In our actual business, some products have just business email list been put on the shelves and have only been sold for a few days, so not all products have sales in the past 30 days. According to the original logic, the "average" should be used for those with less sales, but according to The logic of the technique, using the "median".
I have seen a very interesting analogy, saying that the code of the program is like "the wires of the village in the city". Although it looks messy, it can ensure that every household has electricity, and there is no problem in using it, but if one day someone moves If a wire is disconnected, the entire urban village circuit will be paralyzed (no ridicule, please don't blame the programmer).
So saying that the system is online and running normally does not mean that our product work has been completed. We still need to follow up and be responsible for the full life cycle management of the product. We must be alert to any small problems and review whether there are loopholes in the system design. .
3. Maintain a "skeptical" attitude towards system data
To maintain a "skeptical" attitude towards system data means that we can't just look at the data, but must carry out a reasonable analysis based on the data background. The data of the system is the most objective, and after a reasonable analysis, it can help us to make a lot of improvements.
The following report is the data of an e-commerce company I simulated (for reference only):
Figure 1
Looking at Figure 1, if we just look at this data and trend chart, the first conclusion we can get is that the company's refund rate is declining, and the customer service department has made a lot of efforts.