For insignificant variables, the matching effect of the ws number list model on this variable can be weakened 2) Quantification of matching results The sample size of the experimental group and the control group is close to 1:1 SMD < 0.1 SMD stands for Standardized Mean Difference One way to calculate SMD is: (experimental group mean – control group mean) / ws number list experimental group standard deviation. The above quantitative indicators meet the rules, indicating that the weighted matching is successful.
3) Quantization gain value and significance check ws number list is_treat = 1.62, indicating that the 30-day gain of users who participated in the activity was ws number list 1.62 compared with those who did not participate in the activity, and the result was significant, and the quantitative ws number list evaluation results were available.
4. Quantify activity gain ROI The commonly used measure is ws number list LTV, and the difference in LTV between the active group and the non-participated group is the LTV gain. The difficulty here is how long it takes to calculate the gain from the event. How big is the gain ws number list and how long will it last?