Case Study: Food and Beverage Pricing At Smith Fun Centers

Recommending a new pricing approach focused on growing food and beverage spend at a well-established family entertainment center chain.

Executive Summary

Smith Fun Centers (SFC) is a chain of family entertainment centers with a need to grow food and beverage spend at their centers. Technology and dated pricing strategies have locked down management's ability to change prices in any targeted manner.

Both quick wins and long term solutions can drive Smith Fun Centers growth by double digits. The key facets are price elasticity estimations and a pricing lifecycle approach. These tools set the company up for success by allowing it to learn from its past pricing changes.

Background

Note: Actual names and financial figures have been anonymized in this case study out of respect for the privacy of companies involved.

SFC is a well-established chain of family entertainment centers across the United States, with over 40 centers in over a dozen states. These centers are managed centrally by a headquarters office, and regional managers oversee multiple centers which range in age from less than a year old to nearly two dozen years old. Each center also has a general manager that oversees operations and has a say in pricing and promotions.

Each center offers a combination of arcade games, bowling, billiards, virtual reality games, laser tag, mini golf, ropes courses, and/or karaoke. Additionally, locations often feature event space for corporate or birthday party settings, as well as a full bar, a table-service restaurant, and grab-and-go food. Most of the cuisine centers around classic American comfort food, such as pizza, burgers, fried chicken, nachos, fries, shakes, and more.

SFC is in the middle of a creative food and beverage (F&B) overhaul that involves rebranding both the grab-and-go and table service offerings in centers across the US. This menu overhaul is ongoing, and several centers are amidst a test to measure the popularity of new recipes, the branding of the changes, and the compatibility with kitchen operations.

Today, centers are bucketed into one of three groups, or price tiers, which determines the price of every F&B item up and down the menu. See the Table 1 below for an example of how this plays out across the business.

	Tier 1	Tier 2	Tier 3
Number of Centers	20	10	15
Hot Dog	$3.99	$4.49	$4.99
Cheeseburger	​$5.99	$6.49	$6.75
Fries	$1.99	$2.99	$3.99

Table 1: Example Menu Price Tiering Across SFC

Additionally, SFC has noted that in recent years there was an initiative to move from two price tiers to three. This effort was purportedly expensive, time-consuming, and difficult to execute in stores where prices changed. In initial discussions with their teams, it was clear that further efforts to add price tiers or otherwise strain their technology system would be difficult.

In summary, SFC is seeking revenue growth through food and beverage pricing changes. Today, SFC stratifies centers' food and beverage pricing across three price tiers today. The SFC culinary team is amidst a menu overhaul, so recommendations on specific items may or may not impact revenues in centers with the new menu.

Evaluation

SFC required a robust assessment of its F&B pricing strategy as a part of a broader project to optimize revenues across their centers. Before recommending changes, there was a need to better understand and valuate the current pricing methodology.

First, note that SFC has started to stratify their F&B prices through the center pricing tiers. This shows that SFC understand that willingness to pay is not the same at their various centers across the US. Customers at rural stores may not be willing to pay $4.99 for a hot dog, while $3.99 feels much more approachable, for example. This tactic is an effective one when fully realized, but is it fully realized at SFC yet?

While this is a step in the right direction, the three national pricing tiers limit SFC's ability to adjust pricing in any targeted way. What happens when SFC wants to grow revenue in the 20 Tier 1 centers? They'll look to marketing, promotions for the core product, and ancillary (think food and merchandise) pricing. Their finance teams might review the menu items and find that the Fries feels like the best opportunity, which they raise from $1.99 to $2.29.

While this feels like a good price change off the cuff, in reality this gets complicated quickly. There are 20 centers that follow Tier 1 pricing. These locations span several different states, new buildings and old, and centers where food is popular and centers where the food sales are lagging. This means that consumer behavior varies across these locations. When consumer behavior varies, so does willingness to pay, which is what ultimately results in a sale.

Now SFC has increased the price of Fries at 20 locations. Some of these centers will likely see success, because the customers continue to purchase at a higher price, while others suffer because customers scoffed at the new price tag. The same challenge arises if SFC wants to move a center from a lower price tier to a higher one. Perhaps the Hot Dog and Cheeseburger customers at a given center are willing to pay a higher price but the Fries customers are not. Management does want want one success was washed out by other failures.

Willingness to pay is something that can be measured through sales transactions, given enough history and variances in price. Price elasticity, or price sensitivity, can be estimated using the following formula for a set period in time:

Change in Units Sold

------------------------------ = Price Elasticity

Change in Price

Price elasticity can be intimidating mathematically, so the below example uses visuals to tell the story. Say a center in Texas sells three food items, A, B, and C. The regional manager calls the general manager (GM) for this Texas center and gives them revenue growth goals. The regional manager recommends achieving these goals through a number of methods, including pricing.

The Texas GM has no data to work with other than what's shown in Image 1 below. All three items are priced the same (P), and each has the same weekly sales (Q). Revenue is equal to P * Q, and is represented by the areas of the rectangles in the images below.

Image 1: Base food sales example at a Texas center

The GM looks at this and decides to increase prices on all three items. Unsure of exactly how much to increase each item, the GM decides to increase all three prices by the same nominal amount. After four weeks of sales at the new prices, the GM's team comes back with the results, see Image 2 below.

Image 2: Sales from four weeks of sales at a higher price

Whenever price is increased in a food or retail setting, a drop in units sold is expected. However, this Texas center saw a small drop in sales for item C, a moderate drop for item B, and a large drop for item A. Did this result in net revenue growth for the GM?

The revenue before the change is the same for all three items. Look at the colored shading in Image 3, which visualized the changes in revenue after the price increases. Revenue equals price (P) multiplied by units sold (Q), so these charts visualize the change in revenue via the area of the rectangles.

Image 3: Revenue growth is shown in green (above) and loss in red (to the right)

A quick way to assess the success of these price changes is to compare the size of the shaded rectangles for each item. The upper, green rectangle represents the growth in revenue from customers purchasing at the higher price. The lower, red rectangle represents the loss in revenue from fewer units sold. If the green rectangle is larger than the red one, the price increase was successful for the time period measured.

Looking at Image 3, Item A's red rectangle is clearly larger than the green, resulting in a loss. Item B's rectangles are about the same size, this price increase was likely a wash. Item C is the only item that appears to have had a successful price change. Why is that?

Well, the most important information that the GM needed was hidden from them: the price elasticity of these products. Take a look at Image 4, where curves represent the price elasticity of demand (e) for these three products. The more flat a curve, the more elastic the product: customers will react negatively to price increases. The more vertical a curve, the less elastic the product: customers will react significantly less to price increases.

Image 4: Changes in item sales with elasticities shown

This information would have been key for the Texas GM to have before making price changes. If they had known ahead of time, they would have made different decisions on the price increases. With the elasticity information ahead of time, the following changes may have been more effective:

• Increase the price of Item C by the same amount as in the example

• Hold Item B at its current price

• Consider lowering the price of Item A through a discounting or promotional tool

---

The example above focuses on item-level elasticity. Store-level, or center-level in SFC's case, elasticity works the same way, but is broader and represents the average for all customers at the center. Item elasticity is more precise, but can be difficult to chase down if there is not enough history in the data. In the absence of item elasticity, center elasticity will give pricing managers a direction to start in.

Working with the data available from SFC at the time, there was enough information to estimate the price elasticity of most centers, but not all. This information was compared against the current price tiers to assess the effectiveness of the current pricing structure. When lined up against the pricing tiers, centers spanned low, medium, and high price elasticity in each tier, as seen in Image 5. Note that not all centers could be measured due to lack of data due to center age, variances in pandemic closures, and more.

Image 5: Center-level elasticities by price tier with annual F&B revenue

First note that this is an expected spectrum of elasticities for this kind of pricing setup. To see all Tier 1 centers as highly sensitivity would be shocking, for example, and SFC would have an issue on their hands. When charted this way, the reality sets in that high sales does not equate to low price sensitivity. Certain items sell well, certain centers do good business, but that information alone does not guarantee that price increases will be successful for those items or centers.

Proposed Solutions

Locking centers into price tiers clearly creates a challenge for the teams in charge of pricing food and beverage at SFC. The best solution is for SFC to eliminate the price tiers, but as shown in the above walkthrough of price elasticity, that's only part of the answer. Also, keep in mind that there isn't an appetite for individual menu item price change recommendations while the creative overhaul is taking place. So, both quick wins and long-term solutions are needed for SFC.

First, there are some quick wins that will drive revenue in the short-term for SFC while long-term solutions are implemented. These quick solutions rely on the comparison of center-level elasticities within the price tiers. Consider the centers in Tiers 1 and 2 that are lowest in sensitivity. Those centers are so insensitive to price changes that there is a chance they can withstand an otherwise jarring price change.

That chance can be measured by using a similar methodology to the price elasticity charts shown in images 1-4. From the discovery work in SFC's data, there are centers' price elasticity (e), the average transaction revenue (P) for a given center or the average center within a given tier, and the number of transactions over a year (Q).

Input e and a new price, P2, into the elasticity formula and the output is an estimated Q2. The difference between Q1 and Q2 * P1 is the volume risk from the price increase, think the red box from the example above. When pitted against the increase in revenue, (P2 - P1)*Q2, the result is the net estimated impact from moving a center from one price tier to another.

After running this estimation for each center, a minimum risk threshold can be set. All centers in price tiers 1 and 2, with low sensitivity that beat this threshold can safely move up to the next tier. This assessment at SFC found about 13 centers that could increase price tiers safely, which was executed to great success in the months following these recommendations. The impact was estimated to drive millions in incremental revenue for SFC, a several ppt growth in EBITA.

---

Next, and arguably more importantly, how should SFC move forward with a new pricing strategy for the long-term? This recommendation was two-fold. First, eliminate the pricing tiers by "unlocking" pricing at each center, and changing prices on a case-by-case basis. Second, create a team to manage a pricing lifecycle for these unlocked prices.

Recall that a major hurdle to simply "eliminating" the price tiers was technology, time, and effort. What all of that boils down to is costs. Purchasing new technology has a cost, spending labor hours on a new system has a cost, and focusing energy on nit-picky price changes all but guarantees the new system will not save time.

The recommended change was framed in a way that showed a huge return on investment for SFC. Even with conservative estimates, the technology and time costs would have been somewhere in the low eight-figure range for a company the size of SFC. However, by using similar math to the examples above, and eliminating price changes that result in losses for SFC, the return on investment could be shown to be paid off within two-three years of implementation.

What does a "pricing lifecycle" look like though? This approach is a proven method of iterative pricing for a large F&B system through four simple steps: analyze, execute, measure, and repeat. See Image 6 below for an example of how this might play out over the course of three months. It's perfectly normal to vary this structure some to make it work well with your business.

Analyze the sales data to gather prices, count transactions, and estimate elasticities. A pricing manager can use this information to run simulations of price increases, estimating the net impact of the change for every menu item. They then prioritize the least risky, most impactful changes for recommendations. Set a cutoff for the percent of items that will change at a single time, somewhere in the range 5-15%, and shoot for low increases. Pricing managers should review these recommended changes with key stakeholders before executing.

The next step can be difficult for many organizations. After landing on the recommended changes, use technology systems to put them into place. Then wait. Wait at least six weeks, ideally 10-12. This step is crucial because the initial impacts of price changes often appear to lead to loss, especially in the first four weeks. During that time, work on preparing for step three.

Image 6: A three-month example of the pricing lifecycle

Third, measure the results of the recommended changes. Experiment with the right measurement methodology for your business, a double-delta method is a good starting point if possible. Try to keep the methods as objective as possible, and be clear about assumptions and gaps in knowledge. Present the findings to stakeholders, note all the learnings and move to step four.

Finally, repeat! Go back to step one, using all of the learnings generated in step three. Update methodologies, and continue to grow and learn with each iteration. Importantly, the low percent of the total menu that was changed in the first cycle means there should still be plenty of opportunity in the next cycle. Effectively, the same store-item will likely not have its price changed more than once a year, because there will always be other centers and other areas of the menu to visit.

Conclusions

SFC had some hurdles to get over, but pricing opportunity was still on the table. Using data to guide the decision-making, both quick wins and a long-term solution are estimated to grow revenue by double digits year-over-year.

While the current price stratification is a good step in the right direction, locking centers into pricing tiers creates situations where increases might be just as likely to cause losses as they are to cause growth. Unlock the pricing and take changes case-by-case to find the best opportunities to increase price. Avoid risky, sensitive items to protect against loss.

Finally, use a pricing lifecycle to identify those opportunities and risks. This proven method uses a breathing tool that focuses on the math, and is importantly a learning tool for your teams. A price test on an item thought to be highly sensitive in one center creates a low-risk learning environment. If the price increase fails, the financial loss is insignificant in the bigger picture. Compare this to a test for a single item at 20 centers in the same price tier and the risk skyrockets.

Pricing changes should create growth without risking the hard-earned transactions generated by marketing and operations teams. By working together, being open to learning, and thinking outside the box, many businesses can grow profitability through pricing.