Pivot!

It's a tale old as time. You start off with an idea. It sounds great. The feedback says it's great. So the plan proceeds. But then you realize this is happening: 

We (myself, the ITP and NYU Poly collaborators) had developed visual prototypes for a moviegoing and movieticketing app. Its features included variable pricing, the social event aspect of moviegoing, and other features that acknowledge that the decision to attend a movie is not motivated only by price. 

It seems we made a car and forgot the engine. The heart of the project remains answering (or giving the best go of it possible) whether or not dynamic or variable pricing would bring in uncaptured demand. Although great research has been uncovered throughout the last several quarters, we haven't clearly answered that question. Part of it was due to lack of data. But all our work isn't for naught, it's just that a shift, a pivot is needed. 

For the remainder of the fellowship, we're building the engine: building models and running simulations and making assumptions about what variable pricing would do to attendance and revenues. "If X is reduced by Y%, what happens to Z?" Financial 10-Ks for theater exhibitors will be useful to building hypothetical profit-and-loss statements, and we've found some comparable industries (that do variable pricing). In the end, we'll build infographics of our results and a comprehensive whitepaper of this year's research. And then, present our prototype. 

It's tempting to say, "why didn't we do this first?" but there were some big challenges. Firstly, when we couldn't get the data sources required, I was uncomfortable with just building simulations off of assumptions. Plus, there was already academic literature out there about variable pricing's merits; what else could we add? So I jumped straight into the prototype phase. But as some wise advisors mentioned, these simulations are just a start. We'll keep it as simple as possible. We're opening up the conversation and gladly invite others to point out what is wrong, how we can improve our assumptions, and adjust accordingly. 

Stay tuned for more progress updates! This is what we want in the end:

A team, two visits, and the hunt for data

Elastic Movies has a team! The goal is to build a demo prototype by mid-July and we kick things off in earnest next week. We've got an NYU Poly student  from the Integrated Digital Media, two ITPers (one current, one graduated), and one Stern student (myself). Concurrent to our work building a dynamic pricing app, I am trying to procure more data about attendance. Not just daily attendance by movies, but also, the number that go at any given point of day. Visits to theaters have yielded interesting, anecdotal research about the movie theater business. However, we still need more data! In addition to continuing to reach out to theaters, I will also be reaching out to Rentrak.

Some recent learnings:

• People who jump from movie to movie, but pay only once are called "hoppers." Pre-assigned seating makes it easier for staff to find hoppers.
• Most of the people that come before 12 pm are the elderly and regulars. 12-2 pm moviegoers are killing time. 4 pm onward is the post-work crowd. A lot of couples attend from 8 pm onward.
• As expected, big event movies draw larger online ticket sales. Competitors, though, also impact a consumer's decision to pre-buy. The fewer competitors around you, the more market power you have, the more likely consumers are to pre-buy for big movies. It is more difficult to find another nearby movie theater that might be playing that same big movie.

New partners-in-crime: NYU Poly Integrated Digital Media + IFP Media Center

It takes a crew to build a ship, right? The project (tentatively titled Elastic Movies currently) is now in the "build" phase. There are two big priorities over the next several months: gather data about movie ticket attendance (as granular as possible) to develop an algorithm, and prototype an app. This week, I joined the IFP Media Center as a fellow.  Newly opened just 7 months ago in Dumbo, the Center is a partnership between Made in NY and IFP. There are a number of gaming, new media, and entertainment companies already in the co-working space. I've got some office space in the incubator program and hope to find some great collaborators here. Or at the very least, smart folks to bounce ideas off of and frequent commuting walks over the Manhattan Bridge.

To build the app and rethink movie pricing, a partnership with NYU Poly is in the works. The Integrated Digital Media co-director Luke DuBois gave a tour of the year-old floor and the work that they're doing is pretty awesome. The best part is that this space holds not just students from the IDM program, but from other NYU groups that are interested in cross-collaboration. The MFA Gaming group is here, instance. A blurb about this dynamic ticket pricing project goes up today, and hopefully, some great Poly collaborators will be joining the project soon.

Case study series on dynamic pricing: MLB

Ticket brokers like StubHub and scalpers brought dynamic pricing into sports ages ago, but not until recently did officially-sanctioned dynamic pricing emerge. The MLB has been the most aggressive over the last several years and 26 of the 30 teams now use dynamic pricing. Last time, the airline industry was placed under the microscope. This week, let’s assess the MLB and what ticket pricing principles in that realm can be transferred to the theatrical business. Unlike the movie industry, there is a lively and active secondary ticket selling market, most prominently on StubHub. But teams became concerned this resale market devalued tickets and responded with dynamically priced tickets and other programs for ticket resales. The SF Giants paved the way four years ago. Variables used included: weather, winning streaks, starting pitchers. The algorithm came from ticket-pricing firm Qcue, who built it with data on popularity of opposing teams and prices from online secondary ticket markets.

The pilot was a resounding success - it pulled in an extra $500K over the season. With a World Series win and solid performance since then, the Giants continue to capture more market value than under a static pricing system. Ticket revenues are up by 7-8% per year. Though season ticket sales skyrocketed, the team still caps a quarter of the ballpark to dynamic pricing. The prices varied by $1-2 at the beginning, but can now go as high as $5-15 more for high-demand games.

Between studying the MLB and the airline industry, the project continues to narrow down into the exact type of test we’ll conduct:

Here are the major takeaways from the case study of the MLB:

• The relationship between sports games and movies draws more similarities than airlines. However, movies – unlike sports – are not one-time events. Watching a movie is still less “must-see” because you can see it in many other formats later.
• The goals of this project are similar to what the Giants and MLB wanted: more revenue from high-demand movies (games) and ensure that seats at less popular movies (games) that would otherwise go vacant are sold.
• Contact Qcue for a conversation. Que also sets prices for 30+ NHL, NBA, MLS, and MLB teams.
• Start finalizing the short list of variables we want to consider.
• Where can we get data to develop a simple-as-possible algorithm?
• Set a ceiling and a floor for ticket prices to preserve revenues and to ensure people do not think you’re greedy. Consumers will not buy into perceived greed.
• The type of test the project continues to narrow. The current intended focus: reward those who buy tickets early. Prices increase as ticket inventory decreases or as the date of the movie approaches.

Case study series on dynamic pricing: Airlines

There’s no business like show business, but to better inform the shape of the dynamic pricing in this project, I decided to take a closer look at dynamic pricing in the other industries in a series of case studies. The airline and hotel industries famously use dynamic pricing (tens of thousands of times daily, according to some reports). Dynamic pricing has also spread to other sectors: some sports teams (the SF Giants), at least one restaurant (Next in Chicago), and transit (Uber). At best, this series allows the project to shamelessly crib best practices and sidestep pratfalls. We kick things off with a study of the airline industry. Just like movie theaters, airlines seek to maximize the revenue they can generate per person and balance the challenging operating metrics they must meet to maintain their thin, thin margins. Bad news: the differences between the airline and theatrical exhibition mean we should not replicate airline dynamic pricing practices. Good news: research on dynamic airline pricing has indicated price discrimination grows both revenues and consumer surplus (aka satisfaction) for airlines. This is encouraging news for the scope of our project.

Let’s dive into the pricing mechanics.

Up in the Air

Airline prices usually follow this trend: prices are cheap the earlier you buy, peak when demand is particularly uncertain, then dip again as the airline attempts to fill as many seats and minimize unfulfilled demand, and spike to take advantage of last-minute purchases and very inelastic demand.

The components of the airline ticket price are: cost of service, cost of not selling that seat on a substitute flight, and the forgone option of selling the seat later on the same flight. Airline purchasers must weigh the expected gain from delay against the cost of failing to acquire a seat. This is something that moviegoers are unlikely to spend much time considering due to a number of inherent differences.

The biggest difference is demand elasticity. The power of demand for the ticket is quite different. Theater attendance is more responsive to price changes than the airline industry because movie tickets are discretionary. Airline tickets are often purchased because the consumer must absolutely get there. Consumer demand is often unfulfilled until they buy the ticket, offering some leverage to airlines. In contrast, movie theaters can go empty because the audience just chooses to go elsewhere or see it on another platform.

The power of competitors varies greatly between the airline and movie businesses. The airline industry has very few competitors and few substitutes. Buses and trains tend to be perceived as cheaper and worse because they are quite demanding on personal time. This leaves airlines as the dominant mode of long-distance travel. The same can’t be said for movies in the content landscape. Everywhere you turn, there’s a new competitor for your hard-earned leisure time: television, video games, online video, and plays, just for starters. This ties back into how price sensitive moviegoers are versus fliers.

Though airline industry is a bad proxy for this project, there were some good takeaways from this case study, including:

• Airlines tend to do: peak user pricing and time of purchase – how can this be applied?
• Focus on the customer dynamics and use this to shape how we build this model
• Cannot assume that an efficient market will include fully sold out seats. The optimal amount may not be the case.

Next up in this research series on dynamic pricing: sports.

New web apps off Rotten Tomatoes' API

You and a friend are talking about a movie. It looks great. Rotten Tomatoes says it's fresh. Inevitably, one of you will say, "I really want to see that!" For every time you say that, how many times did you actually buy the ticket? Would you if you knew it was leaving the theater soon? This week, my focus shifted from pricing to the movie consumer's decision process. There are multiple pain points, but for the purposes of this week, the focus was information. It's not effortless to keep track of what movies are still in or not in theaters. Plus, if you aren't sure what movie you want to see, but just want to see something, a quick-and-dirty display might be more effective than a comprehensive, detailed display.

Here are two quick web apps off of the Rotten Tomatoes API that might help. The first is called QuickRT. It's a quick snapshot of current releases and how their scoring, presented in a simple, user-friendly design.

The second web app

is a sand timer for movies. It's the result of my final project for an ITP class. Here's a snapshot:

CRI Post #3.2

First, it grabs all the in-theater moves from the Rotten tomatoes API and filters out movies greater than 90 days old and rotten movies, leaving only the best and freshest. Then the movies are sorted by how long they've been in theaters. The current display is "90s chic" but it could be redesigned to a user-friendly, visually appealing design. "Almost gone" movies could be displayed in red boxes, "On-its-way" movies in yellow boxes, and "Just in" movies in green boxes. Would knowing a movie is almost gone entice you out of couch inertia? While this feature might not be baked into the final product, it's a start to understand how movie habits might be shifted. The app assumes that most people want to see movies that are "certified fresh." It certainly doesn't account for dissatisfaction with the movie itself. After all:

CRI Post #3.4

Misanthropes should go to movies on Tuesdays

The movie theater business is a weekend-centric one. It’s no surprise that movie tickets sold on weekend (Friday to Sunday) are triple the number sold on weekdays (Monday to Thursdays). To better define what type of dynamic pricing this project will pursue, I dove into data about attendance by date. The trends of the past tend not to chart the course of the future, but I hoped the data would yield some clues.

There are two primary sources online – BoxOfficeMojo.com and BoxOffice.com. I chose the former because the site also includes a cornucopia of data on attendance per title per date that will be helpful going forward. The data set is estimated tickets sold (top 10 gross divided by average ticket price) by day of the week from 2002 to 2013. Total estimated tickets is preferred, but since the top ten movies account for as much as 90% of total gross, most courts would say it’s a pretty solid dataset for industry trends. Here are the results:

CRI Post #2 Weekday Top 10

CRI Post #2 Weekend Top 10

Source: BoxOfficeMojo.com

Some takeaways:

-Ticket sales averaged lowest on Tuesdays. One of the most surprising trends is that over the last two years, ticket sales on Tuesdays have ticked up. It’s not clear if this is a shift from folks tired of wrestling with large weekend crowds.

-Ticket sales are highest on Saturdays, followed by Friday and Sunday.

-Monday and weekend attendance trends are fairly stable

-Tuesday, Wednesday, and Thursday trends show fluctuation and volatility (as confirmed by calculating standard deviation)

-Wednesday ticket sales have been the most volatile over the last decade.

If a dynamic pricing model does not cannibalize the preexisting audience, especially the weekend crew, a scheme focused on Tuesday-Thursday tickets for non-blockbuster event movies may capture extra demand. The goal and the big question remains: how can we grab those that hadn’t planned on going to the movie theater at all?

When was the last time you asked someone if they went to the movie theater?

And they answered, “I’ll wait for it to come out on Netflix or DVD?” They said, yes, right?

The big-screen experience remains an undeniably powerful way to tell stories. But occasional moviegoers defected by 72.2 million tickets from 2010 to 2012, which represents an estimated $575 million loss to the industry. This trend doesn’t seem to be going anywhere. Forecasts peg attendance will decline yet again in 2013.

box office 1995-2012

Full disclosure: I fell in love with movies after my first movie theater experience (the Lion King!). And I still believe there’s nothing better than watching a great movie in a great theater with a rapt audience.

But this fellowship project is not borne out of nostalgia.

the price is right

"The Price is Right" will test whether some type of dynamic ticket pricing for movie theaters can grow attendance, particularly for non-blockbuster event films. In addition, can pushing this over a mobile platform capture extra demand by drawing in people with no previous intention of going to the movie?

The first phase is to gather as much research data as necessary and define the scope and parameters of the test. To achieve this, we will seek out partners and collaborators interested in working with us or sharing their thoughts on the project. This includes exhibitors, distributors, online movie websites, academics, mobile app and tech companies, and other non-entertainment firms with dynamic pricing experience.

The motivation is to answer: Is there any more money left on the table for movie theaters? Is this implosion inevitable? Is my generation permanently attached to on-demand viewing habits? The deck is certainly stacked: there’s fierce competition from other types of media, customer dissatisfaction with the movie theater experience, long-standing industry practices between exhibitors and distributors, and a fear that any change will exacerbate the problem.

If the answer is, “No, I want to watch this movie in my sweatpants and barring the apocalypse, I am staying put,” that is ok. An answer from a real-world, real-time test is good. It’d be even better to move the needle in a meaningful way.