Something the CEO of McDonald's said about rewards doesn't add up.

I thought this would be a short post, but turned out to be complex and longer than I hoped... As part of my ongoing fascination with McDonald's and what impact their rewards program has on overall results, I note they announced Q2 2025 earnings on 8/6 (https://corporate.mcdonalds.com/content/dam/sites/corp/nfl/pdf/2025%20Q2%20Earnings%20Release.pdf).

Good news! Reversing a recent negative trend, they reported an increase in overall revenue and same-store-sales growth of 2.5% in the U.S., with the CEO, Chris Kempczinski attributing the increases to "...the power of compelling value, standout marketing, and menu innovation". He referenced the $5 meal deal, McValue, and the re-introduction of the Snack Wrap as important drivers, although the Snack Wrap really didn't impact Q2. CNBC said (https://www.cnbc.com/2025/08/06/mcdonalds-mcd-q2-2025-earnings.html)

"This quarter, the burger chain’s U.S. sales received a boost from a tie-in meal with the “Minecraft” movie and the launch of the McCrispy Chicken Strips."

Did anyone mention the impact of loyalty? "We’re seeing that loyalty program drives significant increases in frequency." Any specifics? Yes, finally! They've been very cagey. Here is what the CEO said:

"In The US alone, on average, the same customer visits 10 and a half times in the year before joining the loyalty program and then 26 times in the year after joining."

Whoa. That's a huge number. Does it make sense? I don't think so...

This is somewhat complex to follow, but if you assume that the same numbers apply in all markets and 100% of the increase in sales is due to the loyalty program, jumping from 10 to 26 visits the year after joining is not possible. I've come at it three different ways and always end up with the same result.

From McD's Q4 earnings we get the following:

2023 2024

Systemwide Sales 125.3B 130.7B

% of Sales from Loyalty Members 16% 23%

And we can define total revenue as the sum of member and non-member revenue, as follows:

Revenue = % of non-members x customer base x AOV x visit frequency

+ % of members x customer base x AOV x visit frequency

Using this formula, data from quarterly earnings reports and the CEO's frequency numbers as input, the worksheet below shows what we should expect in terms of visits and visitors, using the $10 average check number that is frequently cited. The key data points are % of revenue attributable to the rewards program.

One problem - these calculations suggest that even though overall sales increased 4.3%, if members visit 26 times after joining the rewards program, given the increased share of sales attributable to loyalty members, non-member sales would necessarily have to decrease 4.4% and total visitors would have to drop by 0.5% to make total sales add up, meaning that new program members would be crowding out non-members. That doesn't make sense - behavior of non-members should not be impacted by change in behavior of members.

Using this same worksheet, if you were to reduce the member visit frequency from 26 to 23, the number of overall visitors goes to flat from 2023 to 2024, with the entirety of the sales increase attributable to new rewards program members and their increased frequency. Although that still suggests a drop in non-member visits. One more wrinkle - U.S. same-store-sales increase was only 2.5%, not 4.3%. So if we drop that down, the model projects a larger 2.2% drop in visitors. Again, not realistic, but requires using a member visit frequency of 15.75 instead of 26 to get to a flat number of visitors. One more twist - if you assume only half of the increase in sales was due to the rewards program, and the other half was due to promotional activity and menu innovation, using 1.25% growth requires rewards program visit frequency of 12.75 to get to a flat number of visitors.

Now, this is based on a number of assumptions, and they don't report numbers specifically for the U.S. market for systemwide sales, % of sales from rewards, etc. (If we apply the 40% number that is generally used for U.S. contribution to overall sales, we get similar results.) So at best it is directional, but suggests the frequency increase from joining the rewards program can't be as high as 26 visits the year after joining, vs 10 the year prior. In fact, it is more likely closer to 12 or 13 visits the year after joining. That seems more realistic - I'm a member of this (and most other) programs, but there is nothing compelling about this program that would encourage me to increase my visits by 2.6 times.

Quick Update:

In the original model shown above I used systemwide sales and 23% as the share of sales from rewards program members, because that was what was quoted in their earnings release. Digging into the earnings call transcript, I saw that the CEO said that in the U.S., rewards program members account for "...roughly a quarter of our business".

So below is a rework of the model using numbers just for the U.S., assuming 41% of systemwide sales take place in the U.S. The higher percentage of sales from program members actually makes his frequency numbers even more unrealistic. Meaning that if true, there would be an even more significant decline in non-member visitors, a loss of 8MM visitors who stopped visiting between 2023 and 2024, above the number who converted to members. The bigger issue, which only becomes clear once you limit results to the U.S., is that the required number of visitors to achieve reported revenue is greater than the U.S. population. This means non-member frequency can't be as low as 10.5 annual visits.

Per my adjacent piece putting McDonald's earnings report in the context of the recent Numerator research (https://www.linkedin.com/feed/update/urn:li:activity:7359640334328373249/), the frequency for non-members has to be much higher than 10.5. If 25% of revenue comes from members, visit frequency of 29.75 for non-members and 32 for non-members, an increase of 7.6%, makes the numbers work - a far cry from incremental sales of 250%.