Second, we are unable to trace the data from the next day because Quantum clears out the previous day's data once it hits midnight. This means that if a customer sees the product on the first day through a search and then directly enters the website on the second day, we will be unable to trace the user's real source of origin. Taobao online store promotion should be based on data, rather than going by gut feelings. If we follow our gut feelings, we can only go further off track, but if we follow the data, we will have a broader path ahead. How do we analyze Taobao store data? Here, I would like to recommend an analytical tool called Quantum Hengdao Statistics. Let's talk about Quantum. Quantum, probably developed by a team aiming to create microscopic data at the quantum level. From the perspective of the store, Quantum is currently the best data statistics tool available. I remember during a communication meeting at the Shanghai Crown Club, when the Crown store asked Ni Hao from Quantum's team the question "Why does some of Quantum's data sometimes not match with other tools' data?", he confidently replied: "This is normal because Quantum is the most accurate." This left a deep impression on me; and when I consulted Bingyi and Ni Hao from Quantum on Wangwang, they basically answered all my questions, which moved me greatly. How to open an online store; recently, Quantum has opened up the sub-account login function, showing that the Quantum team is indeed continuously striving towards the path of quantization. Many friends have already clearly explained Quantum's functions, with screenshots as well, so I won't elaborate much here.
We specialize in producing collagen, and currently have three products. Although this is quite a number in the collagen industry, compared to other industries, the product count is still relatively low. We were once eliminated in the second round of applying for Taobao's Cloud Brand due to the issue of product quantity, which was quite regrettable. The small number of products, high cost, and high per-customer spending make us particularly focus on data mining. Because of the limited number of products, we cannot afford to operate loosely, thinking that if one market doesn't work out, another might; because of the high cost, every penny spent must consider the return on investment; because of the high per-customer spending, losing any customer means a significant loss. We have always desired such data analysis that could directly tell us the source of each transaction, so we could quickly adjust our advertising strategy accordingly. However, we haven't found this yet. We even thought whether Taobao could open up user visit record interfaces, allowing us to develop our own system, but we haven't found the entry point yet, and even if we did, development would require manpower and resources. Quantum Statistics provides real-time visit data reports, so we use this report combined with Excel to create basic data reports that can analyze promotional effects. I'd like to share our reports with you, and please correct any parts that are not used well.
Step One: First, let's familiarize ourselves with the various columns in the real-time customer visit report. Serial number, recording sequence number; visit time, self-explanatory; store entry source, this column shows the source of visits. The Quantum team lists common sources such as Direct Express, Diamond Booths, etc., while less common ones are grouped under 'other sources'. What we really want to use isn't what we see directly, but the specific address of the source that appears when the mouse hovers over it. For example, when we hover the mouse over the 'other sources' position in the first record, we can see the specific URL address of the source in the browser's address bar or prompt box, which is very important. Through these, we can know where the source is; visited page, we can see the current location the user is visiting; visitor location, self-explanatory, the visitor's address, where they come from; customer tracking, this column is very important and can be used for tracing; repeat customers, can be used to determine if the user is visiting for the first time.
Step Two: When a customer places an order, we can immediately obtain the customer's order time and location. It's best to check the chat records between the user and customer service to see the start time of the chat. Between the order time and the start of the chat, we take the earlier time, which theoretically is closer to the time the customer visited our store. With the time and location, we can trace back in the real-time customer visits to locate the access record of that customer just now, find their first visit, see their store entry source, and then record it.
Thirdly, Quantum only retains 40 pages, i.e., 1000 records of data for the day. If the store has a large volume of visitors, tracing may become problematic unless the data is regularly saved in Excel or imported into a database.
Step Three: Based on the recorded data, we can create horizontal and vertical analysis reports of the effectiveness of sources, which is very helpful for our advertisement placement. We can intuitively see the effectiveness of each source and the trend chart of each source's effectiveness. I won’t go into detail here since Excel is quite powerful; as long as you can think of it, you can uncover meaningful data.
This method is quite helpful, but it is only suitable for temporary use because it has significant limitations. It is more suitable for stores in the initial stage or stores with daily visits not in enormous quantities and transactions not too frequent.
Firstly, it requires relatively high tracing skills from the data analyst;
Fourthly, we found that data may be lost, causing tracing difficulties. For instance, if a customer generates several page views (PV) around 16:00 without purchasing, then revisits around 20:00 in the evening, generating several PVs and placing an order, when we trace back, we find that the PV records generated during the 16:00 timeframe are already missing, even though the total number of visits (PV) at that time had not exceeded 1000.
The most ideal situation would be for the Quantum team to provide an effect analysis report. I believe the Quantum team will eventually develop it. But as for us, we can't afford to wait. I'm sure all buyers are eagerly looking forward to it. When the analysis tools aren't up to par, being able to lead competitors in operational details gives us an earlier chance for self-breakthrough and a higher chance of success.
To run a store well, we need both macroscopic and microscopic reports. The process of gradually digging from the macroscopic aspect to the microscopic aspect is the process of a store or enterprise becoming increasingly refined in its management. Undoubtedly, Quantum Statistics is the sharp tool that helps us refine our operations. Thank you, Quantum Statistics, and hope the Quantum Statistics team becomes stronger, with greater breakthroughs in the expressiveness, accuracy, and development speed of the reports.