How to analyze data in product operations?

Today we will talk about the last element of the four elements of product operation , data analysis . Speaking of data analysis, I believe that no matter whether you are doing product operation , user operation or promotion, it is no stranger to it. Data analysis can fully reflect the effect of your operation. And through it, you can also detect the problem, so as to find a solution to the problem in time.

Earlier we talked about products, users, and channels. Each element has its corresponding data indicators for effect feedback. If we talk about concepts, I believe that you don’t understand it very well. Let’s use a case to illustrate. For example, your boss gives you an information APP project. Before he gives you the task, the APP is still in the idea stage. What should you do at this time? At this time, you need to use data analysis to help you complete this task:

Before making this APP, we must first understand what our business goals are, that is, what is the purpose of making this product? This is easy to understand. We need to establish a large user base to compete for venture capital financing and related value-added services (referring to advertising services). Once we have this goal, we can proceed to the next step. What is the next step? Of course, it is competitive analysis and market research. Through these two aspects, we can obtain the APP-related needs, including user groups, interests and hobbies, terminal devices, content direction and other demand positioning. Then we can proceed to the next step of user experience layout and prototype design; the rest is the technical development work. After a period of time, this information APP will be launched soon.

The preliminary work may not have much to do with data analysis, but the goals we set are related to data analysis to a certain extent, because the goals are the direction we optimize and improve through data analysis. Of course, after going online, we will go through various tests and fix bugs before promoting and advertising it in major app stores to ensure that the user experience of this APP on users' mobile phones is the best. After a period of operation and promotion, we will extract the relevant data, provided that the accuracy of these data is very high. Let's extract relevant data from this APP for analysis:

1. Product data items:

Core indicators:

• Product scale: including download volume, number of registered and activated users, and average daily active users
• Market operation: including the proportion of active users, main sources of users, and retention rate
• Commercial effect: average daily turnover, conversion rate of value-added users, amount of value-added services, etc.

Derivative indicators:

• Browsing direction: average page views, average browsing time, number of launches, and access frequency
• Registration direction: daily number of APP downloads and opens, daily number of new registrations, registration conversion rate
• Retention direction: usage retention, purchase retention
• Interaction direction: number of users commenting daily, number of interactive feedback (collection, sharing, likes, etc.)

2. Channel data items:

• Consumption data: consumption, impressions, clicks, average click price, average ranking
• Traffic data: number of visits, number of visitors, number of IP addresses
• Conversion data: conversion rate, revenue, ROI

3. User data items:

• User experience data: bounce rate, visit rate, dwell time, visit depth
• Visitor attributes: gender, occupation, education, age, region, device used, operating system​

After we get the above three data, of course the work here is what the data analysis specialist needs to do, and statistics need to be done every day, and the accuracy of the data must be guaranteed.

Next, let’s talk about several methods of analyzing data. I have 5 years of experience in promotion and operation, and the two most commonly used methods are chart comparison analysis and attribution analysis.

Chart comparison analysis, this method is to first generate charts for batch or time period data. There are many types of charts here, including pie charts, bar charts, curve charts, etc., which can be based on the needs of the data demander. Although the graphics are different, they all reflect the same problem, which is the core point of attention.

So how do you compare? Comparison does not mean you have to make a verbal comparison, but the commonly used comparisons are month-on-month and year-on-year. Of course, if the data is more sensitive, you may not need to use graphs to see the problem, but for the sake of intuitiveness and easy understanding, the generation of charts is very necessary.

The data items for comparison are not all those listed above, but the core items are taken, so that it does not appear too complicated; after analysis, the problems found are listed one by one in words, and at the same time, your own opinions and suggestions for solving the problems are attached. Finally, they are sent to the demander via email. The whole process tries to reflect the professional aspect of data analysis as much as possible.​

Attribution analysis refers to analyzing causes through results. It may be easier to understand with an example here. If the registration conversion rate of the app you are responsible for is 0.5% in a certain week, while it is 1.2% under normal circumstances, a decrease of 0.7%, how should you analyze it? We first need to look at how the registration conversion rate is calculated, that is, the number of registered users/number of downloads.

Analysis idea: The registration conversion rate is directly proportional to the number of downloads when the number of downloads remains unchanged, and is inversely proportional to the number of registered users when the number of registered users remains unchanged. There are two situations here: either the number of registered users decreases while the number of downloads does not change much; or the number of registrations does not change much but the number of downloads increases. In this case, it is definitely because some details were not done well when the visitor was experiencing the APP, which led to the bounce.

When the number of registered users decreases, first check whether there is a problem with the APP registration process. Secondly, check whether the number of people opening the APP has decreased. Whether the decrease is due to a drop in ASO ranking or an increase in competitors, then make corresponding adjustments. After eliminating these problems, observe the data changes. Also at the end, list the problems found in the form of a document and communicate it to the demander by email.

Well, that’s all for data analysis. If you have better ideas and suggestions, please leave me a message and share them. I wish you a good harvest~

APP Top Promotion (www.opp2.com) is the top mobile APP promotion platform in China, focusing on mobile APP promotion operation methods, experience and skills, channel ASO optimization ranking, and sharing APP marketing information. Welcome to follow the official WeChat public account: appganhuo

[Scan the top APP promotion WeChat QR code to get more dry goods and explosive materials]

This article is compiled and published by (APP Top Promotion). Reprinting this article must be approved by Top Promotion , and please attach the link to this article!