Find the abnormal data (high retention/low retention) through data comparison, and quantify the specific times and time. To use specific data for the indicators, you can also make a executive email list cycle of 7 days or 14, by drawing Line charts and other methods to find obvious inflection points. Find the intersection of retained users and users who have a certain behavior, that is, users who have done a certain behavior are more likely to be retained; When we used to executive email list create a website, we found that users clicked the button for instructions, and more users stayed. Later, we strongly guided users to click the button of "instructions.
And the retention rate of the website gradually increased. After we determined the core indicators, we then found quantitative indicators through comparison, and analyzed three executive email list types of users through data. No key behavior but stayed There are key behaviors but they don't stay had key behavior and stayed Calculate the proportion of various types of users, and then determine specific quantitative indicators by analyzing the relationship between users with key behaviors, the number of behaviors, and executive email list time imes within Suppose we found that the key behavior is "favorites" + "comments" through comparison.
Validate key behaviors through hypothetical experiments We found from the data that the key to affecting retention is "favorites" + "comments" After the initial verification, conduct a executive email list small-scale equal test, and attract more users to complete the corresponding behavior "collection times" + "comment" within a limited time. 6. Guide users through key actions Strategies You Might Think of Make favorites and comments a novice task Visuals that highlight favorites and comments Complete collections and comments, and give some rewards Enhanced executive email list instructions for leading favorites and comments.After passing a small-scale test and verification, it will help to improve the retention of the product.