For sellers, in addition to analyzing user portraits through indirect data, they can also build user portraits through direct order report data, so as to optimize links and advertisements. You can find the order report in the store order-order report.
Select the corresponding date and click Download to get an order report containing product SKU, price, order time, delivery address and other information. By analyzing this data, you can get a more detailed user portrait.
1. Regional distribution
Summarize all order information of the store in the past year, select the two columns of data of ship-state and quantity-purchased for analysis, and analyze the regional positioning of the purchasing customers. By making a visual diagram through Excel, you can analyze more intuitively.
For all products in the store, you can analyze them one by one and make horizontal comparisons. Once the influence of operations is eliminated, the shopping preferences of users in different locations can be confirmed, which will help with the later selection and listing of products.
2. Time distribution
Because orders come from traffic, traffic comes from exposure, and exposure comes from device use (App or PC use), the peak order period is the active time of the buyer’s own product audience, and the trough order period is the inactive time of the product audience.
Filter the order time purhase-date and product quantity quantity-purchased columns in the order report for analysis to get the number of orders generated each hour. Set the total orders for one day to 100%, and then set its specific value/active number according to the number of orders per hour. Assuming the total orders are 1,000, and the number of orders at 11:00-12:00 noon is 100, then the active value for this period is 10%.
In this store, users’ active ordering time is mainly concentrated between 8:00 and 11:00 a.m. Pacific Time. Converted to Beijing time, it is 0:00 to 3:00 in the morning. For operational activities, the start time of flash sales and promotions can be set 1 to 3 hours before the time period for easy preheating. Ads also need to ensure that there is enough budget during this period to improve exposure.
However, for user portrait analysis, more factors need to be considered. Since the continental United States spans 4 time zones, the time in the order report is Pacific Time UTC-8, which corresponds to the local time of California (CA). However, Texas (TX) belongs to Central Time UTC-6, which is 2 hours earlier than Pacific Time; Florida (FL) and New York (NY) belong to Eastern Time UTC-5, which is 3 hours earlier than Pacific Time. Since some states cross different time zones, the order time can be converted into local time according to the location of their capitals, so as to more accurately analyze the user’s purchasing behavior.
After converting to local time, it can be clearly seen that the peak time for users to place orders is between 10:00 and 12:00 local time, and there is a small peak at 15:00, 18:00 and 23:00 in the afternoon. It is not difficult to see that these time periods are the concentrated rest time for the working class, and 15:00 in the afternoon is afternoon tea time. Based on this, it can be inferred that users who browse and order during these time periods have a casual attitude towards shopping. In addition to cost-effectiveness, they also pay attention to the overall shopping experience. Optimization from the perspectives of product images, graphic layout, video dubbing, etc. will have a higher possibility of promoting conversion.
Similar to the analysis method of order regional distribution, the order location of all store products can be analyzed uniformly. By comparing links with similar prices and styles, and combining the order regional distribution analysis, you can more accurately locate user behavior, thereby helping yourself to get optimization inspiration.