As restaurants move more into delivery, catering and other new sales platforms, the sales per transaction amount typically rises. That will result in lower customer “traffic” that many restaurants report. This is to be expected. Investors must take a more detailed look at the same store sales components Running positive and ever-growing traffic, particularly in the US, may be too high of an expectation now. This is not to rationalize materially lower traffic; that is and always has been a negative indicator of brand deterioration.

The risk of misreading restaurant same-store sales results has increased as both restaurants and the consumer space has become more complicated.This involves not fully airing out restaurant sales, traffic and average ticket components.The risk of meaningful misreading is more consequential now that program trades are such a large portion of stock trading.^[1] Truly powerful restaurant holding companies and brands, like Olive Garden (DRI), Starbucks (SBUX), and Panera (PNRA) have all been affected; those working catering and delivery initiatives will be somewhat over time.

Restaurant same-store sales results can’t be read in isolation. The same-store sales comp is a very narrow reed upon which to make meaningful analytical calls. Something always happens this year or last to skew the effect: weather, holidays, promotional patterns, and foodborne outbreak tragedies, national or international effects all happen. More companies have been reporting the two-year comp, which might optically remove a prior year distortion.   

Must consider all of the restaurant sales components: Restaurant sales are multifaceted as the following chart shows:

What can make this even more complicated is that some restaurants report only their company operated unit sales. Many report both, but trends in the two subsets can move differently for a host of reasons. Some brands open new stores that are in the non- comps database but unfortunately cannibalize same-store sales at mature stores.

Of same-store sales, there are three major components^[2]:

Price is the weighted amount of menu price increases taken during a period. In recent years this has ranged app. 2% annually, but higher and lower in some brands. There is a line of logic that price increases can be too high, especially for discretionary consumer businesses like restaurants. We all want to go out to eat or get our food delivered, but we don’t have to.Once the price increase is estimated, then everything else becomes mix: new menu items added, more or less discounting, purchasing fewer items, etc.

There is no consistency how restaurants report traffic: There is no standardization in the industry how “customer traffic” is defined. Is “traffic” when one person comes in and dines inside? Or is it the “table”, when a group of people come in together. Is traffic one or four for a party of four? Some restaurant chains have held to their history of “party size” even though they could count people. One brand, Starbucks in the past was OK with store cashiers running up two “transactions” as a customer tried to maximize loyalty card stars. That has changed at Starbucks, but you guessed it, there is a distortion between now (fewer rings) and earlier (more rings). What about drive-thru and catering transactions? How are those counted?^[3]

Restaurants that Implement Delivery, Catering and Other Off Premise Strategies Will Have Higher Mix But May Record Lower traffic: Every restaurant is looking for more sales.Active restaurants that have added delivery platform strategies have added to online, digital and catering initiatives underway for years. All of these initiatives should increase the restaurant average check component. In some cases, these initiatives will decrease traffic, because logically a large catering order—or delivery order—would feed several people that would be taken out of the dine in” category. The idea is to use the “vacuum cleaner approach”—gather as many sales dollars as possible, through a variety of tactics.^[4] Said another way, there is only so much food away from home or brought in demand.

As delivery continues to build, this effect on traffic will continue. Restaurants can report a proxy value for traffic—using a catering or delivery conversion factor for example. But they might not because of tradition. Any US restaurant working delivery initiatives should see this higher mix effect over time. 

^[1]A few restaurant “shorts” around do nothing to help this matter and sow confusion purposely.

^[2]I’m blending into mixing the difference between gross sales ticket and net sales ticket, which could be a proxy for discounting changes. Some restaurants count the number of food units sold per transaction, which also would be useful, but almost never reported.

^[3]At this point, most restaurants report either a drive-thru or a catering transaction as just one “customer”.

^[4]Term as popularized by Panera CEO Ron Shaich, 2016.