MKT 450 Exam 4

Generally speaking, what type of attributes are prioritized for improvement opportunities?

attributes with high importance and low performance (this is the first priority), the second prioritized one is low performance and low importance

Write out the the stated improvement scale?

1 No Improvement Needed 2 3 Minor Improvement Needed 4 5 Major Improvement Needed We will rescale this improvement scale in the database to ensure that these numbers will line up with the rest of the data.

What are the objectives of sample analysis?

1) Who is the sample? 2) Does the sample look like the population? -If both of these objectives are met, then you can precede with the analysis. If not, you must reevaluate and conduct another survey with the right population.

For attribute performance scores, how many decimal places do we typically use?

2 decimal places

Explain NPS scores to a 5th grader.

A NPS is a survey scale that is used to see how you are doing compared to everyone else in the industry. It is a percentage from -100 to 100 to see where you are at on the scale and if you need to improve on anything. You are basically answering the question of would you recommend this product/service/company to your best friend.

How would you explain sample analysis to a 5th grader?

A sample analysis is used to see if a smaller group looks similar to a larger group or population. To do this we ask two question: Who is the sample and does the sample look like the population?

Are attribute importance scores inferred or stated? Do we ask attribute importance questions on a CX survey?

Attribute importance scores are inferred. On a CX survey there will be no attribute importance questions, but rather questions that have you rate independent variables on a scale of 1-10 (poor to excellent). From this information, we then figure out how the independent variables (attributes) affect dependent variables (NPS). We do not use a correlation analysis or a multiple regression analysis, but we do use a relative weight analysis.

Explain attribute performance scores to a 5th grader.

Attribute performance scores is comparing your company to their competition. To get these scores you have to get customers to participate in a survey. The survey asks the same satisfaction questions and overall satisfaction questions for both your company and the best competitor. It is important to compare the scores relative to the competitor.

Does the sample look like the population? What does this mean?

Does the small people who were picked look like the whole population that we are looking for.

For NPS, how many decimal places do we typically use?

For NPS we do not use any decimal places because there is too much variation in the data.

If the sample looks like the population, does this mean and what do we do?

If the sample looks like the population, then you can proceed with the rest of the analysis. Is that the target? Is that who were going after? Always make sure the sample looks like the population. Can share findings if these two match.

In RWA analysis, what do we use for the dependent variable(s) and what do we use for the independent variables?

In RWA analysis, we use NPS for the dependent variable(s) and we use attributes for the independent variables. Dependent variables= How likely are you to recommend us to a friend or colleague? Independent variables= taste, location, overall performance, etc.

What do we do if stated and inferred improvement analysis identify different improvement opportunities?

It could happen! The best case scenario is that the inferred and stated improvements will match perfectly, but this is not always the case. We will be more confident if both match. But when they do not, we must take our best judgement on what we should really focus on. There will be like 2-3 keys attributes to really focus on. It is important to remember that a lot of people just say price is a needed improvement because it is easy, so we have to keep an eye on this. Price is a little overstated in stated improvement.

How would determine if an attribute is a major strength, major weakness, minor strength and minor weakness? Are there any other categories?

Major strength is when a company has an advantage, it is is really important to the customer, so it is high performance and high importance. Keep up the good work is another name for this. Minor strength is when the attribute is not as important to the customer, but we still have an advantage, so low importance and high performance. Possible overkill is another name for this. Major weakness is when we know it is really important to the customer and we know we have a disadvantage in the marketplace. Another name for this is concentrate here. This is our first focus for improvement opportunities. Minor weakness is when the attribute is not as important to the customer and we have a performance ratio below .97. This is a Low priority but should still be the second option we go to. The other category is the zone of indifference this is when the importance and performance attributes are tied.

For attribute importance scores, how many decimal places do we typically use?

One decimal place

What are performance ratios and how do we calculate them? How do we run these scores in SPSS? Do we use means, frequencies or valid % for attributes? How do we write the formulas in Excel?

Performance ratios are how we really interpret the attribute scores. We compare our scores relative to the competition. We calculate by dividing (us/them). We take the attribute scores from the survey questions that were answered for our company and our best competitor to figure out how our customers are feeling towards our company and our competitors (This is information is in SPSS). Then, we take these numbers and divide the scores (us/them). We are only looking at the mean scores. We copy and paste the information into excel. We write the formula in excel by (=us/them).

How do we calculate NPS scores? How do we run these scores in SPSS? How do we write the formulas in Excel? Do we use means, frequencies or valid % for NPS?

Promoters are scores from the CX survey that select 9's or 10's as the answer. Passives are 8's and 9's. Detractors are 6 & below. To calculate the NPS we take the valid % of promoters- the valid % of detractors to the total NPS score. To run these scores in SPSS we go to analyze->descriptive statistics->frequencies->select all of the NPS scores that you are looking for->okay Then you copy and paste this data into excel. The only important things that you will need is the valid percents and the score. You do not use decimal places for NPS scores! Formulas: =(7's+8's) , =(9+10), =(6+anything below 6)... to get NPS score =(answer to 9+10- the answer to 6&below). We only use valid percentages

Is RWA included in SPSS? What is RWA-Web and how do I access it?

RWA is not included in SPSS. The RWA-web is a way to select different types of data after you log in and the results will simply be emailed to you. You go to www.relativeimportance.davidson.web Once on the website there will specific sections that need to be filled. Click multiple regression, select correlation or raw data, then use list wise deletion compared to parities deletion, then enter the name of criteria: dependent variables, predictor: all independent variables that were used, then bootstrapping, then finally hit submit.

What does RWA stand for and why do we use this statistical technique for attribute importance analysis?

RWA= Relative Weight Analysis We use this statistical technique for attribute importance analysis because it overcame the problems of multicollinearity and large number of independent variables. It is similar to a multiple regression analysis, it looks at independent variables to predict a dependent variable. The larger the numbers, the more relative the importance is. This analysis looks at raw weights (coefficients=R^2, and if weight contains 0 then it is not statistically significant) and rescaled weights (will add up to 100%, will rescale as importance scores).

Understand and be able to explain CX attribute importance analysis to a 5th grader.

The CX attribute importance analysis is figuring out what's most important to our customers. This is not straight up asking what is most important, but rather figuring out how the different variables affect one another. There are two different types of ways to figure out importance: stated importance or statistically inferred importance. Both tells you what is most and least important, but do so in different ways.

If the sample doesn't look like the population, does this mean and what do we do?

The data is useless because the wrong people are answering our survey questions. Do not waste your time. You cannot proceed , you must fix it. You can go back to different people to take your survey or find new people in your database.

What is the difference between inferred (derived) improvement opportunity analysis and stated improvement analysis? Be able to explain and do both of these analyses.

The inferred improvement opportunity analysis is we are not asking the customer directly what we should fix, we infer this from the different types of data that we recieve. The inputs to an inferred analysis is the CX Attribute Importance Analysis and the CX Performance Analysis- the performance ratios. The graphs used for this is a combo chart and a scatterplot. The stated improvement analysis is a question that asks straight up how we are doing. "How important is speed of service?". Max Diff is used to ask the customers what is more or less important to you. We do not assume from data we just ask them straight up. The customers can give use direct qualitative answers in the form of stated improvement. Is not used as often but it gives is a different look and can help to build confidence.

Understand and be able to interpret RWA raw and rescaled scores. How do we know if the raw scores for an attribute are significant?

The original data that is emailed to you after submitting on the RWA-Web website. The first three pages we do not pay attention to, it is all of the coding stuff on how to get the results. We are first looks for R^2 for the model (this will be the same score as the multiple regression), ___% of the variance in NPS is explained by these ____ variables. Next we look for raw and rescaled weights (this is what we will be graphing). The 1st number is the raw weight and the 2nd number is the rescaled weight. We have to make sure all of the number are significant predictors/variables (are they even important). In order to do this we look at the confidence interval test, if zero is not included than the weight is significant.

What results or data do we use to create a "Performance-Importance" analysis graph?

The results that we use to create a Performance-Importance analysis graph is CX Attribute Importance Analysis, which are the RWA rescaled data points that we calculated from the previous weeks assignment. Then we use the CX Attribute Performance Analysis, where we collected the performance ratios. The performance ratios ares important to use because we see the best competitor results. We bring these two results together into a graph to see what we should improve on. We only come up with he top three improvement ideas.

Be able to create and interpret a "Performance-Importance" analysis graph in PPT. What type of graph do we create in PPT?

The type of graph we use in PPT for a Performance-Importance analysis is a performance matrix (which is a scatterplot in powerpoint) and a delta chart (which is a combo chart in powerpoint). Both of these graphs display the same results, but in different ways.

How do we find outside benchmarks of NPS performance?

To find outside benchmarks of NPS performance we simply go google it. For example: if we wanted to know the NPS score for a pizza place that was not in the original survey we just type into google "NPS score for pizza in 2020". You might have to read a little bit, but NPS scores are out there since they are so widely used.

From the "What is Marketing?" class, we know that Coach Garver is NOT a big fan of using demographics to understand the marketplace? If so, why does he suggest using demographic variables to analyze the sample?

To make sure the sample looks like the population

How do we calculate attribute performance scores? How do we run these scores in SPSS? Do we use means, frequencies or valid % for attributes?

We calculate attribute performance scores by looking at answers from survey questions. The survey questions are anchors only scales (angry/delighted, poor/excellent, very dissatisfied/satisfied). We look at where we are compared to competition. We run these scores in SPSS by going to analyze->descriptive statistics->descriptives. Then we select all of the attributes that we want to compare. We are looking at mean scores. We do this for both us and the competitors. When we do not know what competitor we want to pick we can look ash NPS scores, can choose we want to compare our company to or can take the best we can (per class).

With stated improvement analysis, do we calculate means, frequencies, or valid percents?

We calculate frequencies in SPSS, but the results that we are looking for are the cumulative percentages. In SPSS this is where we reverse the code, so the stated improvement scale will now read: 1. No Improvement Needed 2. 3. Minor Improvement 4. 5. Major Improvement Needed From this data we will look at the cumulative percentages from the SPSS database.

How do we interpret performance ratios? What is the range of values for the "zone of indifference"?

We interpret performance ratios by looking at the zone of indifference. The zone of indifference is .97 to 1.03. If our company is above 1.03, then there is a significant advantage for our company. If our company is below .97, then our company is at a significant disadvantage. In the zone of indifference, the company would not have a significant disadvantage or advantage. It is possible that companies could be tied, this does not mean anything because the company would not be at a competitive disadvantage or advantage. If the performance ratio is .92, then we would be at a disadvantage and 1.02 we would be in the zone of indifference (no advantage or disadvantage). 1.07 we would have a significant advantage over the competition.

Should we use multiple regression for attribute importance analysis? Why or why not? Should we use correlation analysis for attribute importance analysis? Why or why not?

We should not use multiple regression for attribute importance analysis. Everybody loves this, but it is not good at getting the proper CX results.... the problem is this analysis does not handle multicollinearity (independent variables are highly correlated) well, never was designed to work with a large number of independent variables (attributes). We should not use correlation analysis for attribute importance analysis either. Although it is better than the multiple regression, it is still not great. This analysis looks at 1 V 1 relationships, which means it cannot look at multiple relationships like we need it too.

How do we "visualize the data" for inferred improvement analysis (what charts, graphs, or tables do we create)?

We use two charts to visualize the data for inferred improvement analysis. The first one is a quadrant analysis, which a scatterplot is used for. This scatterplot shows where the attributes for performance and importance fall. The scales must be consistent for previous assignments, along with this assignment. The y axis is importance and the x is performance. Make the midpoints have a line going through it so it is obvious where the quadrants are. It is important to also add the zone of indifference in the graph. The second chart that is used is a combo chart. In order to start a combo chart, you must know that it is two graphs overlapping each other. We used the CX Attribute Performance ratio graph that we already created and turned it into a combo chart. This chart allows for two different scales (or two different inputs) on the same graph. The importance scores are added with a line graph. Again, the scales must remain consistent.

How do we "visualize the data" for NPS (what charts, graphs, or tables do we create)?

We visualize the data for NPS by creating charts, graphs, and tables. The table was used to show how the NPS score was calculate: it showed the promoters, detractors, and passives, then of course the NPS score. For looking at current NPS vs. competitors we use a column chart.. this chart look at all three pizza places. For the data over the last 5 years we use a line chart. We must sort using the TIME not the NPS scores. Each company will have separate chart for this. Can use markers for this to emphasis what years you were doing good or bad. When comparing the NPS scores to the benchmark we also use a column chart. We can easily show this by duplicating the slide and editing the excel file. We just have to make sure that we edit the new add ins by using the graphing checklist. Show that it is the benchmark compared to the other NPS scores, by adding a line or text (we want a little bit of separation). Each graph needs to be checked with the graphing checklist. It is important to implement presentation zen.

How do we "visualize the data" for attribute performance analysis (what charts, graphs, or tables do we create)?

We visualize the data for attribute performance by using a column chart, this is just the mean score of our company. The scales for this graph is 5-10, change the major units to 1, the decimal places are rounded to 2 decimal places. Looking at the side by side attribute of performance, we use another column chart. We compare us vs our competitors of the attribute of performance. In excel we must find the differences between us and the competition and sort by that before we make our chart. Use the same scales as the chart before! Remember if data labels make the graph too busy, do not include them. The performance ratio graph is another column chart. Make a formula in excel (US/Them), then copy and paste so the formulas are not used in the graph. This graph should show the zone of indifference and the axis of the scale can be different then, the other two graphs because of the results. 2 decimal places if we use them!

How do we "visualize the data" for stated improvement analysis (what charts, graphs, or tables do we create)?

We visualize the data for stated improvement analysis by using a radar chart and a bubble chart. The radar chart is specifically just from the stated improvement portion of the data. This chart is unique, different, and cool. It also works really well with stated improvements. It communicates the message clear, simple, and elegantly. There is no decimal places in this data for the graph. Arrange the data from smallest to largest and read from the smallest to largest on the graph. As we go around the scale more and more improvement is needed. A bubble chart takes inferred and stated importance and combines it together. We will use the data from importance and performance from the previous week and the stated improvement is the third variable. Looks similar to importance-performance analysis. X= performance, Y=Importance, Z= Stated improvement. the bubble size is the % of stated improvement.

How do we "visualize the data" for attribute importance analysis (what charts, graphs, or tables do we create)?

We visualize the data for the attribute importance analysis by using a bar chart and a table. The bar chart is used to change things up since we have been using a column chart all along (not to confuse the audience). We are using the same information for both the chart and the table. We will be using 1 decimal place for the attribute importance scores.