Contrast is the basic method to identify things-horizontal, vertical and multi-dimensional contrast, the logic of logical indicators behind the ratio ratio and the power of management indicators on the level and dimension of the target, benchmarking management and example.

The most common comparison is the comparison of scale and quantity, such as the comparison of sales, the comparison of number of people and the comparison of time. Using different comparison indicators will lead to different conclusions.

We call the choice of contrast standard the perspective of contrast, and different perspectives will draw different conclusions. For example, if we compare Xiao Qiang and Xiaoming from the perspective of height, we can judge their height. We can also compare it from other angles such as academic performance and age. When comparing people, we may have a more comprehensive dimension. For example, when comparing customers, various factors will be considered comprehensively; We also have various models when comparing the causes of various changes. Contrast happens all the time. All we have to do is to find a suitable perspective of contrast and find different opinions on the same problem.

After comparing the variables at the first level, comprehensive variables can be formed. A variable obtained by processing first-order variables (variables that directly describe things, such as length, quantity, quota, width, height, etc.). ) is called a secondary variable. When comparing secondary variables, the commonly used indicators are growth rate, efficiency and benefit.

Growth rate refers to the ratio of quantity change within a certain time range. Two companies, two products, two markets, two customers and two channels can compare the growth rate. The comparison period can be set by month, quarter and year. In 20 15, the GDP of the United States was $65,438 +07.4 trillion, and that of China was $65,438 +00.4 trillion. The annual economic growth rate of the United States is 2.4%, and that of China is 6.9%. This is the speed comparison, as shown in the figure below. Naturally, we will ask, if this growth rate continues, how long will it take for China's GDP to surpass the United States? 2027!

Efficiency is the ratio of input and output, and it is an evaluation index of resource utilization ability. Efficiency comparison is to see who can produce more value with less input. Comparing the efficiency of the two enterprises, we can see which enterprise has more development potential and competitiveness.

The commonly used indicators to measure the efficiency of human resources are the per capita output value (the average output value of everyone in a certain period) and the yuan equivalent output value (the output value brought by the company's salary per yuan), as shown in the following figure. The former evaluates the number of people as an input factor, while the latter evaluates the salary of personnel as an input factor. If a company's sales and profits have achieved rapid and stable growth, but in the process of company growth, the efficiency of material factors is improving, while the efficiency of labor factors is declining, that is, the per capita output value is declining, which means that with the development of the company, the average level of new recruits is declining, while wages are increasing, and the output of each yuan of wages is greatly reduced.

Some seemingly thriving companies have a huge human resource crisis behind them, and the overall level of human resources in the company is declining. It is difficult for such a company to develop continuously, especially in the increasingly fierce market competition, and people's ability can't keep up with the development of the company. If the company can't see this problem in the process of development, then this problem will become a very key problem in the future-when the company develops rapidly, all problems are not problems, but when the company's development is blocked, small problems will become big problems. The problems faced by many companies in times of crisis stem from prosperity. Therefore, it is very important to track some efficiency indicators.

For relatively complex things, the analysis of a certain dimension can only represent one side, but not the whole picture of things. And if there are too many dimensions of comparison, we often can't get a clear answer.

For example, Xiao Ming is taller than Xiao Qiang 10 cm, but Xiao Qiang is more handsome than Xiao Ming; Some customers don't have strong purchasing power, but they can drive friends to buy the company's products and services. This requires comprehensive consideration of various factors, and it also brings new challenges to data analysis-we need to find a simpler way to evaluate things.

The so-called index is a meaningful evaluation system obtained by weighting and synthesizing various evaluation standards. For example, the consumer price index is an indicator to measure price changes. The price of vegetables is rising, but the price of rice is falling, and the price of meat, poultry, eggs and milk is also falling. It cannot be said that the overall price is rising. So how to evaluate the fluctuation of prices? Price fluctuation can be measured by comprehensive indicators such as CPI (Consumer Price Index) or PPI (Producer Price Index).

Consumer price index (CPI) is the combination of most people's consumption habits, and the comprehensive fluctuation of consumer price is calculated according to the proportion of consumer products. The composition of CPI in different periods is also different, and different countries will combine different consumer price indices according to different consumption habits of consumers. Because China is a vast country, consumers' consumption habits are very different, so CPI will bring you different feelings. The essence of CPI is an index used to measure the change of cash purchasing power owned by consumers in a specific period. PPI refers to the comprehensive purchasing power of industrial producers.

In enterprise management, some comprehensive evaluation indexes are also used for comparison. For example, the most typical one is KPI(KeyPerformanceIndicator), which is a comprehensive scoring index formed by setting different weights according to the company's requirements for a certain position and the different importance of each dimension. The KPI settings of different companies and different positions will definitely be different. In order to make the performance indicators better serve the company's strategy, a well-known consulting company once put forward a KPI indicator called BSC(BalanceScoreCard), which comprehensively considered the financial indicators, customer indicators, growth indicators and process indicators of each position. The BSC of different companies and different positions is definitely different, but it basically covers the comprehensive weighted average of four types of indicators.

A key function of data analysts is to design "indicators" for comparison. There is a big difference between design indicators and application indicators, and many people will make mistakes when applying other people's indicators. If we really want to design indicators, we need to have a deep understanding of the logical relationship between things.

For example, I work for a company that deals in infant food. The products include formula milk powder, rice flour, fruit paste and safe and nutritious infant food supplements. Their business is directly related to the birth rate of a market, the overall purchasing power of this market, consumers' concept of eating packaged food for infants, and the government's control over food safety in this market. After analyzing so many relationships, they hope to build an index to reflect the attractiveness of a market to companies, so that companies can invest according to this market attractiveness index.

The attractiveness of the market is also related to the intensity of market competition. If there are many competitors in this market, the top competitors are strong and the background capital is abundant, then the attraction of this market is small; If this market is almost blank, then this market is attractive.

Considering the above factors, we need to establish a comprehensive index to evaluate the attractiveness of this market, and it is best to get a score directly for intuitive judgment. For example, a market with 80 points is 5 points less attractive than a market with 75 points, and a market with 60 points is twice as attractive as a market with 30 points. So how to design this indicator? We need to weigh all kinds of data.

Without considering the market size, we can first establish an exponential model:

y = ax 1+bX2+cX3+dX4+eX5+fX6+…

These include:

Y: market attractiveness index value. X 1: infant birth rate (or the number of babies born each year). X2: market purchasing power parity index. X3: Consumer's attitude towards baby packaged food. X4: The influence of corporate reputation on consumers' purchase of infant products. X5: Government control of baby food.

A, b, c, d, e, f… are coefficients, representing the degree of influence.

We can build an addition model. The additive model indicates that there is no mutual influence among the factors, and each factor has an independent influence on the market attractiveness.

Of course, we can also build a multiplication model:

Y=aX 1×X2×X3×X4×X5×X6×…

This multiplication model assumes that every element is interactive. For example, if consumer confidence is insufficient, purchasing power will be greatly reduced.

We can trace all kinds of historical data, put the data of different stages together, form multiple data combination equations, and realize the model construction through approximate solutions. Finally, the mathematical model for calculating the market attractiveness index is obtained:

Y market attraction = f (x 1, x2, x3, x4, x5, x6, ...) This mathematical model can be used to guide the company's future market investment practice, and it can also be continuously verified in the company's continuous market expansion process, constantly improving various assumptions, indicators, coefficients and calculation methods, and finally forming an expansion model suitable for the company's own development process.

This is the first time that this book mentions the construction of mathematical model, which requires readers to have a mathematical foundation. If the above content makes you feel difficult to understand, then you can skip this part completely; If you are a professional in data analysis, then this part should not be too difficult for you; If you are not a professional in data analysis, you only need to know these contents, you don't need to delve into them, and you don't need to build your own mathematical model. If you are a senior manager of the company, reading this section can help you better understand the process of generating a data model, so as to understand the daily work of data analysts.

The following is a demonstration with CPI. CPI is essentially a comprehensive data index formed after building a mathematical model.

Suppose a resident eats 5 kg of pork, 3 kg of chicken, 2 kg of beef, 1 kg of mutton, 0.5 kg of duck, 0.5 kg of goose, 5 kg of white flour, 5 kg of cucumber every month ... After a lot of statistical research, we have come to the conclusion that the diet structure of the whole country is the above factors.

We investigated the prices of all these products in all vegetable markets (in fact, sample representatives) and got the consumption expenditure of this resident this month, assuming that it is 1000 yuan; Next month, he will also buy these products to meet his daily needs, but the prices of various products have changed. He recalculated his consumption expenditure with the new product price of the month, and the result was 1050 yuan. Then, compared with last month, the consumer price has increased this month (1050- 1000)/6544. If last month's CPI is determined to be 100 yuan, this month's CPI is 105 yuan, and the CPI rises by 5%. This is an image description of the consumer price index CPI. Of course, the actual CPI calculation will be more complicated than this, because we monitor more products than I listed, and there are also many channels and monitoring points to obtain product prices.

Indicators play an important role in the whole economic field, and some indicators directly represent the economic vane and even affect the economic development. Indicators representing an economy, an economic entity or a company include credit rating indicators, price indicators include CPI and PPI, and short-term economic development indicators include PMI (Purchasing Managers Index). ...

When we compare and analyze the data, we should not only simply compare the data directly, but also construct some indicators that can be used or evaluated repeatedly in a certain department, a certain business field and a certain situation. These indicators can be the comprehensive analysis results of multiple data or the collection of business indicators. Data analysts should model various indicators according to business needs, form long-term observation data sets, and verify the rationality of such indicators. Only indicators that have been tested by long-term practice can become comparative indicators for enterprises to continue to use. The longer a comprehensive index enterprise takes, the more perfect it will be, which can better reflect the characteristics of company management.

When the management indicators of enterprises are gradually enriched, you will find that the management culture and management system of enterprises are undergoing subtle changes. In the past, the responsibility of managers was to make decisions according to the production and operation conditions and ensure the implementation of decisions, form feedback before, during and after implementation, and constantly adjust the implementation process of decisions.

When data is more responsible for this process of analysis and decision-making, the responsibilities of managers gradually shift from "thinking" to "directing", and the requirements for managers' intelligence and experience become weaker. The number of managers with the same ability is gradually changing, the scope of management is increasing, some complex management work is gradually exerted by data and data indicators, some analysis and judgment work is completed by intelligent systems, and enterprise organizations are gradually turning to flattening and community.

The full text is taken from Zhao Xingfeng's Business Data Analysis-Ideas, Methods, Applications and Tools.

Contents of the last issue:

A common analysis idea in data analysis-comparative analysis (1),

Contrast is the basic method to identify things.

② Contrast-horizontal, vertical and multidimensional contrast

Next content preview:

Common Analysis Ideas in Data Analysis —— Comparative Analysis (Ⅲ)

① the level and size of the target

② Benchmarking management and the power of example.