This article mainly introduces some commonly used models for data analysis: event analysis, funnel analysis, heat map analysis, retention analysis, event flow analysis, user group analysis, user scrutiny, distribution analysis, and attribution analysis. 1. Event Analysis In the analysis of user behavior data, an event refers to a certain behavior of the user operating the product, that is, what the user does in the product, which is translated into descriptive language as "operation + object". Event types include: browse pages, click on elements, browse elements, modify text boxes, etc. A complete event should contain the following aspects: User Information: Information describing the user. For example, user access or login ID Time information: the time the event occurred Behavioral information: what behavior the user did Behavior object information: on which objects the user's behavior acts.
For example, if button A is clicked, page B is browsed, and text box C is modified, then the distribution of A, B, and C is the object of user behavior. Event analysis is the most basic of all data analysis models, which refers to analysis operations such as statistics, latitude subdivision, and screening of indicators of user behavior events. For example, for the event "click to add to cart button", we can use "clicks" or "number of clicks" to measure, the mobile number list corresponding indicators are "click to add to cart button" and "click to add to cart button" respectively number of people”. Measurement results can be presented in line charts, vertical bar charts, horizontal column charts (bar charts), cousins, numerical values, bubble charts, etc. The line graph of event analysis can be used to observe the trend of continuous change of one or more data indicators, and also can perform year-on-year data analysis with the previous period as required.
Through event analysis, we can accurately understand the amount of events that occur in the product, reasonably configure tracking according to product characteristics, and easily answer questions about changing trends and latitude comparisons, such as: How many clicks are there on the product promotion page of a certain event segment? How much has it improved compared to yesterday? What is the cumulative number of product registrations for a channel? What are the top ten product registration channels in the first quarter? For the UV time-sharing trend of a certain activity page of a product, what is the proportion of Android and iOS? 2. Funnel analysis Funnel analysis is a set of process data analysis model. It measures the conversion effect of each node by using each behavior node starting from user behavior as an analysis model node, which is generally presented by a horizontal bar chart.