OM in the News: Jet Blue’s Unique Revenue Management Strategy

jet blueOne of the financial tricks in an airline’s tool kit is to sell more tickets than there are seats on a plane. If there are 150 seats, sell 175 tickets—people miss flights for a myriad of reasons and gate agents can typically muster enough volunteers who will take a later flight for a discount voucher. Yet this process doesn’t play out at JetBlue Airways, reports BusinessWeek (Feb. 5, 2014),  which has shunned “bumping” since its first flight 14 years ago. “Our traditional mission is to bring humanity back to air travel, and we feel that customers that purchase a seat should get a seat,” says the firm’s spokeswoman. It seems like a kinder way to treat travelers, but it might not be a smart way to run an airline.

The ultimate goal is to fill every seat on every flight, preferably in the order of who paid the most. Travelers flying on the lowest fares are those who also tend to volunteer their seats for compensation, while customers who pay the most—usually business travelers—can’t be tempted out of their seats. Overbooking pays off too: airlines almost always make more from the extra fares than they give back to volunteers in future-travel vouchers.

Yet because airlines have amassed years of detailed data on passenger no-shows—down to days, times, seasons, and specific routes—they only rarely need to write customers checks. The data also help them to know how to tweak their oversales for each flight, part of the complex algorithms that power revenue-management systems, the backbone of airlines’ fare pricing. Because it doesn’t overbook, JetBlue enjoys the lowest rate of involuntary denied boardings in the industry: only 18 people out of 21.3 million passengers through the first three quarters of 2013. On the other end of the spectrum, AirTran Airways had 1.28 passengers bumped for every 10,000 travelers (or 1,800 customers in total during the period).

Classroom discussion questions:

1. Should Jet Blue overbook, like other carriers do?

2. What other service industry overbooks?

Guest Post: Using Excel Teaching Videos at U. of Dayton

steve harrodDr. Steven Harrod is Assistant Professor of Operations Management at the University of Dayton. If you teach Excel modeling in your OM class, you will enjoy Steve’s detailed “how to” videos.

Videos are a great  way to help students learn the complexities of modeling in spreadsheets. Students can be overwhelmed when this material is presented live in lecture. Too often, students experience computer technical problems, and become lost in lecture while they attempt to fix their technical issue. These videos allow students to follow the lesson at their own pace, as they resolve problems such as installing add-ins (or even charging their battery!).

The first video supports Chapter 4, Forecasting. Titled “Creating an Exponential Forecast in Excel, Including Error Statistics” (http://youtu.be/uHy5tG1Rdvg), it walks
students through the process of formatting a spreadsheet, generating an exponential forecast, and calculating error statistics for that forecast. This video, and the others, was purposely recorded at a low screen resolution and high magnification, so that the visual clarity is high. The runtime for this video is 23 minutes.

The second and third videos are fundamentally linear programming exercises. “How to Create a Linear Programming Transportation Model” (http://youtu.be/RZX2bmoCzLI) is self-explanatory (runtime 14 minutes). Sharp eyed readers will recognize this as Problem 16 from Module B.

“Aggregate Planning on Microsoft Excel, Transportation Model” (http://youtu.be/m44gSMpb3Ic) is another transportation model, this time from Chapter 13, Aggregate Planning. This problem is “Planning Example 1” from the Heizer/Render text (Tables 13.2 and 13.3). This video is longer, at 36 minutes, because the data entry and model structure for the aggregate planning problem are more complex.

I hope you find these videos useful in your course presentation, and I welcome your comments at steven.harrod@udayton.edu.

Teaching Tip: The Secret to the Airline Pricing Model

This will probably be the 1st blog we do that can save you and your students money!  The Wall Street Journal (Jan.27,2011) just reported  that whatever you do, don’t buy your next airline ticket on a weekend.

We all know that airlines use revenue management (Ch.13) to maximize seat revenue. But the airlines don’t manage their inventory as actively on weekends, so if cheap seats sell on some flights, prices automatically jump higher. When is the best time to buy? The answer is Tuesday and Wednesday—and to be exact, it is at 3 pm on a Tuesday. “That’s when the maximum number of cheapest seats are on the marketplace”, says the CEO of FareCompare.com.

Though prices fluctuate frequently and the ups and downs of air fares can frustrate and anger us, it turns out that pricing has  followed  the same cycle during the week for many years. Discounts of 15-25% for seats are typically launched on Monday nights. “There is a method to the madness… behind the moves for airlines”, adds Expedia’s strategy director. “But for consumers it does seem crazy”. A ticket can be $199 on certain days and $499 other days, even months ahead of a flight.

Two weeks ago, reports the Journal, a Chicago-Atlanta  round-trip ticket for April  (on both American and Delta) cost $209 on Tuesday and Wednesday, but then $301 for the next 4 days. When Tuesday rolled around one week ago, the fares went to $219 —then back up to $307 by Friday.

This is certainly a topic that will interest your students.

OM in the News: Matching Supply (of seats) with Demand at Delta Air Lines

Have you ever been bumped from a flight that was over capacity? It’s a drama that plays out at departure gates every day in airports around the world. Typically, gate agents on overbooked flights embark on last-minute negotiations with passengers who might be willing to take a later plane. The agents broadcast their offers– vouchers worth $200-$400–and keep ratcheting up the price until enough passengers accept. Customers involuntarily bumped get an $800 voucher ( which the Transportation Dept. is proposing to raise to $1,300).

With 541,000 US passengers bumped in the 1st 9 months of 2010 (53,000 involuntarily), there must be a better way to manage capacity (Supp.7) and manage revenue (Ch.13). According to today’s Wall Street Journal (Jan.14,2011),  Delta Air Lines thinks it has the answer.

Delta’s high-tech new system (opened last month) asks passengers who check in online or at kiosks before going through security, what dollar amount they would accept to be bumped from their (overbooked) flight. Delta can then accept the lowest bids, eliminating a lot of uncertainty early. Not only does this give Delta a negotiating edge–passengers won’t know how low others are willing to go. But, in addition, “saving 3 or 4 minutes at the gate has a big operational impact”, according to Delta.  Delta calls it a “win-win” for both consumers and the airline.

Is this good customer service–or do any one of us even expect customer service when we fly? The topic fits well when discussing capacity and yield management issues in both Supp.7 and Ch.13. Given that 8-10% of passengers with reservations do not show for their flights, what other suggestions do students have?

Discussion questions:

1. Which system is better–Delta’s or its competitors?

2. What options do airlines have for capacity and demand?

OM in the News: Robots vs. People for Holiday Staffing

On-line retailers staffing up for the holiday rush are testing a critical hiring decision this year: man vs. machine. Ch.13 lists the options open to firms facing capacity issues, but robots introduce a new wrinkle. The Wall Street Journal (Dec.20,2010) highlights 2 firms with totally different staffing approaches: Crate & Barrel and Amazon. Both firms face a similar holiday crunch and e-commerce sales are up 12% overall, now accounting for 8% of US retail sales.

At Crate & Barrel, which sees holiday sales quadruple, the warehouse gets by with just double the number of employees, thanks to a cadre of 35 robots from Kiva Systems, a firm we highlighted in a recent blog. The machines carry racks of company products (8,000 SKUs) to people who pick and pack—no walking around the building at all.

By contrast, Amazon takes a more human-oriented approach. There, employees walk 18-20 miles a day down aisles lined with shelves, filling shopping carts with orders and carrying them back to packing stations. Amazon also quadruples its holiday staff—hiring a massive force of 11,500 seasonal employees at $11/hour. The firm tries to remove waste from the  hand-picking process with weekly  “kaizen” sessions. Amazon believes that people, not robots, give flexibility to handle its wider variety of products. It also uses hand held computers to guide workers to walk the shortest distance for each order, as described in an article in London’s daily, The Telegraph (Dec.21,2010).

“Which approach is better is a matter of debate in the 15-year-old e-commerce industry”‘ concludes The Journal. But Amazon does point out that it can now take orders up to Dec.19 and still guarantee Christmas delivery, a full 2 days later than last year.

Discussion questions:

1. Discuss the advantages vs. disadvantages of the human vs. robots staffing decision.

2. Where else has Kiva make progress in the fullfillment industry?

OM in the News: Yield Management Turns to Sports

We discuss the subject of yield (revenue) management in detail in Chapter 13, Aggregate Planning. Examples are provided from the airlines(American), hotels (Marriott), car rental companies (Hertz), and even Disney’s theme parks. But the latest issue of Operations Management/Management Science (OR/MS) Today (Oct., 2010) turns to an interesting and relatively new application of revenue management that may interest your students, namely, major league baseball.

Ticket prices to sporting events have always been priced to depend on the seat’s location. But the San Francisco Giants have discovered that  dynamic pricing of game tickets has increased 2010 revenues 6%. Ticket costs now depend on the opposing teams,  pitching match-ups, day of the week,  and even the weather forecast.

For example, a ticket in the Field Club, behind home plate,  for the Oct. 1st game between the San Diego Padres and the host Giants cost $68 at the start of the season. It went to $92 on Aug. 1st,$121 a week later, $145 on Sept. 4th, and $175  just before the game!

Discussion questions:

1. Are other sports and teams replicating this concept of dynamic pricing ?

2. Will there be fan pushback to the idea?

3. How did the 2010 pennant race impact on the Giant’s decision to use yield management?