Among the many data of Internet companies, business data and financial data are often relatively easy to apply, but website visit data is often the least used. What is the reason for this, and how should website data be used?
Let’s first talk about the reasons why there are so few applications:
1. First, understand the importance of data.
Transaction data and financial data are often related to the life and death of a company. If there is a problem with this data, the company's operation will be tense. Therefore, everyone from the CEO and CFO to the specific employees of the company are concerned about this data.
Website data is only related to the quality of the company's products. At the stage of company development, this is not obvious, especially in China. Many companies can rely on first-class sales and third-rate products to seize the market. A website doesn't even need an excellent product. It can raise a large amount of money just by relying on a lot of investment, advertising, attracting users, and generating business data. Therefore, the data of the website itself has been used to a certain extent. After being ignored, companies that can invest heavily in products rely not on the data of the website, but on the strength of the product personnel themselves.
2. The second is the cost of data processing.
Each business in Business Data Pass is a separate data, so the data processing is much simpler and the amount of data is much smaller than the website data. At the same time, in terms of ensuring data accuracy, since business data is related to financial issues, there are multiple servers and multiple systems to ensure its accuracy.
Website data is often a sequence of data composed of users, independent sessions, and visited pages. These matrix data increase the difficulty in processing. In terms of data volume alone, it is often dozens of times the transaction volume. The data volume is The increase is not the fundamental difficulty. The difficulty is from analyzing a piece of data to analyzing a sequence of data. At the same time, another difficulty is that with such a huge amount of data, it is difficult to ensure that data is not lost. In fact, most website records It is normal to lose 1%-5% of the data, and even 1% will have a considerable impact on the analysis of the sequence.
3. Website data is more difficult to understand.
Since trading/finance exists in traditional industries, it has relatively clear definitions. Most people can understand the meaning of data such as transaction volume and transaction value (except financial data); for professionals, these data do not exist. any ambiguity. Moreover, in the process of obtaining data, the acquisition methods and statistical methods are relatively standardized, and doubts rarely arise during this period.
Website data only developed with the development of the Internet. There is no clear definition of data. Often in two statistical systems, the same nouns represent different meanings; different nouns represent the same meaning; the same nouns represent different Meaning; the name and meaning are the same, but the way of obtaining them is different; and there are few professionals in this area of analysis. Relevant personnel, such as product managers, interaction designers, even have the number of independent users, independent IPs, and independent sessions. But no one can figure out the traffic number. At the same time, in the process of data acquisition, due to differences in corresponding development languages, jump methods, and server configurations, the methods of obtaining data are not uniform, which adds an additional layer of difficulty to understanding.
4. It is difficult to apply website data.
Traditional data is often easy to apply in practice. The comparison of input and output makes many judgments easier to make decisions. And the business situation is often a situation of choosing 1 from 2 or choosing 1 from N. When the data proves that one is better than the other, it is easier to make a decision. At the same time, due to long-term accumulation, transaction data has a formed data interpretation structure. When you see XX data, you know that this data is affected by XX data.
It is often difficult to make decisions based on website data, and it is often difficult to directly apply the data to actual modifications. Except for A/B test, which is a comparison of plans, most of the other data is not a comparison of plans, but the optimization of the same product. If the conversion rate of a certain page is low, the product manager cannot directly abandon this page, but still needs to optimize it, which makes the judgment and processing of data much more difficult. At the same time, the data on the website is affected by both the system and human consciousness, and there is no established structure to explain the impact of related data on the data. Moreover, once the judgment result is that the user does not continue to operate, the data will fall into a trap. Arrived in the wasteland.
The above are the disadvantages of website data compared to business data, but this does not mean that website data is meaningless, but it has not yet come into play. When the territory is demarcated and each company has to work hard to optimize business data, the website data The effect is also reflected. (Text/Lance’s record book)