Download Tb Applied Marketing Analytics Using R Gokhan Yildirim, Raoul Kübler and more Exams Marketing in PDF only on Docsity!
Test Bank For
Applied Marketing Analytics Using R Gokhan Yildirim, Raoul Kübler Chapter 1-
Chapter 1: Introduction
Multiple choice questions
- Which of the following best defines marketing analytics? a. Marketing analytics is the process of analysing datasets in a systematic way to improve sales outcomes with the help of machine learning models. b. Marketing analytics is the process of collecting and analysing datasets in a systematic way to improve sales outcomes with the help of analytic tools and techniques. c. Marketing analytics is the process of analysing datasets in a systematic way to draw conclusions on customer acquisition and retention strategies with the help of analytic tools and techniques. d. Marketing analytics is the process of collecting and analysing datasets in a systematic way to draw conclusions on marketing strategies and improve business outcomes with the help of analytic tools and techniques. e. Marketing analytics is the process of collecting and analysing datasets in a systematic way to improve business outcomes with the help of text mining and image analytics tools.
- The diagram below illustrates the typical five-stage process that marketing analytics applications go through, but it lacks the names of these stages.
Which of the following captures the process best?
a. 1: Analytic models, 2: Insights, 3: Action, 4: Cloud storage, 5: Data b. 1: Data, 2: Analytic models, 3: Cloud storage, 4: Action, 5: Insight c. 1: Cloud storage, 2: Analytic models, 3: Data, 4: Insights, 5: Action d. 1: Insights, 2: Cloud storage 3: Action, 4: Data, 5: Analytic models e. 1: Data, 2: Cloud storage, 3: Analytic models, 4: Insights, 5: Action
- Incorporating the results of quantitative models into intuition-led decision-making is referred to as: a. Face validity b. Quantification c. Quantitative intuition d. Content validity e. Predictive analytics
- You have observed that the model-based recommendations align with your knowledge on the topic and intuition-led recommendations regarding the direction of change for your marketing decisions. In marketing analytics, this phenomenon is known as: a. Data mining b. Face validity c. Quantification d. Logical thinking e. Market trend analysis
- The model precisely indicates the extent to which each marketing action should be increased or decreased. This is known as: a. Quantification b. Data integration
some of our analytics projects remain incomplete and some have failed to meet our expectations. We have not observed a significant improvement in how analytics contribute to our company’s performance really’. What are the common reasons behind the failure of marketing analytics applications? Please discuss your answer briefly.
- You are being interviewed for the position of marketing analytics director at a fintech company that has not previously had a marketing analytics team. The role entails building a successful analytics team. Which factors or key players would you consider when forming such a team? END of TEST
Answers
Chapter 1: Introduction
- Ans: D
- Ans: E
- Ans: C
- Ans: B
- Ans: A
- Ans: D
- Ans: E
- Ans: Students are expected to discuss the applications of consumer profiling, text mining, real-time targeting, media planning, measuring return on marketing investment and predicting future demand. Some students with a background in analytics or data science could give examples of customer analytics tools such as recency-frequency-monetary and customer lifetime value.
- Students should discuss the following points in their answers: Marketing analytics fails if:
- you do not define the business problem clearly
- the model and data do not map to the business problem
- data collected by different systems is disjointed
- you neglect unobserved factors
- you do not apply the model correctly
- you focus on performance metrics that do not matter
- you do not have the right talent to leverage marketing analytics
- The marketing analytics director (MAD) is a leadership role that requires strong domain knowledge as well as strategic decision-making skills. Building a successful analytics team requires careful consideration on the following factors:
- Mindset. The team members must have the right mindset and embrace the fact that analytics and AI-supported tools can resolve marketing issues and/or unearth new opportunities for growth. The MAD should consider the following essential roles for her team: Data analyst. People working in this role are typically in charge of (i) collecting, integrating and maintaining large-scale datasets and (ii) developing algorithms, executing analytical models and preparing data visualizations. This role is quite technical and requires advanced data, software and programming skills. Data translator. This role plays a bridging role between data analysts and data directors. Data translators help analysts (i) look at the right marketing problems and how to approach them, (ii) define the scope of a project and outline the deliverables and (iii) interpret the results and determine the next best course of action. Also, they generate project reports and communicate the results to marketing analytics directors. These people should have training in marketing analytics and possess some strong team and communication skills. How big the team should be, of course, depends on the size of the organization, the long-term needs and other financial constraints. To be able to get analytics projects up and running, the team should have access to data analytics software such as R, Python, MATLAB or SAS along with high-powered computers.
Testbank
k-means:
- What is the primary goal of k-means clustering? a. To minimize within-cluster variance b. To maximize between-cluster variance c. To find the optimal number of clusters d. To assign data points to predefined clusters
- How does the k -means algorithm initialize the cluster centroids? a. Randomly select data points as initial centroids b. Place centroids at equal distances from each other c. Assign each data point to the nearest centroid d. Use the mean values of each feature as initial centroids
- What happens in each iteration of the k -means algorithm? a. The cluster centroids remain fixed, and data points are reassigned to the nearest centroid b. The cluster centroids are recalculated based on the current assignments c. The number of clusters is adjusted based on the within-cluster variance d. The algorithm terminates if the within-cluster variance reaches zero
- How is the optimal number of clusters determined in k -means clustering? a. By visually inspecting the scatterplot of the data b. By using the elbow method and evaluating the within-cluster variance c. By performing a hierarchical clustering analysis d. By setting the number of clusters based on domain knowledge
- What is a limitation of k -means clustering? a. It is sensitive to the initial placement of cluster centroids b. It cannot handle high-dimensional data c. It only works with numeric data
d. It is computationally inefficient for large datasets
STP analysis
- What is the primary purpose of segmentation in the STP analysis? a. To identify the target market for a product or service b. To position the product in the market c. To create a unique value proposition d. To differentiate the product from competitors
- What is the key outcome of the targeting phase in the STP analysis? a. Developing a positioning strategy b. Conducting market research to understand customer preferences c. Identifying specific segments to focus marketing efforts on d. Analyzing the competition and market trends
- What is the main objective of positioning in the STP analysis? a. Identifying potential customer segments b. Establishing the target market for the product c. Creating a distinct and favourable image of the product in the minds of consumers d. Developing a pricing strategy for the product
- Which factor is important when selecting segmentation variables for STP analysis? a. Variables should be measurable and observable b. Variables should have a strong correlation with customer preferences c. Variables should be relevant and meaningful to the target market d. Variables should be unique to the product being marketed
- What does STP stand for in marketing? a. Sales, tactics, promotion b. Strategy, targeting, planning
the insights gained, develop a targeting and positioning strategy. This strategy should be presented in the form of a report or presentation. Please note that many of the Kaggle datasets include vignettes, where Kaggle users explain how to analyse these datasets. These vignettes may provide students with solutions or answers. One possible approach to address this issue is to remove certain observations from the original dataset, thereby altering its distribution. By doing so, students cannot solely rely on the tutorials and must develop their own solutions. Retail and Customer Spending Data: https://www.kaggle.com/datasets/vetrirah/customer Indian Bank Customer Data: https://www.kaggle.com/datasets/shivamb/bank-customer-segmentation Wine Types: https://www.kaggle.com/datasets/sadeghjalalian/wine-customer-segmentation Hotel Bookings: https://www.kaggle.com/datasets/nantonio/a-hotels-customers-dataset Online Retailing: https://www.kaggle.com/datasets/yasserh/customer-segmentation-dataset
Answers
Chapter 2: Customer segmentation
Cluster numbers
- Ans: A
- Ans: B
- Ans: B
- Ans: C
k-means
- Ans : D
- Ans: C
- Ans : B
- Ans: B
- Ans: A
STP analysis
- Ans : A
- Ans: C
- Ans: C
- Ans: C
- Ans: C
Euclidean distances
Ans: Consumer 6: Age 35, Income $50,000, Education 12, Household size 3
Testbank
Chapter 3: Marketing mix modelling
Multiple choice questions
- Which of the following best defines a marketing mix model? a. A model that evaluates the impact of marketing decisions on key performance metrics b. A model that focuses on advertising effectiveness
- ‘ Initially, spending more and more money on advertising is beneficial, but after a certain point, the additional value gained from an extra spending will be very small‟. Which of the following best describes this phenomenon in a marketing mix model? a. Reciprocal transformation b. Synergistic effect of advertising c. Average returns to advertising d. S-shape functional form e. Diminishing returns
- ‘______ represent the “unexplained” part of the model, that is, the impact of other factors that are not explicitly included in the model‟. Which of the following correctly completes the sentence? a. Residuals b. Predictors c. Slopes d. Response variable e. Model fit
- Suppose that you have estimated the following log-log regression model and found 1 to be 0.75. ln(Salest) = 0 + 1 ln(Advertisingt) + Which of the following represents the correct interpretation of this finding? a. A one-unit increase in advertising is associated with a 0.75 unit increase in sales b. A one-unit increase in advertising is associated with a 0.75% increase in sales c. A 1% increase in advertising is associated with a 0.75 unit increase in sales d. A 1% increase in advertising is associated with a 0.75% increase in sales e. The coefficient of 0.75 is not meaningful and cannot be interpreted
- Suppose that you have the following log–log model. ln(Salest) = 0 + 1 ln(Advertisingt) +
Which of the following equation shows the unit (marginal) effects of advertising correctly? a. 1 0 b. 1 Salest c. 1 Advertisingt d. 1 (Advertisingt Salest) e. 1 (Salest/Advertisingt)
- In marketing mix models, the purpose of model diagnostic checks is: a. To identify the impact of external factors on marketing variables b. To measure the effectiveness of marketing campaigns c. To assess if model assumptions are met d. To determine the optimal allocation of marketing budget e. To analyze customer segmentation and targeting strategies
- In a marketing mix model, if the R -squared value is 0.65, what does this indicate about the model’s performance? a. The model has a 65% accuracy rate in predicting customer purchases b. The model explains 65% of the total variation in the response (dependent) variable c. The model has achieved a 65% increase in sales due to marketing efforts d. The model demonstrates a 65% fit to the marketing mix variables e. The model has an error rate of 65% in predicting sales
- When developing a marketing mix model, the marketing scientist says: ‘TV advertising may serve a dual role. It may increase sales directly but also it may increase the effectiveness of Google’s paid ads’. How can this be addressed in a marketing mix model? a. By including TV advertising and Google’s paid ads as separate variables in the model b. By creating an interaction term between TV advertising and Google’s paid ads to capture their combined effect
- The coefficient in a log-linear marketing mix model (i.e. the log–log specification, which involves taking the logarithm of both sides of the regression equation) can be interpreted as elasticities. a. True b. False
Open-ended questions
- You are recently hired as a marketing scientist at Adidas, a global brand that specializes in sportswear, footwear, and accessories. As part of your role, you have developed a marketing mix model to evaluate the effectiveness of the brand’s TV advertising campaigns. Now, the question arises: How would you determine if your model is a good one?
- You work as a marketing analyst for a UK-based e-commerce company focused on fast- fashion products. Recently you were tasked with developing a marketing mix model for the sales performance of one of the popular basic crew neck T-shirts. You have tried several models with different sets of variables and now need to decide which model to use based on their predictive performance. How would you address this?
- You are being interviewed for the position of a data scientist at Apple, an American multinational technology company. During the interview, you are presented with a case study on a marketing mix model and tasked with proposing possible improvements to the model. As you examine the model output, you observe that the root mean squared error of the model is significantly higher for the test set compared to the training set. A closer examination of the model equation and the model fit plot based on the training set indicates that the model was excessively tailored and overly complex to fit the training data. (i) How would you identify this phenomenon? (ii) How would you address it? END of TEST
Answers
Chapter 3: Marketing Mix Modelling
- Ans: A
- Ans: D
- Ans: C
- Ans: B
- Ans: E
- Ans: A
- Ans: D
- Ans: E
- Ans: C
- Ans: B
- Ans: B
- Ans: E
- Ans: B
- Ans: A
- Ans: A
- Ans: The answer should focus on the model diagnostics. In a marketing mix model, we make assumptions about the residuals of the model. Residuals should be uncorrelated (i.e. independent), have zero mean and constant variance. Students can briefly discuss the meaning of these diagnostic checks. Once the model passes these diagnostics, it is decided that the model is not misspecified. At this point, one can examine the R -squared and adjusted R -squared measures to determine the extent to which the model explains the variation in the response variable. A more detailed interpretation of these measures is provided in the R tutorial of the chapter on marketing mix modelling. Furthermore, although not explicitly asked, students may also discuss the assessment of the model’s predictive ability. This can involve techniques such as data splitting into training and test sets and evaluating error measures to test how well the model predicts unseen data. These aspects can complement the evaluation of the model’s performance.
- Ans:
Chapter 4: Attribution modelling
Multiple choice questions
- Which of the following best defines a marketing attribution model? a. Attribution is a statistical technique used to analyse consumer behaviour and predict future purchase patterns based on historical data b. Attribution is defined as the identification of a set of customer touchpoints that contribute to a desired outcome (e.g. purchase) and the assignment of a credit to each of the touchpoints on the customer journey c. Attribution is a marketing strategy that focuses on assigning credits to individuals within a team for the success or failure of a campaign d. Attribution is a method of tracking the number of times a marketing message or advertisement is viewed by potential customers e. Attribution is a process of creating personas to represent different segments of a target audience and tailoring marketing messages accordingly
- Which of the following marketing attribution models do not aim to answer? I. Which marketing channels or touchpoints played a significant role in a customer’s journey towards conversion? II. How much credit should be given to each marketing touchpoint in the customer journey? III. What is the target audience for the marketing campaign and how do they interact with digital touchpoints? a. Only III b. I and II c. I and III d. II and III e. I, II and III
- Consider a marketing campaign that includes various touchpoints such as social media ads, email newsletters and a website landing page. The campaign resulted in a customer making a purchase. The marketing team wants to analyse the impact of each touchpoint on the conversion. They decide to use an attribution model that assigns credit to only one touchpoint. Which attribution model(s) are they most likely using? I. Bathtub II. Time-decay III. Even split IV. First-touch V. Last-touch a. Only I b. I and IV c. II and V d. III and IV e. IV and V
- The following image represents a customer’s journey over multiple touchpoints.
What type of marketing attribution model is this? a. First-touch b. Last-touch c. Even split d. Time-decay e. Bathtub