Target practice

This piece has been edited since its first appearance, to reflect a math error a sharp-eyed reader spotted – you can read more about that here.

Steve asked, in response to my piece about robot restaurants

Let's start with how much money staff would have to be paid in order to remain consistently with a restaurant as a career, how much of a profit a restaurant would have to earn to remain in business for years, etc. Take these as given. Now back out to how much revenue that translates into, and into how much meals need to cost. How much of an increase in prices would we need? … I'm just trying to understand the scope of the problem. Is the problem of restaurant financing completely insoluble within a plausible range of customer prices? Or is it only soluble within a range of prices such that only the wealthy will be able to afford non-robot food?

His email was actually somewhat longer than that and touched on many aspects of restaurant operations (the full text with his other questions is reproduced in a footnote), but looking at the relationship between current and sustainable wages and prices seems like a good place to start. 

I’ve been meaning to write a piece working through what it might take, from a financial and business standpoint, to run what might be termed a regenerative restaurant. I’m borrowing this term from the farming world, where “regenerative agriculture” refers to a set of practices that  

not only “does no harm” to the land but actually improves it, using technologies that regenerate and revitalize the soil and the environment … ultimately leading to productive farms and healthy communities and economies … to increase food production, farmers’ income and especially, topsoil.

There’s no reason restaurants can’t operate with similar goals in mind — supporting their staff, neighborhood, business partners, and environment in ways that are holistic rather than purely financial. The idea is tantalizing, because independent restaurants, as a group, have significant economic and social reach. As one pundit put it:

For better or worse, we are an archipelago of tens of thousands of little islands spread across every corner of the country — literally the Indonesia of the American economy. But, it bears repeating — in every corner.

So here we are. Hold on for the ride. I want to say at the outset that I’m not a professional sociologist or economist, and I know some of the folks reading this are. I would very much welcome constructive critique of these numbers from social scientists, other hospitality professionals, or really anyone at all. I’ll publish all correspondence (unless you ask me not to). 

I’m going to start by looking at this question as it pertains to a neighborhood restaurant in NYC in 2019. If there’s interest, I’ll try to examine other scenarios — smaller cities, rural areas, fast casual places, high-end restaurants, etc. — in a future edition. 

“How much money staff would have to be paid in order to remain consistently with a restaurant as a career?” 

For someone to “remain consistently” with a restaurant, or in restaurant work, it has to pay enough to allow people to prepare for and afford some of the major life stage markers that society expects of us — long-term partnership, home ownership, child rearing, retirement. When my wife and I ran our restaurant, we had the great fortune to work with numerous talented, driven, and incredibly decent people who loved restaurants and believed in hospitality. Many ultimately decided to leave the industry because they were making a wage that sufficed at the time, but not one that really felt sustainable for them in the long term. 

How much money is this? This isn’t minimum wage, and it’s not even living wage. It’s a higher figure, a benchmark we don’t yet have an accepted term for (Stability wage? Dignity wage? Equitable wage? Comfortable wage?). Since we don’t even have a name for it, we don’t have a widely accepted way of calculating it (to the best of my knowledge anyway. Social science professionals, please point me to anything you know of). 

What we do have are a couple of different Living Wage Calculators on the internet. Both assume zero savings and zero discretionary spending, which is clearly not a situation most people would find acceptable in the long term. Per the technical documentation from the MIT version, “the living wage [returned by the MIT calculator] is perhaps better defined as a minimum subsistence wage for persons living in the United States.

So to get from Living Wage to something that allows someone to even imagine buying a home or having a child or paying a hospital bill, I’m going to apply a common personal finance rule of thumb, the 50/30/20 rule. This rule says that you should spend 50% of your post-tax income on necessities, 30% on discretionary spending, and save the other 20%. 

For this exercise, I’m going to take the living wage for a household with 2 working adults and 1 child as a baseline. In NYC, the MIT living wage calculator pegs post-tax income for this household at $59k. This means we need a net income of $118k, which, split between two earners, means a pre-tax income of approximately $76k for restaurant work in NYC to really be a solid lifelong career. 

So now we know what we need to be paying people at this hypothetical regenerative restaurant. But we can’t work backwards into revenue figures, and consequently the question of what a meal in a restaurant needs to cost, unless we know how many people work here.  

The closest restaurant to my apartment in Harlem is a pretty typical NYC boîte. It has about 60 seats split between a dining room and a bar, and occupies probably 1200sf at ground level, maybe a bit more. There are mussels and a burger and frisee aux lardons on the menu, and they are clearly trying to conjure up what they think Americans think a French bistro is like. If you’ve lived in NYC for any length of time at all, you’ve probably been to a place like this. 

The last time I went was on a Tuesday night. There were 10 people working that night, plus a delivery rider. This is what we’d call a staff:seat ratio (remember this concept) of 1:6 — enough to deliver pretty basic levels of food and service. In a dive bar where you get your own beer and there’s one cook flipping burgers you might see a ratio of closer to 1:10. At a slightly nicer restaurant, one where the burger is actually cooked to the doneness you ask for, or where servers check in on you without being flagged down, it’s probably more like 1:5. At a place with 2 or 3 Michelin stars, it’s somewhere between 1:2 and 1:0.5. 

This place near me is actually open all day, every day, but on the first pass, let’s assume that they do only dinner, 7 nights a week. Given a 5-day workweek, and assuming that they need to add a person or two to handle a busy night, they’ll need 16 full-time employees (FTEs) to cover dinner shifts. The detailed math for this is in a footnote* below if you’d like to read it. 

Now we have 16 employees who need to make at least $76k a year, for a base wage bill of $1.216M, on top of which we need to add payroll tax and social security (together these are 11% of base wages) and health insurance for everyone (sustainable, remember?). Assuming health insurance is $5k/person, our total labor cost is now $1.43M.  

The most common benchmark for independent full-service restaurants is that Prime Cost — cost of goods sold (COGS) + cost of labor — should be around 65%. People used to assume that this would be split evenly between COGS and labor, but in the 20-teens, it looks much more like labor at 40% or more and COGS at 20-25%. Bear in mind that this is for a restaurant where tipping is practiced, so that 40% of revenue is just money that the house is paying workers, and doesn’t include tips.** Once you factor tips in, 18% of revenue being a reasonable estimate, this 60-seat restaurant would need to clear $2.465M a year in sales, or $47.4k a week (the math is in yet another footnote***, which also shows that running hospitality-included doesn’t actually shift the target by very much at all). 

Assuming we are in the happy position of filling, or “turning” every seat in the house twice a night on Sunday through Wednesday, and three times a night on Friday and Saturday, this means we’re filling the dining room 16 times a week, which means we need an average check of just about $50 per person. 

There are two important caveats about this estimate that may not be obvious to the lay reader.

First, 16 turns a week is really a pretty busy restaurant. We’re not talking “four-star review in the NYT last week” busy, but this is certainly a very popular neighborhood spot, one that draws people from outside the immediate area. By definition, only a small number of restaurants see this level of traffic. (I am very open to a hospitality professional who’s more familiar with the NYC market correcting me on this.)

Second, this $50 figure (which let us remember, is sketched in with a fat grease pencil on the back of a napkin) assumes that everyone in your 16 turns is there for the full ride — drinks and dinner. In truth, especially in a place like this, a substantial number of people come in for just drinks, or drinks and appetizers, or dessert, or whatever. Basically sitting at your tables but not eating a full dinner. 

So this $50-per-person price point is actually pretty optimistic, because one of two things is true: either the restaurant is popular enough to be not just full, but full of people eating a full dinner; or, a full dinner costs more than $50, so that the many tables not eating a full dinner still leave the average at $50. 

Given the current menu at my neighborhood “French” bistro, I would guess that their average check for someone eating dinner is between $40 and $50. So the very condensed, first pass, one-example discussion tells us this: in the most optimistic scenario, and if they are paying everyone on staff the same wage, they just might be able to pay all their staff what they need to live on with their current business model. In truth, prices should probably rise – by how much is dependent on how far they are from having 16 full turns of dinner a week, but I would be surprised if it were less than 20%.

Is this a “plausible range of consumer prices”? I don’t really know, but this boîte feels correctly priced for the neighborhood, a quite unfashionable part of Manhattan that is slowly becoming more fashionable. Pricing higher, even at $50 for a full dinner, seems to me the kind of pricing that would keep people away and sink them.

Once again, this is me working through a single example. Constructive input is very very welcome, and I’ll publish any feedback I receive.

* Here’s the math for the number of FTEs: 

7 dinner shifts a week x 10 people a shift = 70 shifts a week
Assume they need to add 1 person every Friday and Saturday = 72 shifts a week
72 shifts a week x 52 weeks a year = 3744 shifts
Assume each FTE works 240 shifts (50 working weeks of 5 days each, less sick and personal days)
3744 shifts / 240 = 15.6 FTEs, round that up to 16.

** I’ve railed against tipping in the past, but the fact is that most people in America still judge a restaurant’s price point assuming that they’re going to be tipping on top of the check, and I’m just rolling with this assumption so the mental picture is clearer. I still believe that hospitality-included is a better way to run.

*** For the restaurant in this example, knowing the labor cost, we can back into revenue using the following equation: 

Where R is revenue:
Labor cost of 1.43M = 0.4R (from the house) + 0.18R (from tips) = 0.58R

My conversations with people who run hospitality-included restaurants lead me to think that their labor costs converge on 50% of their revenue. It’s hard to give a clean number on what kind of premium they charge vs. their former pricing, but hospitality-included pricing is at least 15% higher than their former pricing, this makes sense, per the following equation:

Where R is revenue before the hospitality-included adjustment:
Labor is 0.5 x R x 1.15 = 0.57R

And here is the relevant portion of Steve’s email:

Let's start with how much money staff would have to be paid in order to remain consistently with a restaurant as a career, how much of a profit a restaurant would have to earn to remain in business for years, etc. Take these as given. Now back out to how much revenue that translates into, and into how much meals need to cost. How much of an increase in prices would we need? (Suppose we're talking about a standard fine-dining establishment, rather than a robot joint.)

I'm just trying to understand the scope of the problem. Is the problem of restaurant financing completely insoluble within a plausible range of customer prices? Or is it only soluble within a range of prices such that only the wealthy will be able to afford non-robot food?

Also makes me wonder whether the problem has to do with the price of real estate in major American cities. If so, I wonder whether large chains are the only places that will be able to make a consistent buck in large U.S. cities are those that can subsidize their expensive outlets with outlets in cheap locations. (Boise subsidizes Boston. Or something.)

One way out of this would be to make it easier to open restaurants in cities, make it easier for them to sell alcohol, etc. How much of the problem do you think that would solve?

Thank you for reading let them eat cake, a weekly newsletter about food systems and food. And as always, a super-special thank you to my pre-release readers, Jen Thompson and Diana Kudayarova. Jen, who is almost a professional social scientist, also proposed the “comfortable wage” term, so double super-special thanks this week!

This newsletter is free and a labor of love. I’m doing this to develop a platform as a writer, and if you like it, the best way to show your support is to send this along to someone who might enjoy it, or give it a shout out on social media, or simply click that little heart at the top of this email to like this post. You can find me @briocheactually on both twitter and instagram.



p.s. I’d love to buy you a coffee. Drop me a line!