The importance of data has been recognized and accepted by more and more companies and individuals, and even has gone too far. The concept of big data is flying all over the sky. It seems that everyone is talking about big data overnight. They meet and don't need big data to say hello. It seems that they are not doing very well in the data circle. So, from what aspects can the data spread by the outside world promote the business to take off? Or in other words, what level of data does the business need? In what ways can data help enterprises?
Combined with the author's years of work experience and understanding of data and business, the business demand for data can be summarized into four levels.
Layer 1? Knowing this, we can know what happened and how much happened by establishing a data monitoring system, so as to achieve "know fairly well".
Specifically, the angle of cutting into data mainly includes these aspects. The first is "observing the sky" to observe the overall trend of the industry and the impact of the policy environment; Then "know the place" to understand the performance of competitors; Finally, there is "introspection". How are you doing and how are your data performing? Looking at the data cycle, "looking at the sky" can be a quarter or longer; "Know the place" is weekly or monthly, except for special time points and special events; The data in Introspection is the most comprehensive, and it needs to be read every day. Some people watch it specially and some people study it specially.
At this level, two data views are shared:
1. The data is scattered and needs a framework to look at the data.
How to treat data is very particular, and it is difficult for piecemeal data to play its real value. Only by putting the data into an effective framework can the overall value be exerted. An effective framework has at least two functions:
(1) There are a lot of data, and different people have different needs for data. For example, CEO, middle-level managers and bottom-level employees usually pay attention to different data, and an effective framework allows different people to get what they need.
(2) An effective framework can quickly locate the problem. For example, the trading volume index that everyone cares about. If the trading volume index drops by 20% on a certain day, it is likely that there is something wrong with the business, but where is the problem? If there are only a few highly abstract indicators, such as conversion rate, number of transactions, customer unit price, etc. It is impossible to locate the problem. A good framework can support us to drill down, find out problems from categories and flow channels, and let the board hit the specific person in charge. This is what we usually say, depending on the data.
2. Data, only comparison can tell the truth.
I have 120 kg. Do you think it is heavy or light? A single data is difficult to explain the problem. To judge the growth rate of an index, it is necessary to choose the correct comparison object and reference system, that is, the baseline. This baseline can be a preset target, the average level of the same industry, or the historical data of the same period.
Second floor? It was not until I knew why I saw the problem through the data that I knew that it was not enough to be so far away. Data is just a representation, which is used to find and describe problems, and more importantly, to solve problems in practice. Combine data with business, find the real reason behind the data appearance and solve it. The process of solving problems will involve data and data processing, and may also involve methods or tools such as data models, which are relatively high in technology and will not be introduced here.
There are also two points to share on the second floor:
1. The data is objective, but the interpretation of the data may have a strong subjective consciousness.
Data itself is objective, but people who consume data have subjective initiative. People often bring subjective factors into the interpretation of data: the same data may have a good conclusion in A's view, but it may have the opposite result in B's view. It's not such a bad situation. The more you argue, the clearer the truth is. However, if we don't look for problems through data, we should first identify the problems and then use data selectively to prove our views, which is not advisable. But in fact, this kind of thing often happens around us.
2. Understand the business to truly understand the data.
Teacher Che Pinjue's blog post "Don't talk about data if you don't understand business" has profoundly expounded this view, so I won't say much here. Just because of the importance of this view, the author deliberately took it out to emphasize it.
Third floor? Using data to find opportunities can help enterprises find opportunities. For example, Taobao has a middle-aged and elderly clothing market and a plus-size women's clothing market. These markets can know that some middle-aged and elderly people or fat MM around us have not met their needs through their perception of the surrounding environment. So is there any other channel to find more market segments?
The data is ok!
By comparing the keywords searched by users with the actual transaction data, it is found that many demands are not well met, reflecting the strong demand but insufficient supply. If you find such a market segment, will it be announced to the second generation of the industry and to the sellers, and will it help you better serve consumers? This example is the "potential market segment discovery" project we are doing now.
Telling this case is not to brag about how powerful the data is, but to tell everyone that the data is there, some people turn a blind eye, but some people can dig out the "baby" from it. What is the difference? Business sense. Many people can see the search data and transaction data just mentioned, but no one has linked these two data before, which reflects the business feeling behind it.
The fourth layer establishes a data operating system. The data operation I understand contains two meanings: data as indirect productivity and direct productivity.
1. Data as indirect productivity.
The so-called indirect productivity means that data workers pass data value to consumers through operations, which is commonly called decision support. The output report and analysis report of data workers are for the reference of business decision makers at all levels. I call it decision support 1.0 mode. However, with the development of business and the enhancement of business personnel's awareness of the importance of data, the demand for data will spring up like mushrooms after rain, which obviously cannot be met by just a few analysts. It is better to teach people to fish than to teach people to fish, so that students in operation and products can analyze the data. This is the decision support 2.0 model in my mind.
The decision support 2.0 model has three key words: product, ability and willingness.
For operators and PD, it is unrealistic and unnecessary to master retrieval languages such as SQL and analysis work such as SAS and SPSS. Providing data products with low threshold and good user experience is the basis for realizing the decision support 2.0 model. The products mentioned here are not only a set of operational functions, but also need to carry analytical ideas and practical cases.
However, the threshold of data analysis always exists. This puts forward new basic ability requirements for operation and PD, namely, basic mathematics ability, logical thinking ability and learning ability.
The final will, perhaps the most crucial, can only be done well if there is a strong internal motivation.
2. Data as direct productivity.
The so-called direct productivity means that data workers directly act on consumers through front-end products. Say fashionable, call data realization. With the advent of the era of big data, the management of the company has paid more and more attention to this. The era of big data has brought great opportunities, but it may also be a catastrophe. If you can't use data to generate value, it's a disaster-the more data you generate, the more storage space you have, and the more resources you waste.
An application that is easy to understand now is related recommendation. After you buy a product, I recommend a product that you are most likely to buy again. Personalization is a new wave of data as direct productivity, which is getting closer and closer. Data workers, get ready.
The above are the four levels of business data requirements shared by Bian Xiao. For more information, you can pay attention to Global Ivy and share more dry goods.