Case Study (Background to AusPaper)
AusPaper, a subsidery of Pinnon Paper Industries, is an Australian company with a long history in local manufacturing of paper products. In 2013 only, AusPaper produced 619,000 tonnes of paper products, and sold more than 690,000 tonnes of products to local and overseas markets. They export their products to over 75 countries in Asia, USA, Europe, Middle East, the Indian subcontinent, Latin America and Africa.
AusPaper sells paper products to two market segments: the newspaper industy (e.g. Australian Finacial Review, Herald Sun etc.) and the magazine industry (e.g. Mens’ Style Magazine, Homes and Gardens etc.). Also these products are sold to these market segments either directly to the customer or indirectly through a broker.
Despite their successful operations and solid financial turn-overs over the last two decade, AusPaper is forecasting a major shift in business climate within the next seven years. This is a result of a change in end-consumers preferences (i.e., readers’ preference to access newspapers, magazines online or via e- readers, and social media). Now more than ever, AusPaper management feels the need to ensure a strong customer base and ideally a strong strategic alliance with their clients in newspaper and magazine industry. In addition, they are planning to put in place a formal procedure to be able to project future financial turnovers using historical data.
Consequently, AusPaper has approached ANALYTICS7 (a Market Research Company) and asked them to conduct a large-scale survey of their clients to better understand the characteristics of AusPaper customers, their perceptions of the company, and the likelihood of customers building long-term strategic alliance with AusPaper.
Data Collection Process (Conducted by ANALYTICS7)
To address AusPaper concerns, ANALYTICS7 has contacted purchasing managers of firms buying from Auspaper and encouraged them to participate in an online survey. The collected data are then supplemented by other information compiled and stored in AusPaper’s data warehouse and accessible through its decision support system.
The primary databse consists of 200 observations on 18 separate variables. Two types of information are accessible in this database. The first type of information is perceptions of AusPaper’s performance on 13 attributes. Purchasing managers of firms buying from AusPaper were asked to rate the company on each of these 13 attributes using a 0 – 10 scale, with 10 being “Excellent” and 0 is being “Poor”. The second type of information relates to purchase outcomes and business relationships (e.g., satisfaction with AusPaper and whether the purchasing firm would consider strategic alliance / partnership with AusPaper). A third type of information is available from AusPaper’s data warehouse and includes information such as size of customer and length of purchase relationship, as well as quarterly turnover of AusPaper operations. A complete listing of variables, their definitions, and an explanation of their coding are provided in file.
Your Role as an ANALYTICS7 Data Analyst Intern
You are a master of business analytics student doing an interenship at ANALYTICS7. The research team manager (Hugo Barra, with PhD in Data Science and a Master Degree in Digital Marketing) has asked you to lead the data analysis process for AusPaper project and directly report the results to him. You and Hugo just finished a meeting wherein he briefed you on key purposes of AusPaper research project.
Hugo explained that an important prerequisite in building a strategic alliance in a B2B environment is “customer satisfaction” with a firm’s operations. Therefore, the first goal is to identify key factors that predict customer satisfaction with past purchases from AusPaper. He is also interested in gaining deeper insights into factors that predict the “likelihood of AusPaper customers building strategic alliance” with the firm. The final analytics goal is constructing a forecasting model to predict AusPaper’s turnover in the upcoming three quarters of 2017. From these understandings, Hugo and consequently AusPaper will be in a good position to develop plans for the next financial year.
In addition to briefing you about key research questions, Hugo also allocated relevant research tasks and explained his expectations from your analysis. Minutes of this meeting are available on the next page.
Now, your job is to review and complete the allocated tasks as per this document.
Providing an overall summary of two outcome (dependent) variables of interest: One variable captures the extent of satisfaction with previous purchases from AusPaper. The other one indicates whether a client perceives his/her firm would engage in strategic alliance or partnership with AusPaper.
Identifying 15 factors (from AusPaper_Data) that may influence customer satisfaction with
An appropriate statistical technique could be used here to identify a list of predictors that could be included in the regression model.
Building and finalising a model (Model building process) to predict customer satisfaction with
Hugo has done a separate analysis and found that the depth and breadth of AusPaper ‘product line’ is a significant predictor of ‘customer satisfaction’. In line with his findings, prior research shows that the strength of this relationship may vary according to ‘customer location’. That is, customers from global markets have needs that are more diverse compared to those from a specific region such as ANz.
Thus, the relationship between ‘product line’ diversification and ‘customer satisfaction’ should be more prominent for customers in global markets (i.e. outside ANZ region). Your task here is to test Hugo’s assumption by modelling an interaction effect between the above-mentioned independent variables and ‘customer satisfaction’.
Building and finalising a model (through a model building process) to predict the “likelihood of customers developing strategic alliance/partnership” with AusPaper.Hugo has already done an initial analysis for this task. Based on his analysis, Hugo has narrowed down the key predictors of “customers likelihood to build a strategic alliance/partnership” to the following:
Product Quality, Product Line, Personnel Image, Price Flexibility, and Competitive Pricing
Your job now is to continue his work by finalising a predictive model of key factors that influence the “likelihood of building a strategic alliance/partnership”.Hugo would like to gain a deeper understanding of customers’ likelihood for building strategic alliance/partnership with AusPaper. He is specifically interested in understanding the probability of having strategic alliance with AusPaper for customers who meet the following criteria:
a. Feel neutral (i.e. score of 5 on the relevant scales) towards AusPapers’ image and its product
b. Varying levels of perception towards product quality (i.e., scores from 1 to 10) and price flexibility (scores of 0, 5, and 10).
Hugo believes that ‘price flexibility’ and ‘product quality’ would define AusPaper’s success in building strategic alliance with its customers. Therefore, it is important for AusPaper to know how flexible its sales representatives should be in negotiating prices on the one hand, and how much effort should be put in improving perceptions of product quality in order to increase the probability of building strategic alliance.
Accordingly, your job is to visualise the predicted probability of building strategic alliance with AusPaper for customers with varying levels of ‘product quality’, and ‘product flexibility’ perceptions and fixed perception towards ‘product line’ diversity and personnel ‘image’.
Developing a time-series model to forecast AusPaper financial turnover in the next 3
It is your job to decide which time-series model is most appropriate in this scenario.
Produce a written report detailing ALL aspects of your analysis. Your report should be as detailed as possible and should describe ALL key outputs of your analysis.
The “Aus Paper” is actually a subsidiary of the “Pinnon Paper Industries”. The Australian Company “Pinnon Paper Industries” possesses a long history in the local production. The industry is seen to produce paper only. The subsidiary sells products to two major industries including the two industries namely the newspaper and magazine industry. “Herald Sun” and “Australian Financial Review” are the two industries which receive the paper products. The magazine industries that get the paper products are “Homes Gardens” and “Men’s style magazine”. Paper products are sold to the customers either directly or indirectly.
Despite the fact that the subsidiary has achieved immense success over the last few centuries the organisation is eying a shift in the coming years with respect to the business environment. The management is of the opinion that the need for ensuring a stable customer base and solid strategic alliance with the consumers of the industries is essential.
Data Collection Process:
In this report the contracted firm managers are reported and the gathered data is given along with certain amount of assembled information. The data is tabulated with respect to the sales of the AusPaper warehouse. The sales are manageable by means of the decision support system.
There are 200 perceptions along with the 18 factors which are included in the data file. It is seen that majorly two types of data are included in the required database. The first is concerned with the performance of the organisation and this is on the 13 characteristics which are measured with the help of the 0-10 scale. This is used to imply that 0 is poor and 10 is excellent. The other information is linked to include certain outcomes and business connections. This can include the number of consumers and the length of purchase association and also the quarterly turnover operations of the “AusPaper”.
“The analysis with respect to the assistance of the considered factors is concerned with the clarification of the consumer loyalty with the firm operations. The information is aimed at discovering the major factors helpful for the estimation of satisfaction of the customers. “AusPaper” customers building strategic associationwith the firm. The report analytically focuses to develop a predictive model to anticipate the turnover of AusPaper in the 2nd, 3rd and 4th quarters of 2017. For the forthcoming financial year 2017, the analysis would demonstrate whether “AusPaper” would be in a decent position or not. The researcher clarified his objectives from the analysis to summarise the major research questions.”
Findings of the data analysis:
Summary of Customer satisfaction:
- “The average customer satisfaction with past purchases from “AusPaper “is found to be 6.95”.
- “The standard deviation indicates the spread of the distribution of customer satisfaction with purchases from “AusPaper” that provides the value 1.24”.
- “The median of the satisfaction data set is 7.05”.
- “The measure of location states that 25% of the bottom values are less than 6 and 25% of the top values are more than 7.9”.
- “The mode is 5.4 that indicates that the frequency is found highest at the time customer satisfaction rate 5.4”.
- “The minimum customer satisfaction of previous purchases from “AusPaper” is 4.7 and maximum customer satisfaction of previous purchases from “AusPaper” is 9.9”.
- “Hence, the customer satisfaction with previous purchases from “AusPaper” ranges between 5.2”.
- “The customer satisfaction level of slightly rightly and positively skewed. The graphical visualization indicates that its left tail is longer than its right tail”.
Summary of Strategic alliance:
- “The frequency distribution and frequency table of the variable “Extent to which the customer/respondent perceives hisor her firm would engage in strategic alliance/partnership with AusPaper” determines that out of 200 samples, 114 samples incurred thestrategic alliance or strategic partnership”.
- “On the other hand, 86 people conveyed that they are engaged in strategic alliance or relationship. The percentage share of these two cases are 57% and 43% respectively”.
There are fifteen samples which are chosen for the purpose of analysis. The tabulated samples are attached in the last section namely the appendix at the end of the assignment. “These chosen variables are – Cstmr_Type, Strat_Alliance, Prdct_Qual, E_Comm, Tch_Supp, Cmplnt_Supp, Advert, Prdct_Line, Image, Pricing, Warranty, New_Prdct, Billing, Price_Flex and Delvry_Flex.“Sats” is assumed to be a dependent variable. All the variables except“Sats” are assumed to be independent variables.” The analyst in this case is interested in building a suitable predictive model with a single dependent and several independent variables.
It needs to be noted that in this specific analysis the value of R-square is not authenticated. This implies that there is a chance for it to contain a certain amount of multi-co linearity. Regression analysis is a powerful statistical tool which helps in the examination of the relationship between two or more variables which are to b understood and analysed. This analysis is important for providing detailed into the method of research.
It is known that the method of finding the association among the variables along with the goodness of fit comes under the adjusted R – square calculations and interpretation. It is clearly seen that there are certain variables which are significant out of all the variables as the values of the values of the variables are greater than 0.5. These variables are Cstmr_type, Strat_alliance, Prdct_quality, Cmplnt_res, prdct_line, billing and delivry_speed.
It is known that negative association is used to indicate that with the decrease in the values of the independent variables the values of the dependent variables increase and vice versa.
The analyst in the next part of the analysiscaused that the depth and breadth of “Product line” of “AusPaper” is a significant estimator of the variable “Customer Satisfaction”. The previous analysis referred that the strength of this association may vary according to the location of customers.
Among 200 consumers, 81 customers are from in Australia and New Zealand. 119 consumers are from outside Australia and New Zealand. Three multiple regression models are executed with the help of four variables that are product line, region, customer satisfaction and interaction effect of region and product line.
“The very first multiple linear regression model is build assuming“Product line” as predictor variable and “Customer satisfaction”as dependent variable. The analyst for the regression calculation transformed the variable “Region” as binary variable. Here, 0 = Outside ANZ and 1 = ANZ region. The second multiple regression model is concerned with two “Region” and “Product line” and there is one dependent variable, “Customer satisfaction’. The third multiple regression model has considered the interaction variable of “Region” and “Productline” as a new predictor variable. The interaction variable is calculated multiplying two predictor variables of model 2 “. The predictor variables are “Region”, “Product line” and “Interaction effect”. The dependent variable is as usual as “Customer satisfaction”. There is a help in the representation of the variables with the help of the multiple regression models.
“In the first predictive model product line is linearly, positively and significantly associated with customer satisfaction with co-efficient 0.608915 andsignificant p-value = 0.0.In the second predictive model, product line and Region both are linearly, positively and significantly related with customer satisfaction. Product line whose co-efficient is 0.713734 and significant p-value is 0.0 and Region whose co-efficient (-0.54615) andsignificant p-value is 0.0005. In the third multiple regression model, all the three variables Product line, Region and Interaction are linearly, positively and significantly associated with “Customer satisfaction” where product line has co-efficient is 0.865218 and p-value is 0.0. Here,“Region” has co-efficient 2.927282 andsignificant p-value is 0.0003 and “Interaction” has co-efficient has(-0.55504) and significant p-value is 0.0.Thus, among the three variables ‘Product line’, ‘Region’ and the ‘Interaction effect’, ‘Product line’ and ‘Region’ have linear significant influence on the dependent variable ‘Customer satisfaction’.”
“The first regression model which is simple regression model has maximum“AIC” value and least p-value among three models. Therefore, the model is best fitted.In this case, the effect of “Region” and“Interaction” effect is not present.”
The most important analysis is regarding the understanding of a predictive model which is advanced and this utilizes major variables which impact the chance of impacting a strategic relation to the partnership of the “AusPaper”. There are five factors which are taken into consideration in the advanced model building analysis namely “Product Quality”, “Product Line”, “Personnel Image”, “Flexibility” and “Competitive Pricing”.
“Personnel Image” and “Product Line” are the predictor variable and “Strategic Alliance” is dependent variable in the predicted logistic regression model. The predictive regression model could be stated as-
Prob. (Strategic Alliance)
The p-values of two predictor variables in this logistic regression model are “Product Line” and “Personnel Image” are both 0.0. The p-value is much less in comparison to the level of significance. Both the variables are significant. The level of significance in this connection is 5%.
“Product Quality” and “Price Flexibility” are predictor variables and “Strategic Alliance” is dependent variable in the predictive regression model. It is given as-
Prob. (Strategic Alliance)
The calculated p-values of two predictor variables in the logistic regression model “Product Quality” and “Price Flexibility” are 0.0 in both cases. Both the predictor variables are significant at 5% level of significance. It is seen that the calculated values are less than the level of significance.
The method of forecasting is used for making the predictions of the future on the basis of past and present data by method of trend analysis. This is suitable in statistical analysis and helps in the forecasting of data on the basis of the provided information.
To predict the future turnover amount, the analyst considered “Time” as explanatory variable and “Turnover($’000)” as response variable. The variable “Turnover ($’000)” indicates for total turnover amount within the time span. The variable “Time” isactually the chronological frequency of quarters starting from 1st quarter of 2008 to 1st quarter of 2017.
It is seen that the estimated turnover amounts of the 2nd, 3rd and 4th quarters of 2017 are $4531.955, $4991.969 and $5043.73.
It is seen that the customers of the subsidiary have a significantly high level of satisfaction. Certain variables namely the quality of the product, the region, the effect of interaction, the price flexibility have a noteworthy influence on the dependent variable which is that of the strategic alliance. The discussion is clearly indicative of the fact that there is a predicted possibility of the creation of a strategic alliance with differing levels of the product quality and product flexibility. The continuous proportion to the line of the product and the image of the people are also asserted. Thus, it can be stated that, “quarterly turnover amounts of 2nd, 3rd and 4th quarters are greater in 2017 than any other quartiles of previous year”. Similar to the 1st quarter, the turnover amounts of other quarters also have grown in a significant manner.