number of views, number of purchasing users, number of orders, and sales of each product today; 2. By product group, the number of views, number of purchasing users, number of orders, and sales of each product every day in the past week; 4. Category perspective 1. Group by product category, including the number of views, number of purchasing users, number of orders, sales, and order unit price of each category today; 2. Group by product category, with the number of page views, number of purchasing users, number of orders, sales, and order price of each category every day in the past week; 5. Store perspective 1. According to the store group, the number of visits, number of purchasing users, number of orders, sales, and order unit price of each store today; 2. Group by store, the number of visits, number of purchasing users, number of orders, sales, and order unit price of each store in the past week; After conducting data statistics based on the above indicators, we will conduct data analysis: First, data analysis and modeling. We know that e-commerce product models are generally modeled based on event (click, browse, etc.) user attributes. We then proceed to formal data analysis: Method 1: Multi-dimensional data analysis We need to define some events, such as canceling an order, submitting an order, paying an order, browsing products, adding to a shopping cart, and so on. Then based on these events, we need an indicator, such as the number or the total. But we know that having a data compass is not enough, we need to segment the data. Here we have created events and indicators. Therefore, it is also necessary to filter the user's attributes, such as city, device used, payment method, and source channel. We often use this method for user profiling, user behavior analysis, data anomaly detection and analysis, etc. Method 2: Conversion rate data analysis When we conduct an event, we need to evaluate how many people have registered in the past two days, what the order conversion rate is, and what the payment rate is. We need an analytical method. From what I just said, we are based on event analysis, so we can define an event, filter the time, first define the event (registration), then define the event (submit order), and then define the event (pay order) again, we can get a conversion rate. This method is often used in conversion rate analysis, also known as funnel analysis. Method 3: Retention Data Analysis Retention analysis means retention as its name suggests. We need to analyze the data of users over a period of time. For example, if we have organized an event, we need to look at the registered users during that period and the order submissions, how many submissions were made on the first day, how many submissions were made on the second day, how many submissions were made on the third day, etc. Generally speaking, for data analysis, we first define an event (such as registering a user) based on our data model, and then define an event (such as submitting an order) to obtain a certain ratio of events. We often use this method to observe the retention of certain events. Method 4: Active or return visit data analysis We know that we have defined the user behavior data analysis, but we need to look at the number of times an event is used over a period of time, or the user usage in a certain place, so what should we do? This is what I call active data analysis. We first need to define an event (such as registering a user), define a situation for an event (such as submitting an order) (usually the number of days here), and then we filter the user's event and get a data. We often use this method to investigate user usage, and it is also a key data analysis indicator for measuring user activity. Having said so much, what can we do? 1. Check data anomalies and check them one by one 2. Improve conversion rates on key pages 3. Activity evaluation and channel data analysis and evaluation ASO optimization service click link: ASO optimization service introduction IOS accelerated review click link: //www.opp2.com/8854.html APP Top Promotion (www.opp2.com) is the top mobile APP promotion platform in China. Welcome to follow our official WeChat public account: appganhuo [Scan the top APP promotion WeChat QR code to get more dry goods and explosive materials] |
>>: 5 simple visualization tools to make your data charts look great instantly!
Recommend
Example: How to design an operational activity H5?
This article can meet the following needs: Want t...
The game genres with the most development potential in the next ten years
In May this year, I mentioned at the end of the a...
Guangxi border residents can have four children! What are the specific regulations? Will there be a fine for having a fourth child now?
At present, my country has relaxed the three-chil...
Nanny-level Zhihu Marketing Strategy [7000 words full process]
Zhihu marketing is an issue that almost every bra...
Understand the rise, turning point and breakthrough of Internet finance in one article!
The rapid rise of any industry is not accidental....
What are the benefits of bidding hosting and agency operation?
There are also four major advantages of bidding h...
Coca-Cola Marketing Law
When you see this billboard , can you recognize w...
What is the algorithm of Tik Tok’s recommendation mechanism?
The reason why Tik Tok is so popular is definitel...
6 ways to use countdown posters, how many have you seen?
Starting from July 21, Mr. Jia Yueting, Chairman ...
Writing copy is not to attract attention, but to occupy your time!
The purpose of writing copy is to gain as much us...
Insurance industry uses Tencent Guangdiantong as an example!
Driven by the improvement of national insurance a...
Analysis of Tencent Advertising’s Bidding Mechanism
With the development of Internet technology, more...
New App Promotion Techniques: Video Promotion Makes Your App Popular Overnight
Everyone on the Internet is familiar with Chai Ji...
Teach you how to write a perfect product promotion plan in 6 steps! (2)
We will teach you how to write a perfect product ...
Gaizhou SEO training: Keyword SEO optimization is a necessary step to accelerate the development of corporate websites
Most entrepreneurs always have such questions: wh...