Why do we need to analyze user behavior?

1. What is user behavior ?

User behavior is composed of the five simplest elements: time, place, person, interaction, and content of interaction.

1. What is user behavior?

To analyze user behavior, it is necessary to define it into various events. For example, user search is an event, which includes the time, platform, ID, search result and search content. This is a complete event and a definition of user behavior; we can define thousands of such events in a website or APP.

With such events, we can connect user behaviors and observe them. When a user enters the website for the first time, he is a new user and may need to register. In this case, the registration behavior is an event. Registration requires filling in personal information, and then he may start searching for things to buy, all of which are user behavior events.

2. How to obtain user behavior data?

So, how do we monitor these user behavior data?

A very traditional and common way is to define the event by writing code. Load a piece of code where the website needs to monitor user behavior data, such as a registration button, an order button, etc. After the monitoring code is loaded, we can know whether the user clicks the registration button and what order the user places.

All these methods of describing events and properties in detail by writing code are collectively referred to as "point embedding" in China. This is a very labor-intensive project, and the process is very tedious and repetitive; however, most Internet companies still employ a large number of tracking teams.

2. Why do we need to do user behavior analysis ?

If it is so troublesome, why do we need to do user behavior analysis?

Because only by doing user behavior analysis can we know the user portrait and the business truth behind the users' various browsing, clicking and purchasing on the website.

Simply put, the main method of analysis is to focus on churn, especially for websites that require conversions . We hope that users will not leave and will not leave after they come here. Like many O2O products, there are a lot of subsidies for users at the beginning; once the money is used up, the users will leave. Such a product or business model is not good. We hope that users can truly find the value of the platform and keep coming back instead of leaving.

User behavior analysis helps analyze how, why, and where users churn.

For example, the simplest search behavior is: when a certain ID searched for keywords , which page and which results it viewed, and at what time the ID placed an order. The entire behavior is very important. If he is not satisfied with the search results, he will definitely search again, changing the keywords to other ones, and then he will be able to find the results.

What else can user behavior analysis do?

Once you have a lot of user behavior data and defined events, you can split the user data into a table by hour, day, user level, or event level. What is this table used for? One is to know the simplest events of users, such as login or purchase, and also to know which are high-quality users and which are customers that are about to churn. Such data can be seen every day or every hour.

3. Five Scenarios of User Behavior Analysis

Now that we have user behavior data, what application scenarios do we have?

• Attracting new users means acquiring new users.
• Conversion, for example, e-commerce pays special attention to order conversion rate .
• Promotion: how to get users to use our products frequently.
• Retention : identify potential churn users in advance and reduce churn rate.
• Monetize, discover high-value users, and improve sales efficiency.

1. Attracting new customers

When I was at eBay in 2008, my job was to analyze the ROI of each keyword in SEM and SEO . eBay buys 4 million keywords from Google every day. In addition to SEM and SEO, we also need to analyze various other partner channels . For example, if a small e-commerce website puts an eBay link on it, and the user finally completes the purchase on eBay through the link, eBay will give the website a share of the money.

eBay pays special attention to which search engine and which keyword brings the traffic ; whether the keyword is paid or free. Search engine terms from Google bring in a lot of traffic, but whether this traffic turns into orders on eBay, this data must be combined with eBay's own data, and then channel allocation is done to determine which channel actually turns out to be the one that made the order. The entire data chain needs to be connected from beginning to end, which requires the integration of data from both sides.

2. Conversion

Taking the registration conversion funnel as an example, the first step is to know which registration entrances are on the web page. Many websites have more than one registration entrance, so each event needs to be defined. We also want to know how many people and what percentage of people clicked the registration button in the next step, how many people opened the verification page, how many people logged in, and how many people completed the entire registration.

There will be user loss at every step during the process. After the funnel is completed, we can intuitively see the loss rate of each link.

(III) Promoting activity

Another is the fluency with which users use the product. We can analyze specific user behaviors, such as visit duration and which page they stay on for a particularly long time, which is particularly obvious on APPs. Another thing is to improve the user portrait. It is more accurate to use user behavior analysis to make user portraits.

For example, in the United States there is a very famous online video network Netflix. Netflix is ​​very interesting. Through user behavior analysis, it can accurately analyze and define your entire family. How many people are there in your family, adults or children, and what are your three favorite movies ? The more behavioral output you have, the more accurate his recommendations will become.

(IV) Retention

User churn does not mean that they will churn all at once. Some subtle and small behaviors can indicate that they will churn in the future.

At LinkedIn, we need to track user usage behavior. For example, whether you have logged in, whether you have searched for your resume after logging in, whether you have uploaded your resume, etc. Every little behavior of the user is important. With these data support, LinkedIn's product and sales departments have to look at user reports every day. The simplest one is to find out whether user usage behavior has declined, which behaviors have declined, which users use the product particularly well, etc., in order to maintain user relationships.

(V) Monetization

LinkedIn is a 2C and 2B company with 400 million users worldwide and a lot of resume information of real users. The 2B business is sold by LinkedIn to every corporate HR, with the aim of helping American companies find mid- to high-end talent. There are many different product lines involved. LinkedIn itself is a social network . Whether the users are managers, VPs, directors, or business people, marketing, sales, etc., all this data is aggregated into a company's dimension on LinkedIn.

With the company's knowledge, we can quickly have sales take this and sell it to customers. For example, when negotiating business with Starbucks , the data that would most shock Starbucks HR is the list of talent turnover rates.

As shown in the above picture, which employees have joined from other companies in the past year and whose last company they came from are shown in blue. The left side shows the employee turnover of Starbucks, and the companies they jumped to are shown in red.

Through this simple distribution, we can quickly see the situation of talent loss. If there are more blue lines, it means that the company is strong in attracting talents. If there are more red lines, it means that the company is in a downturn in talent reserve and recruitment . We show the data to the end customer and basically get the order. We can tell stories through data. We made a lot of reports at the beginning, which sales could use to tell stories and quickly close deals.

All of this is done through user behavior analysis, not through guesswork or third-party data. The value of user behavior analysis is self-evident.

Mobile application product promotion service: APP promotion service Qinggua Media advertising

The author of this article @ GrowingIO compiled and published by (Qinggua Media). Please indicate the author information and source when reprinting!