You are on page 1of 4

ORIE 3800 Project (Spring 2011)

Instructor: Matthew Maxwell Date: 3/27/2011 Executive Summary


The RM Theater (RMT) company showcases hit theatrical productions in a beautiful outdoor arena. RMT has many years of experience in the performing arts but has no in-house expertise for business development and pricing. RMT has hired you to analyze their historical data and consult them on how to improve their business.

Project Guidelines
Do not procrastinate. You already have all the knowledge you need to complete much of the project. The topics I cover in class for the rest of the semester will mirror the skills you need to do the rest. I advise you to finish what you can now and fill in the rest as soon as you learn the information in class. The project will be graded out of 100 points. For those that go above and beyond what the project asks I may give up to 15 additional pointsthat is 4.5% of your total grade or about 1/3 of an exam. Try to teach me something I do not know or may not have considered about this dataset. Feel free to ask me for ideas or run your own ideas by me. Please be thorough but concise in your project report. Don't give extraneous info. Don't give incorrect info. Don't give misleading info. Write your conclusions and back them up with data. Some questions have exact answers and some do not. Don't be afraid to make assumptions, but clearly state what assumptions you are making. I'm more concerned about you having a logical connection between the parts of the project than making sure you have the exact same answers that I have. You may have to write computer code to accomplish this project. I do not need to see the code in the report (small snippets are fine if relevant). I may ask for an electronic copy of your code however. This is a real dataset and this project has not been done before. I think this is a terrific opportunity to get actual experience in data analysis and data-based decision making. Since we are venturing into unexplored territory, the scope of the project may change, but I will try to keep it from changing too much or too late. If you feel like there may be some additional data in RMT's historical data that you'd like to look at let me know. I'll be able to tell you if it exists or not. Ask me questions early and often!

Group Formation Guidelines


You may work on this project in groups of 3-4 people. Since this project will require significant data analysis it will be useful to have one team member who is comfortable with computer programming or SQL or Matlab or something else similar. Email Eric Guo (ag625@cornell.edu) a list of you team members by April 3 rd. If you do not know people in the class email Eric and ask him to place you in a group. Come up with a group name. It'll be fun. Besides, it will help me keep track of the groups. You may discuss ideas with other groups, but no sharing of code, electronic files, and so forth. When you hand in your project report, each group member will be required to complete a confidential evaluation of the other group members' participation. Negative evaluations will not affect the project score, but they may affect a person's class participation score.

Data Overview
Each year RMT has three shows and each show has multiple show times. In additional to show tickets, RMT also sells other extras such as a catered dinners before the show, backstage tours, and valet parking. RMT currently implements versioning and price discrimination in the following manner: 1. Versioning RMT sells seats in different areas of the theater for different prices: Area 1 are seats near the stage (most expensive), Area 2 are those further from the stage, and Area 3 are seats on the sides of the stage (least expensive). 2. Price Discrimination Within each area there are a number of different prices offered, e.g., child seat, student discount, etc. Thus each seat within an area is not always sold at the same price. Some tickets are also given away free for promotional purposes. RMT has provided you with three years of sales data for these shows. This includes information on when the shows were performed, every ticket purchased for every show, and every extra purchased for every show. Below is a description of the data files and their content. ShowInfo.csv One record for every performance
Time_ID Event_ID Unique identifier of the show time Unique identifier of the play being performed Week Week of the performances. 1 is the 1st week the show was performed, 2 is the 2nd week, etc. Weekday Indicates the day of the week of this show time. 0 is Monday, 1 is Tuesday, etc. Year Year of this show time DayOfYear The day of year of this show time. 1 is Jan 1st, 2 is Jan 2nd, etc. Area1Seats Area2Seats Area3Seats Number of seats reserved for Area 1 Number of seats reserved for Area 2 Number of seats reserved for Area 3

Tickets.csv One record for every ticket


Patron_ID Event_ID Unique identifier for the customer Unique identifier of the play being performed Time_ID Unique identifier of the show time DayOfYear Year Region Geographical region the customer is from. 13 total regions Area Which seating area the ticket is in TicketPrice The price paid for the ticket. The day of the year Year that the that this ticket was ticket was purchased. 1 is Jan purchased 1st, 2 is Jan 2nd, etc.

Extras.csv One record for every extra ticket


Patron_ID Event_ID Unique identifier for the customer Unique identifier of the play being performed Time_ID Unique identifier of the show time DayOfYear The day of the year that this extra was purchased. 1 is Jan 1st, 2 is Jan 2nd, etc. Year Year that the extra was purchased Region Geographical region the customer is from. 13 total regions IsDinner ExtraPrice Indicates The price whether the paid for the extra is a dinner extra. or not. 1 is dinner, 0 is not.

For the analysis in this project you will be required to manipulate, combine, and synthesize the data in these files. If you are using Excel or OpenOffice Calc make sure you are using the latest version otherwise you will not be able to load all of the records in the spreadsheet. Other tools you may use include MATLAB, a SQL database, custom programming, or whatever else your team is comfortable with.

Task 1 Data Analysis


RMT would like to get a good idea of the makeup of their customer base. Here are a number of questions they would like to gain insight into. Questions 1. RMT would like to see their data broken up by a number of factors. Such as when do people purchase show tickets? What percentage of tickets and revenue come from each geographical region? What percentage of patrons buy a dinner ticket as well? What about other extra tickets (i.e., valet parking and backstage tours)? How many tickets does a single patron buy for a show? How many different events do patrons buy tickets for? How has the number of tickets sold and the revenue changed over the past three years? How much revenue can be attributed to each of the three seating areas? i. RMT usually experiences a busy season when they must employ additional box office workers. What period(s) of the year do you recommend employing additional workers?

2. How many days prior to the show do patrons actually buy their tickets? Note that tickets go on sale no more than one year before the show date, so anything beyond that can be assumed to be erroneous data and must be dealt with somehow. For these erroneous entries, it is safe to assume that all the ticket information except the day/year the ticket was sold is valid information. i. Does this pattern change when considering what region the customers are from? Are you able to guess which region(s) are closest to the theater based on both the quantity of tickets purchased and how far in advance people buy tickets?

ii. Does this pattern change based on whether people are buying Area 1, Area 2, or Area 3 seats? iii. Does this pattern change based on what day of the week the show will be playing? What conclusions can you make about the customers' general purchasing behavior? Are there other factors that influence how far in advance customers purchase tickets? 3. How does the demand for show times differ according to how many weeks the show has already been playing and on what day of the week the show is playing? Is this information consistent if you analyze Area 1, Area 2, and Area 3 separately? Note that every show does not have the same number of performances or last for the same number of weeks. This fact must be taken into consideration to correctly analyze this information. 4. How often do shows sell out (i.e., there are no available seats in the theater to sell)? How often does Area 1 sell out? How about Area 2 and Area 3? i. RMT will often reserve some seats for various reasons. How do the above results change when we consider a 95% occupancy to be sold out?

ii. Which day of the week is most likely to sell out? Does the likelihood a show will sell out change based on the number of weeks the show has been running?

Task 2 Decision Analysis


RMT has noticed that a significant portion of their patrons opt to buy the pre-show dinners, but relatively few people pay for additional extras such as backstage tours and valet parking. RMT feels that since these extras are newer offerings many people do not yet know about them. Nevertheless, RMT feels that these extras could be significant source of revenue and would like to take advantage of that if possible. One strategy RMT has considered is creating a VIP pass which would include valet parking, a pre-show dinner, one of the best seats in the house to watch the show, and a backstage tour with the cast after the show. By bundling these products together RMT hopes to encourage more people to experience all that RMT has to offer and increase their revenue as well.

Questions 1. How much do you estimate that patrons would be willing to pay for the VIP pass? Note that the VIP pass has all the qualities of valet parking, a pre-show dinner, a show ticket, and a backstage tour. i. How much should RMT charge for the VIP passes? ii. How many people do you estimate will choose to purchase the VIP pass? How many of these people would normally have chosen Area 1? How about Area 2 and Area 3? iii. How much additional revenue would the VIP pass generate under the assumption that the VIP passes generate no new customers, but they may entice customers already planning to attend spend more? 2. The customer relations department feels that the VIP pass would in fact entice wealthy residents to attend more frequently. They believe that implementing the VIP pass would attract an additional 20 people per show on average who would be willing to pay at least $100 for a VIP pass. If this statement is true what does the revised demand curve look like? i. How much should RMT charge for the VIP passes? ii. How many people do you estimate will choose to purchase the VIP pass? For the existing customers what areas would you expected them to come from? iii. How much additional revenue would the VIP pass generate? 3. After being questioned rigorously, the customer relations department admitted that they are only 60% sure that the wealthy residents would come as they predicted. i. How much should RMT charge for the VIP passes? ii. How many people do you estimate will choose to purchase the VIP pass? For the existing customers what areas would you expected them to come from? iii. How much additional revenue would the VIP pass generate? 4. A local marketing research firm offers to investigate the customer relations department's hypothesis. The research firm claims that if the wealthy patrons would come they will be able to detect this with 70% accuracy, and if the wealthy patrons would not come they will be able to detect this with 90% accuracy. How much would RMT be willing to pay for this research? 5. RMT doesn't necessarily believe the marketing firm's accuracy quotes. Assume that the real accuracy for the two quantities above is uniformly distributed between 60%-70% and 80%-90%. Perform sensitivity analysis on the additional revenue under these conditions. You may assume that RMT continues to base the VIP price on the research firm's quoted accuracy, but the real probabilities of whether or not the wealthy people decide to show given the firm's research are determined via the adjusted probabilities. i. Are there situations where RMT would actually lose money by following the market research firm's advice?

ii. Now how much would RMT be willing to pay for the market research under this assumption?

You might also like