Next Generation Ticket Pricing

Well, we’re 14 months into this pandemic and while a return to full-attendance live events is on the horizon in some parts of the world, most of North America is still somewhere between limited attendance and no fans, and here in Toronto I can’t even go to the damn golf course let alone a crowded stadium.

So while the ticket buying/selling ecosystem is mostly on hold, let’s talk about how it could be improved when (if??) it comes back.

Background

Dynamic ticket pricing is gradually gaining traction in the sports industry. As a general concept, the purpose of dynamic pricing is to vary ticket prices in accordance with anticipated customer demand. Of all sports, baseball seems to be the most advanced with respect to dynamic pricing. However, applications of this concept in Major League Baseball stadiums have focused on identifying differences in demand from game to game using factors such as the day of the week and the opponent. This is useful, but to truly optimize revenues one must also consider differences in demand from seat to seat.

Many businesses use operating models that are antiquated in the modern age of computers and analytics, often without realizing or even thinking about it. The grouping of seats into price tiers is an example of one of those models – its simplicity was required in 1991 but is unnecessary in 2021. A consequence of price tiering is that demand is highly variable among different seats in the same tier. The best seats in the tier are snapped up in seconds by insiders, bots or people with good luck. These seats are underpriced relative to the amount that the market would be willing to pay – resulting in excess demand and “economic surplus” gained by the lucky few at the expense of the seller; in layperson’s terms, money left on the table. The worst seats in the tier are excess supply – sold to late buyers or not sold at all.

The proposal is to eliminate price tiers and find a unique, optimal price for each row in each section. The simplest application will be single-game tickets, but the model could also be extended to season tickets and flex packs.

Analytical Approach

The proposed approach is as follows:

Step 1: Sort every row in every section in order from highest demand to lowest demand, independent of price. This can be done by analyzing historical occupancy rates for each section/row and, if available, aggregated data from secondary markets such as StubHub.

Step 2: Estimate a “demand curve” for each section/row; that is, the projected probability of each seat being sold for any hypothetical price. This can be done by looking at several years of historical ticket sales data.

Step 3: For each section/row, find the price that optimizes revenue using the results of steps 1 and 2 as well as an assessment of the expected “ancillary spend” (concessions, souvenirs, etc) associated with each seat.

Example

The following example is intended to illustrate the concept and approach. All figures are purely hypothetical and would be replaced with actual outcomes of the analysis described above.

Potential Benefits

In addition to the obvious revenue maximization, there are other benefits that could potentially be realized:

• Instead of the best seats in each tier being sold in the first minute, seats would sell more evenly throughout the stadium over time. This would result in enhanced availability over time of seats in a wide range of price levels. For example, suppose a corporate executive is entertaining some clients from out of town for a game next week. They want premium seats and are willing to pay a premium price. In the current system they would be forced to either settle for inferior seats or buy from a secondary market – the best seats would have long since sold out. In the proposed system, the higher pricing of the best seats would ensure that they are available for a longer period of time. 
• As seats sell more evenly throughout the stadium, there would no longer be a clustering of full sections of the stadium and empty sections of the stadium. This would reduce crowding for concessions, restrooms etc and it would give the stadium a fuller look on television.
• This can easily sit on top of any existing game-by-game dynamic pricing systems. The calculated optimal pricing for each section/row can be modified by a factor reflecting dynamic pricing elements such as day of week, opponent, etc. This factor could even change over time as tickets outsell or undersell expectations, and it could differ by section/row – for example, the best 100 level seats on weeknights are presumably occupied by corporate customers who may care less about the opponent than a casual fan in the 500 level may.
• I’m oversimplifying here, but in the best case scenario the upside looks like “the secondary market is put out of business, and you as the primary market gain all of their revenue. Oh and by revenue I don’t just mean the sales commissions but also all of the markup on re-sold tickets.”

Implementation Considerations

Software for box office computers, self-service ticket kiosks and online ticket portals may need to be upgraded for this functionality depending on current capabilities.

Box office staff will have to be trained to operate in the new pricing environment. They would need to ask questions such as “how much would you like to spend for your tickets today?” and it would be helpful to have two-way screens that facilitate an interactive discussion with the customer. A side benefit here may be the facilitation of “upsells” – customers may be persuaded to pay a little bit more than planned if they can see a tangible benefit to the additional spend (in the form of upgraded seats).

There will have to be some customer education, and as with any major change in pricing approach there is the risk of initial resistance and/or negative publicity. PR management will be critical.

What Better Time Than Now?

Post-pandemic, there will likely be a surge of pent-up demand for live event tickets that has never been seen in history. That means a LOT of economic surplus on the table just waiting to be captured by somebody.

Any readers in the ticket industry who want to explore this, you know where to find me. You only have a limited amount of time though before StubHub has me whacked…

Learn Analytics for Sports Business Decisions

This article is a sports business example of the same analytical ideas used in betting markets: estimating demand, modeling uncertainty, optimizing prices, and using data to make better decisions. If you want to build the modeling and analytics foundation behind this kind of work, start with Art of Sports Betting Analytics. If you want a more applied path focused on data workflows, model building, and practical sports analytics projects, continue with Data Science for Sports Betting