AD654: Marketing Analytics essay

Buy ready-to-submit essays. No Plagiarism Guarantee

Note: Every paper is crafted by human writers to ensure authenticity and originality.

AD654: Marketing Analytics

Boston University

The Disney Corporation is considering some tough strategic decisions at this time, and they heard that you and your team knew about analytics, and about theme parks.  They have come to you for help.

Note:  Some of the data used in this project is real, while other data is simulated for the project.  

Your team will submit a written report via Blackboard.  This report should come in the form of one ZIP file.  This ZIP file will include:

• A written document that includes a link to your Tableau dashboard and your write-up/description of the dashboard.
• A PDF with an accompanying .ipynb that includes all of your results for the following parts:
  • Summary Stats
  • Segmentation and Targeting
  • Forecasting
  • Classification
  • A/B Testing
• Strategic Analysis Memo (< 3 pages in length)
• Your conclusion — this can either come in a Markdown cell at the end of your Jupyter Notebook or it can be its own document that you upload into the .ZIP. 

During the last week of the course, your team will deliver a 15-minute presentation in Zoom to your Professor.

Some things to keep in mind:

• As you move through the various tasks, remember to “call ‘em like you see ‘em.”  If you see results that aren’t pretty, it does not necessarily mean that you did something wrong, or that there is a problem with the dataset or the software. 
• Some tasks in this project are very similar to things that we have done in homework assignments, whereas others are unique.  Take advantage of all your available resources…but keep in mind that the teaching team will remain at “arm’s-length” distance from all project teams. 
• Assumptions are okay — if you make an assumption anywhere along the way, you can just state it so that your facilitator can see why you took some particular step.
• A rubric will be available on Blackboard.

Data Visualization (10 points):  Disney’s movies are one of the most valuable jewels in its crown. 

Disney knows that many of its films have been major commercial successes, but it wants to understand this success with even greater detail. 

Using Tableau, build a dashboard that includes anywhere from 4 to 6 visualizations created from the variables in the disney_movies.csv dataset. 

Include a 1-2 paragraph description of your dashboard that talks about the plots you made and some of the valuable takeaways and insights that they may provide for management. 

Summary Stats (5 points):   In Python, conduct some exploratory data analysis of the disney_movies.csv dataset.

You may want to present some summary statistics of the entire dataset, but should also consider some groupings of variables, using groupby() or pivot_table() from pandas.  Create anywhere from 4 to 6 total summary statistics for this section. 

In a paragraph, describe your findings.  What did you learn about the dataset?  Mention any insights that might be particularly useful or valuable for management.  Use a markdown cell to write this paragraph. 

Segmentation and Targeting (25 points):   After a very rough period over the last 15+ months, Disney is hoping to bring more families back to its parks. 

Here’s where your team can be a huge help to the company:   Disney recently reached out to a large data aggregation company to obtain information about households that might  but the park doesn’t know how to begin analyzing it.  The dataset is named family_segments.csv.   There’s a lot of data here, though!  There are more than 9000 observations in the file.  Disney is hoping that your team can help them to separate this huge group of people into distinct clusters, and then figure out how to reach out to each of these groups from a marketing perspective. 

You may wish to use either k-means or hierarchical clustering for this task.  To perform the actual clustering, use only your numeric variables (but when you analyze your clusters, you can include observations about the categorical factors).

Once you have built your clustering model, use anywhere from 4-6 visualizations that help to communicate information about your model.  The visualizations should depict information about your clusters that you can clearly explain, and that park management can understand.  So stay away from things like PCA and t-SNE!  

Name each one of your clusters, and include a few sentences describing/explaining the name that you chose for each cluster. 

Finally, for each of your clusters, talk about targeting.  In a couple of sentences per group, how should park management reach each of these segments? 

Use markdown cells for the cluster names and the targeting section.  Also, use a markdown cell to describe the process that you used for arriving at the number of clusters for your model. 

Note:  As you explore, review, and analyze this dataset, keep in mind that it is a specific sample of household data that was delivered to Disney by a data aggregator.  If the numbers do not seem representative of the entire population, that’s because they aren’t supposed to be!

Conjoint Analysis (20 points):  

As you know from the case, Disney is considering whether to build a new hotel in Hawaii. 

In order to better understand the features that hotel guests might prefer, Disney has commissioned a survey of people who regularly vacation to Hawaii.  The survey respondents answered questions about features that they hoped to see in the hotels.  Disney heard that you can perform some ratings-based conjoint analysis, which they hope you will be able to apply to this dataset.  You can find the ratings in hotel_amenities.csv. 

There’s just one little problem, though — the people who answered the questions did so without needing to consider costs.  In the business world, this is a luxury that no one has — not even Disney! 

The amenity_costs.csv dataset contains information about the per-serving cost associated with each option that has been presented to Disney. 

The number crunchers have determined that Disney can price the rooms in a way that would enable the hotel to incur $250 in costs, per room, per night.  With that in mind, analyze the conjoint model results, and come up with recommendations for the park regarding your recommended set of amenities for this new hotel. 

Strategic Memo (10 points):  You and your teammates should each read the case, “The Walt Disney Company:  Theme Parks.”  This case will not be posted to Blackboard. 

After reading the case, and thinking about the major decisions that Disney faces, write a strategy memo of no more than three pages to advise Disney on a path forward.

Some of the major questions that Disney seems to be facing include:  whether to launch and scale DisneyQuest, whether to construct the standalone Hawaiian hotel, and whether to invest in a technological upgrade of the company’s current parks. 

For this section, consider this case to be a “closed universe.”  You may bring your opinions, and your teammates’ opinions, into the case, but you will not need to do any outside research, and should not cite any outside research here. 

This memo should include some quantitative analysis.  You may include visualizations within your three pages, but this is not a requirement.  If you build any visualizations, use information in the case and its exhibits to do so. 

Forecasting Net Income (2.5 points):   Through an online search, you can find recent annual reports published by Disney. 

Using several years’ worth of data, and any forecasting method that your team deems appropriate for this purpose, generate a forecast for Disney’s 2021 net income.  Use any forecasting tool available in Python to accomplish this.

In a markdown cell, write one paragraph that describes your process and results.  In addition to the net income data from previous years, you may also wish to consider any other factors — but keep the weighting of this section in mind here. 

Classification (20 points):   In this section, you and your teammates will use the cruise_returners.csv dataset to help Disney to better predict whether a particular household who embarks on a Disney Cruise will purchase tickets for a follow-up voyage. 

You may use any classification method that you have seen anywhere in AD654 material.  Show the results of your model, and the steps that you used to build it. 

Write 3-4 thoughtful paragraphs about the conclusions that Disney can draw from this model’s results.  In this assessment portion, you may wish to point out any specific details about your model to help company management to better understand your results.

A/B Testing (5 points):   Disney’s popcorn buckets are a big hit!  Lately, Disney’s popcorn buckets have proven to be very popular with collectors.

Disney is hoping that you can narrow down its offerings, though.  Using data from popcorn_buckets.csv, you will perform some analysis to compare the performance of each of these buckets’ recent sales. 

Using the appropriate statistical test for this situation, determine at a 95% significance level which head-to-head popcorn bucket matchups show meaningfully different results. 

In a markdown cell, write one paragraph that describes your process and your results for this analysis.  What should Disney do? 

Conclusions (2.5 points):  Here, write 1-2 paragraphs to wrap everything up.  Do not merely describe your findings as if you were summarizing a lab report.  Instead, emphasize your team’s insights about how these findings could be helpful/meaningful for Disney, theme parks in general (even including Lobster Land!), or about the process itself.  For this section, you may simply use a Markdown cell, or you may write in a separate document that you include in your ZIP — either way is fine.