How to Measure Customer Satisfaction (CSAT)

The data presented by A1 Hotels customers on the quality of services offered to them was presented in the non-numerical form. To make it applicable for data analysis, Excel IF function was used with the coding =IF(Room Quality="G","2",IF(Room Quality="P","1")) for room quality assessment, =IF(Food Quality="G","2",IF(Food Quality="P","1")) for food quality assessment, and =IF(Service Quality="G","2",IF(Service Quality="P","1")) for service quality. The numerical coding produced a different set of data with number entries where two indicated ‘good' and one showed ‘poor' as per a customer's rating. The following tables indicate the individual frequencies for each variable according to the data collected from the customers. 

Table 1: Descriptive Statistics 

Table 2: Room Quality Frequencies 

Table 3: Food Quality Frequencies 

Table 4: Service Quality Frequencies 

From the above frequency tables, less than 40% of the customers were not satisfied with the quality of the room, food, and services offered by A1 hotels. The percentage is lower than the expected traditional ratio of dissatisfied customers of 40%. The figures indicate that the services, food, and room quality offered by A1 hotels were beyond traditional as per the customers' feedback. From the responses, 23% of the customers were dissatisfied by the quality of the room in the A1 hotels. On the other hand, 24.5% were dissatisfied by the quality of food offered while 27.5% were dissatisfied by the quality of the services provided. 

Confidence Interval Calculations 

Confidence intervals were calculated to determine the lower and upper endpoints for each of the category assessed; room, food, and service. The calculations were done using Excel function with the numerically coded data. The most significant values in the determination of the confidence interval were the mean, sample size, and standard deviation which were all calculated using Excel functions as illustrated in the attached Excel spreadsheet. Since the calculations were done at 92% confidence interval, a confidence coefficient of 1.75 was used in all the three endpoints determinations (Knapp, 2018). The equation used in the determination of the endpoints were; 

 for the upper-end point, and  for the lower end point where  is the sample mean, 1.75 is the confidence coefficient for 92% confidence interval,  is the standard deviation for the sample and  is the sample size. The table below represents the values obtained in the calculations for each of the three variables. 

Table 4: Confidence Interval Calculations 

According to the table above, the error margins obtained for the room, food, and service quality are 0.050964, 0.053354, and 0.055392 where all of them are small values thus giving a guarantee of possibly correct results. For room quality calculations, the upper and lower bounds or endpoints contained the sample mean calculated at 92% confidence interval. Similar cases were recorded by the analysis for food and service quality variables. From the traditional expectation, 40% which is equal to 80 customers from the sample should be dissatisfied by the quality of food, room, and services of A1 hotels. The ratio would constitute to a mean of  as per the coding values used in the present data set. A proposition to test in the dataset is to determine whether the customers' responses meet the traditional expectation or not. The current analysis hypothesized that the recent clients are less dissatisfied than the traditional level of dissatisfaction. 

To test the hypothesis, a critical approach test was conducted at an alpha value of 0.08. The analysis results produced p-values less than 0.0.8 for 92% confidence interval for each of the three variables. The obtained p-values indicate that there is a significant difference in the dissatisfaction level between the recent clients and the traditional dissatisfaction level with the former being less dissatisfied. The findings support the hypothesis as developed by the present analysis. From the hypothesis testing findings, as well as the descriptive analysis results as presented by the tables above, it can be inferred that A1 hotels have overcome the traditional levels of dissatisfaction. However, there is still room for improvement since not all the recent clients are satisfied with the quality of food, rooms, and services offered by the hotels. 

References  

Knapp, G. (2018). An exact confidence interval for a common effect size.  Journal of Statistical Theory and Practice ,  12 (1), 3-11.