First, I am worried about the high income, how high it is, how big the risk is, how big it is, and how to prevent it. So the report needs to write the project income and project risk.
Second, does the cash flow of investment projects occur in the past, present or future? If it is the future, you need to speculate. Therefore, the report needs to have forecast content.
Third, do investment projects exist in isolation, or can many companies do this project? Therefore, the report needs to include the contents of the project company.
Fourth, when will the project be invested and what is the external environment for investment? Therefore, the report should include environmental analysis.
Customer Satisfaction Index System-Scorer Index
Suppose you want to do user preference analysis for a color TV company, what aspects will you investigate?
Because user preferences are reflected in user behavior, the content of the survey depends on how you describe user behavior. Specifically, it is divided into two small problems:
Question 1: What are the specific behaviors of users in the process of purchasing and using products? -& gt； Corresponding time clue
Question 2: How would you describe user behavior? -& gt； Corresponding structural clues
Time clues can be developed by using the five-stage theory of user behavior.
In terms of structural clues, we can use 5W2H to analyze user behavior.
Combining the two, the preference analysis of color TV users is obtained.
1, standard deduction
2. Ordinary deduction
The main form of common deduction is 4W mode:
1, who do you choose?
According to the two dimensions, the attraction and relative competitiveness of the object are divided into three zones (9 quadrants): high, medium and low, with the attraction of the other party as the ordinate and the competitiveness of the other party as the abscissa.
2. Why did you choose him?
3. The track of love development
4. How to win love
● Essential attribute (M for short): the core attribute of a product or service. With this attribute, users can only avoid dissatisfaction.
● One-dimensional attribute (O for short): an attribute that is linearly and positively related to the user's attitude. If you have this attribute, users will be satisfied; If this property is not available, the user is not satisfied.
● Charm attribute (A for short): the attribute expected by users. With this attribute, users will be satisfied, and without this attribute, users will not be dissatisfied.
● dispensable attribute (I for short): users don't care if they have this attribute, which is redundant.
● Aversion attribute (R for short): having this attribute will lead to user dissatisfaction.
The relevant perspective discusses a certain relationship between things, which may be causal or related.
1, scale forecast
Find the factors that affect the target variable, and then establish a regression model to predict the value of the target variable.
Douglas production function: regression model, in which enterprises predict the output scale according to the input of technology, capital and labor.
2. Precision marketing
In order to carry out accurate marketing, it is necessary to understand the relationship between user characteristics and user preferences and judge whether there is correlation between them, such as gender and color preferences. If gender has an influence on color preference, customers' color preference can be judged according to gender.
If the p value is less than 5%, a small probability event actually occurs, then the hypothesis of SSR=0 (that is, there is no difference in color preferences between different sexes) is invalid, and this original hypothesis is rejected. Therefore, different colors can be recommended to users of different genders.
The "target vehicle" is the vehicle for which insurance is purchased. When the "target car" changes more than three times in half a year (going out of danger, changing cards, changing people), it is necessary to pay attention, which is likely to be auto insurance fraud.
Background: Airline passenger satisfaction analysis project. The airline made an analysis with consulting company A, but felt that the analysis of company A was not deep enough, so it found consulting company B, hoping to have more value output.
The analysis and evaluation of Company A provided eight indicators (broadcasting, attitude, catering, safety ...), while Company B added eight indicators (price, image, entertainment ...) through focus group discussion, totaling 16 indicators.
So many indicators, where to start? There is a logical clue called importance, and KANO model is made for 16 indicator.
Rank all indicators according to importance and satisfaction:
The analysis idea is the decomposition process from the research purpose to the research content, and it is the refinement of the demand. Therefore, strategic analysis needs to consider two issues:
The purpose of strategic analysis is to help enterprises make strategic choices and choose their own target markets. There are two questions to answer:
Market environment refers to the current situation of the market, and the main indicators to measure the market environment are: market scale, profit level, growth rate, growth potential and life cycle.
Enterprises certainly hope to find a market with large market scale, high profit level, fast growth rate and strong enough growth, which is in the growth stage. However, such a market is often difficult to satisfy, and it is impossible to perform perfectly in every index, so enterprises need to make choices according to their own resources and positioning.
In SWOT, opportunities and threats are used to judge the attractiveness of the market, and strengths and weaknesses are used to judge the competitiveness of enterprises. Therefore:
Take advantage data as an example to calculate:
Background: In recent years, the competition in the color TV market has become increasingly fierce. Foreign brands dominated by Japan and South Korea have occupied more than 60% in premium market and primary market, and occupied high-end market. The competition of domestic brands is more intense, and the price war is getting worse, squeezing the meager profit space.
The cost advantage of domestic color TV brand A is not obvious. In order to get rid of the price war and enhance the core competitiveness, we must start with users and carry out differentiated marketing according to the preferences of different users. Therefore, it is necessary to analyze the preferences of color TV users.
In order to understand this differentiated marketing, we need to answer two questions:
● Why do enterprises need differentiated marketing? -& gt； Because there are differences between users' preferences and needs, identifying these differences and carrying out accurate marketing can get twice the result with half the effort.
● Why can user preference analysis support differentiated marketing of enterprises?
Based on temporal thinking and structural thinking, user preferences can be divided into five stages and seven elements:
1. Demand generation stage: user demand is different (why). Work, study, play and give gifts?
2. Information collection stage: users have different collection channels (where). Acquaintances, Tik Tok, Zhihu, advertisements?
3. Scheme comparison stage: the key purchase factors (what) of users are different. Price, face value, thinness, brand image?
4. Purchase decision-making stage: There are differences in how users make decisions. All-round consideration and decision-making speed ...
5. post-purchase behavior stage: There are differences in user usage occasions (when/where). Wear sneakers when playing, shoes when running, shoes at work and slippers at home.
Besides Q 1~Q 18, there are some basic information.
leave out
Analytical framework:
Analysis method:
Among them, the percentage is the proportion of this data to the total (generally counting); Effective percentage, excluding the proportion of filtering factors such as missing values.
Where n represents the sample size.
According to the test results of variance analysis, it can be concluded that there are significant differences in the following aspects when various users buy and use color TV products (see Table 5-7), so the comparison of preferences of various users should be analyzed around these aspects:
Analysis of variance+numerical problem = comparison mean range.
It can be seen that the single group is the most sensitive to the promotion (price sensitive), and the old and small groups are the least sensitive to the promotion (guessing that because of the strong demand, more consideration is given to whether to meet the demand).
Analysis of variance+classification problem = the range of cross analysis
Take gender, consideration factors (burning function, power consumption, internet access function, others' recommendation, promotion activities) and specifications as examples to interpret (that is, turn the results of spss into more understandable charts with conclusions attached).
Background: The main business of insurance company A is auto insurance. In recent years, the competition in the auto insurance market has become increasingly fierce. In order to win in the fierce competition, insurance company A has decided to take precision marketing as its development strategy, and plans to carry out customized services according to the needs of auto insurance target customers. Therefore, it is necessary to conduct classified investigation and analysis on auto insurance customers.
In the specific project, the selected dimension should distinguish the differences of customers, so that the differences between various customer categories are large and the differences within categories are small. This needs to be tested by analysis of variance.
Classification dimension: life status+premium amount (corresponding to Q 13 and Q6 of the questionnaire respectively).
Factor analysis: eliminate correlation and reduce dimension. The operation is as follows:
The premise of factor analysis is that the original dimension is related, and the applicability test judges whether the original dimension is related.
The values in the above table are called factor load, which indicates the degree to which the factor explains the dimension (i.e. 9 sentences) information.
As can be seen from the above table, the factor 1 explains 54. 1% of the information in this dimension, while the factor 2 explains 49.0% of the information in this dimension, and the value of 54. 1% is close to the value of 49.0%, indicating the factor 1. Similarly, the table also shows that both factor 1 and factor 2 have the characteristics of "enjoying a quiet life alone" and "going home as soon as possible after work" ......
As mentioned above, the purpose of factor analysis is to eliminate the correlation and make each factor have the characteristics of differentiation. However, the current component matrix has not achieved the established effect, and the rotation factor is needed.
Iterative clustering (such as kmeans) is only applicable to continuous variables, and factor categories are classified variables, so hierarchical clustering is used.
Considering that the number of people in each category should be uniform, the preliminary judgment is divided into five categories.
Analysis of variance: analysis → comparison average → one-way ANOVA.
Factor classes are classified variables, so cross-analysis is used to describe the differences of living conditions of each class: analysis → descriptive statistics → cross-table.
The premium amount is a numerical variable, so the comparison average is used to describe the difference of consumption grades of various categories: analysis → comparison average → average.
According to the cross table (category) and average value (premium) above, name five types of customers:
How to select target customers? There are two issues to consider:
● How attractive are customers? Is it worth doing?
What is the competitiveness of enterprises? Can you do it?
The indicators that describe customer attraction include customer scale, growth rate, profit space, life cycle and so on.
Indicators describing the competitiveness of enterprises include market share, brand reputation and resource strength.
In this case, after discussion, the company decided to use customer size+premium amount to measure the attractiveness to customers, and use market share to measure the company's competitiveness among various users.
Frequency statistics: analysis->; Descriptive statistics->; frequency
Customer size (effective percentage in the above figure)+premium amount (name in the previous category) is used to measure customer attractiveness.
Because insurance companies insist on low-price strategy and pay more attention to customer scale, the weight of customer scale is 60%, and the weight of premium amount is 40%.
Because of the different dimensions, they need to be standardized:
Cross analysis: analysis-> Descriptive statistics->; crosstab
As can be seen from the figure below, the company's target customers are mid-range extroverted.