| Nowadays,we have more access to information,which presents a whole new challenge to the release of their activities.The traditional activity planning model,which relies heavily on the experience of planners,often results in untimely activity releases,large deviations in activity scale estimates and biased activity content.Small-scale activities,for example,“Share and Collect Coupon Campaign”,are becoming a major part of activity planning due to the positioning of daily activities.The short release time of smallscale activities has determined that the traditional inefficient activity release mode has been unable to adapt to the rapid pace of operation.We design and implement an operational growth activity management platform from the perspective of solving the above-mentioned problems.After detailed requirement analysis,we complete the design and development of the activity release module,activity state management module,flow audit module,data prediction module and user management module.By providing the functions of activity management,activity state management,audit management,template management and creating activity with templates,we effectively improve the small-scale activities’ efficiency of planning and releasing.By providing the function of data prediction,we effectively improve the effect of activities.The data prediction module is a module built to meet the specific needs of the online car industry.By building the XGBoost data prediction model,we use the past30 days’ daily order volume data,weather data and holiday data to predict the future 15days’ daily order volume.The obtained prediction data will help operators to determine the activities launch time,activities scale and activities content.At present,the system has been tested and put into operation.The feedback from users using the platform is positive.The time from planning to release for small-scale events has been reduced by an average of 4 days and the results of the events have been significantly improved after release. |