There is a big gap between China and the United States in data analysis, and data analysis can only be highly valued by some particularly large domestic enterprises, such as BAT. Of course, this is the result of their long-term accumulation, and the combination of data and operation is better. This is my overall feeling after returning to China. Domestic enterprises' understanding of the data itself and the value it can provide is not as profound as that of the United States, and the difference is quite large.
Question 1: What kind of companies need to pay attention to data? What's the difference between different stages?
Generally speaking, the companies that pay more attention to data in China are all companies with high customer unit price and heavy transformation, such as internet finance, e-commerce, trading platform, SaaS, online travel and so on. This kind of customers have high unit price and incomplete traffic, so entrepreneurs have the motivation to improve their transformation.
Macroscopically, entrepreneurs will go through four life cycle stages of products.
The first stage is called cold start. At this time, the company was very early, and the angel round was still the A round, and even the financing was not successful. For companies at this stage, it is a false proposition to drive with big data, because the number of customers is limited and the sample is insufficient. They need to know more about the needs of potential customers and ask them to use this product.
The second stage, the initial stage of growth. The cold start is about to be completed. Experienced entrepreneurs will begin to lay out some core indicators related to growth, such as daily/monthly activity and retention. The purpose of these indicators is not to measure the current performance of products, but to provide a comparable benchmark for future growth.
The third stage is the growth period. At this stage, we can see the great difference between a good startup and an ordinary startup. Both public relations and activities require manpower and time costs. How to find the most efficient channel in growth? I think this is the core competitiveness of PK between startups. If you don't do data-driven, you can do it once or twice by intuition, but no one can win 10 thousand times in a row. Therefore, intuition needs to combine data, so that enterprises can quickly optimize various channels and improve the conversion efficiency per unit time.
The fourth stage, liquidation period. The realization of business needs a high user base. General Internet products, a small number of highly active and experienced users, will be converted into paying users. Similar to a funnel, it is constantly screened, which is to strive for the efficiency of operation. For example, the conversion funnel of e-commerce users is generally: visit the registration search, browse and join the shopping cart to pay, or return it in the future. This is a very, very long funnel. In order to do a good job in data operation, every link of the funnel should be continuously tracked.
A good enterprise, especially one that wants to generate income in the future, must pay attention to the transformation efficiency of various departments and links. This conversion efficiency can be achieved by means of marketing, product improvement and even customer operation. Moreover, each link is slightly improved, and together it is a multiple improvement. It's hard to understand how big this multiplication will be without digital operation.
Question 2: What should a good data analysis look like?
Good data analysis can benefit everyone in the company. It is not a privilege, not just the privilege of one or two people in the company, but all the operating departments of the company, especially those fighting in the front line, directly benefit.
Generally speaking, it is not enough to talk about strategy, general direction, CEO and VP or operation. It needs to be given to employees working in the front line for them to use. I think this is a big difference between data-driven enterprises and non-data-driven enterprises. Efficiency is improved by everyone, not one or two people.
To establish a complete data analysis mechanism, a company must first start with business. All data analysis operations or data systems should start with business and customers. This data analysis system should not only solve one or two narrow problems, but also need a system and a big picture. Then, in fact, the most difficult part of data analysis is data collection and data collation, which is the most time-consuming, probably because the initial plan is not comprehensive enough. Therefore, we should pay great attention to collecting and sorting out information in a planned way.
Later data analysis can't just stay on the basis of reports, and its value is still not enough. Finally, it is correct and effective to tell others what to do after those figures come out. In that case, you have deep knowledge and need strong operational ability.
Therefore, an enterprise should not only have the overall situation, but also pay attention to enforceability. I suggest that if a general enterprise wants to build itself, it should first make a breakthrough from a single point, find the transformation point, see the value, and learn the next practice method through this practice. This is also a learning process. Don't start with a huge system, but start with 50 data circles, and want to build a data science framework. I think if you generally do this, unless you have a lot of resources, you will certainly fail.
Question 3: What stages can enterprise data analysis be divided into?
The first stage, nothing;
In the second stage, the company needs to go back to history: it is the most basic and primitive stage to know what is happening in its products;
In the third stage, people who do products, operations and marketing internally need to ask why: this stage is prediction, that is, predicting what a group of people will do next, so as to develop products better and pertinently;
The fourth stage is to have a solution: that is, I predict that this group of people will do this, so I will give it a better solution, make it better transformed and retained, and bring better new effects;
The fifth stage is optimization. How to find the best balance point for diverse product lines? There is a balance between price, marketing, product design and sales. This balance point is the point where the interests of entrepreneurs are maximized, and it is also the point where users like this product best.
These five stages need to take time to accumulate. Don't jump, jumping will often fail, starting from the basics.
Question 4: Why do data analysis of many companies become a mere formality?
This is mainly due to the lack of cognition of many enterprises in three aspects: the value of data, the methodology of data analysis and the practical operation methods.
1) value cognition
Many companies are in a period of crazy growth, and the decisions made by everyone may have produced a lot of value; In this case, it is difficult for them to realize that data decision-making can produce greater value than violent growth.
2) Understanding of basic methodology
It means a core but simple methodology. At present, there is not much understanding of basic methodology in China, which may be because the domestic development time is still relatively short, while the United States has developed for decades.
3) Cognition of practical operation methods
Domestic front-line employees use data to guide their work operations, such as products, customers, sales and other practical experience is relatively small. On the one hand, it is because of the short development time, and on the other hand, the concept of data usage has accumulated less.
However, domestic enterprises have been rapidly raising this awareness. But this cognition is a gradual process. In the United States, cognition and methodology have gradually been well unified between technology and business, and data are used to fuse.
Many domestic entrepreneurs didn't realize the value of data at first; When he realizes the value of data, his expectations are often high. This huge gap can't really reduce the value, and even make people question whether this value can really be realized and lack patience.
Question 5: What are the myths about data in China?
I think the understanding of data analysis by domestic companies is divided into two poles: one is pure technology, and the other is superstition. I believe that as long as the data is large, it will become a tall company. I think there are some misunderstandings on both sides.
At the core, I think this thing you made is valuable and effective. It is the most direct to measure with the effect. Other companies want to build their own platforms and form large teams, with low efficiency and output. I suggest you be careful. With the continuous development of the ecosystem, many tools are now very easy to use, so you should learn to use them. These are some good auxiliary means for entrepreneurs to succeed. You can't say that you can succeed in starting a business just because you can use tools. But a good entrepreneur can certainly use these tools to achieve his goals.
Question 6: How to break the strange circle that data can't make the best use of it?
In the past few months, we have dealt with customers and found that some enterprises use our products very well, while others are just so-so. Usually, an enterprise has a core responsible for data, which will be used well; Some enterprises don't have a core person to pursue this matter, so they do it in general.
Therefore, at least one person in the operation department must have a certain concept of data analysis. It's like moving an advanced surgical instrument to the company, even if no one can operate it.
I think the best way to acquire knowledge is practical operation. The premise of practical operation is that it is best to have someone who knows a little and can do it with him several times. Then turn around and learn, which is the fastest and most effective way to acquire data analysis knowledge. I don't think pure reading or reading some textbooks and reading some big data guidance books outside can have this effect.
With this person, we can get methodological support from people and company products who understand this aspect, and this learning mechanism is established. This is very important, otherwise, no matter how powerful the system is, no one will operate it and make the best use of it.