How to do user behavior path analysis

How to do user behavior path analysis

User behavior path analysis is a unique data analysis method in Internet industry. Based on the user's click behavior log, it mainly analyzes the user's circulation rules and characteristics in each module of an App or a website, and mines the user's access or click patterns, so as to achieve some specific business purposes, such as improving the arrival rate of App core modules, extracting the mainstream path of specific user groups, depicting browsing characteristics, optimizing and modifying the design of App products, etc.

This paper will briefly discuss the user behavior path analysis method, and introduce some business scenarios and technical means of path analysis, which will play a role in attracting jade. Welcome friends who are committed to Internet data analysis to criticize. If there is an opportunity in the future, we can continue to introduce more user behavior path analysis cases combined with actual business.

First, the path analysis of enterprise scenarios

An important ultimate goal of user behavior path analysis is to optimize and improve the conversion rate of key modules, so that users can easily reach the core modules according to the expected mainstream path of product design. There are also the following application scenarios in the analysis process:

User typical path identification and user feature analysis are often used for demographic data such as gender and region or operational data such as order price and order number. User's access path data opens another door for us to understand user characteristics. For example, for an application that makes, uploads and shares pictures, we can divide it into creative users who are willing to make and upload, interactive users who are willing to comment, diving users who silently browse and look at pictures, and consumer users who never upload pictures but download them through their apps.

The path analysis of product design optimization and improvement is very helpful to the optimization and improvement of product design, which can be used to monitor and optimize the conversion rate of each module in the expected user path, and can also find some obscure function points. In a video creation sharing App application, users often carry out a series of editing operations from the beginning of shooting and making videos to the final release of videos; Through path analysis, we can clearly see which editing tools users are familiar with and love, and which operations are too complicated, which can help us improve the editing operation module and optimize the user experience. If the number of users' creations in the process of path analysis is closely related to users' behaviors of being praised, commented and shared, we can consider enhancing the sociality of this App and enhancing users' stickiness and creative desire.

3. Monitoring of product operation process

The conversion rate of key product modules is itself a very important product operation index. The corresponding operation results are monitored and verified by path analysis, which is convenient for relevant personnel to understand the operation effect.

Second, the path analysis data collection

Internet industry has a unique advantage in obtaining data, and the data that path analysis relies on is mainly the log data in the server. Every step of the user's use of the App can be recorded. At this time, we need to pay attention to the excellent distribution strategy, which should be closely related to the business we care about. Zhuge io, a refined operational analysis tool based on user insight, can be recommended here. By integrating Zhuge io's SDK into an App or website, all user behavior data in the application can be obtained. In fact, Zhuge io believes that not all events have the same value in every App. Based on the in-depth analysis of core events, Zhuge io recommends that you use the hierarchical customized event distribution method. Each event consists of three levels: event, key and value. At the same time, Zhuge io also provides data monitoring and distribution consulting services for developers, and can provide personalized event distribution consulting and technical support for customers based on rich industry experience.

Third, the relationship between funnel model and path analysis

The path analysis mentioned above is similar to the well-known funnel model. Broadly speaking, the funnel model can be regarded as a special case of path analysis, aiming at specific modules and event nodes of a few people.

Funnel model is usually a description of the conversion rate of a series of key nodes in a website or App, which is often specified by us. For example, we can see the funnel transformation of the purchase behavior of a shopping App application in Zhuge io. On this shopping App platform, buyers have successfully experienced four key nodes from browsing to payment, namely browsing, adding shopping cart, settlement and payment. From step 1 to step 4, fewer and fewer people experience the key nodes, and the node conversion rate presents a funnel-shaped situation. We can monitor and manage the conversion efficiency, operation effect and process of each link, and make targeted in-depth analysis and improvement for the links with low conversion rate. Other funnel model analysis scenarios can be used flexibly according to business requirements. Zhuge io platform has a very powerful funnel analysis tool, which is a platform for you to give full play to your imagination of data. Please see the analysis case based on the funnel model, Funnel/Keep the new gameplay.

Path analysis is different from funnel model. It usually tracks and records every behavior path of each user, and then analyzes and mines the behavior characteristics of user paths, involving the source and destination of each step and the conversion rate of each step. It can be said that the funnel model artificially and actively sets several key event node paths in advance, and path analysis is an exploratory way to explore the overall behavior path, find out the mainstream path of users, and even find some interesting pattern paths that are unknown in advance. From a technical point of view, the funnel model is simple and intuitive to calculate and display the relevant conversion rate, and the path analysis will involve some broader aspects.

Fourth, the general ideas and methods of path analysis

1, exploration of naive ergodic statistics and visual analysis

By analyzing the distributed user behavior path data, the clickstream data of each user event path can be counted in the simplest and most direct way, and presented intuitively by data visualization. D3.js is one of the most popular data visualization libraries at present, and we can use Sunburst partition to describe the click state of user group event paths. Starting from the center of the diagram, it is extrapolated layer by layer, representing the whole behavior statistics of users from the beginning to the end of using the product; The rising sun event path map can quickly locate the mainstream usage path of users. By extracting the path data between specific people or specific modules, and analyzing the path map of the rising sun event, we can locate the deeper problems. Flexible use of the statistical chart of the rising sun path is a magic weapon in path analysis.

Zhuge io can not only obtain the distributed data conveniently, but also provide customers with personalized analysis of the rising sun event path map and make customized product analysis reports for customers' products.

2. The method of sequential path mining based on correlation analysis.

When it comes to association rule analysis, the classic case of data mining "beer and diapers" is inevitable. Whether "beer and diapers" is a "fairy tale" fabricated and boasted by a manager of Teradata, this case makes people understand and understand the process of shopping basket analysis (correlation analysis) and its commercial value to some extent. All the goods purchased by each customer in the supermarket at one time are regarded as a shopping basket, and the shopping behavior data stored in the database is analyzed by using the association rule algorithm, that is, shopping basket analysis. It was found that 10% of customers' colleagues bought diapers and beer, and 70% of all customers who bought diapers bought beer at the same time. So the supermarket decided to put beer and diapers together, which significantly increased sales.

Here, we might as well regard all the event points operated by each user every time they use the App as "a series of goods" in the "shopping basket". Unlike the shopping basket mentioned above, all the click behaviors here have a strict sequence of events. We can improve Apriori or FP-Growth algorithm in association rules, so that it can mine frequent user behavior paths with strict order, which is an important idea of user path analysis. We can carefully consider the product business logic embodied in the discovered rule sequence paths, and also compare and analyze the rule sequence paths among different user groups.

Social network analysis (or link analysis) Early search engines mainly found relevant web pages for users based on the similarity between the content of search web pages and users' queries or the index pages in search engines. With the explosive growth of the number of Internet pages in the middle and late 1990s, the early strategies are no longer effective, and it is impossible to give reasonable ranking search results for a large number of similar pages. Today's search engine giants such as Google and Baidu all adopt the search engine algorithm based on link analysis as one of the methods to solve this problem. Web pages are linked together by hyperlinks, just like social networks in Weibo are linked together by paying attention to behaviors. There are famous, authoritative and influential people in social networks, and there are also important or authoritative web pages on the Internet. Providing authoritative web pages in front of search engine results makes the search effect better.

We regard people in social networks as nodes, web pages in the Internet as nodes, and even every module event in our App products as nodes. Nodes are connected with each other in their own ways to form a specific network diagram. The analysis methods based on these network structures will be collectively called social network analysis.

There are some common analysis methods in social network analysis that can be applied to our path analysis, such as node centrality analysis, node influence modeling, community discovery and so on. Through centrality analysis, we can explore which module events are in the center, or connect two types of module events to become the hub, or become the final destination of most module events. Through community discovery, we can explore whether there are some "small circles" in this social network, that is, a small part of behavior paths that users always like to operate, which are independent from most other modules.

The above is what Bian Xiao shared for you about how to do user behavior path analysis. For more information, you can pay attention to Global Ivy and share more dry goods.