Case Study 1 — Enlightened Hospitality and the Limits of One Operator

Union Square Hospitality Group and "Hospitality Included," 2015–2020

How to read this case. The events, dates, and the existence and ending of the program described here are matters of public record and widely reported at the time. Where this case discusses reasons — why guests reacted as they did, why certain staff left — it is summarizing what operators and industry press reported and what the mechanics make likely, and it says so. No financial results, turnover figures, or survey numbers are attributed to the company, because none have been verified for this book. Nothing here should be read as a judgment of the people involved; the point of the case is the structure, and the structure is instructive precisely because the operator involved is unusually good at this.

Background: an operator who put the argument in writing

Danny Meyer opened Union Square Cafe in New York in 1985 and built Union Square Hospitality Group around it. In 2006 he published Setting the Table, which is one of the handful of genuinely canonical books in this industry and which is on this book's reading list for a reason: it makes, explicitly and at length, the argument that Chapter 21 has been making in dollars.

Meyer's framing is what he calls enlightened hospitality, and its core is a priority order that sounds wrong the first time you hear it. The stakeholders, ranked: employees first, then guests, then community, then suppliers, then investors. Not because employees matter more than guests in some abstract moral sense, but because — the argument runs — the only reliable mechanism for producing genuine hospitality toward a guest is a staff that is itself well treated. Hospitality flows in one direction and it cannot be manufactured at the point of contact by people who are not receiving it.

That is the same claim this chapter makes at §21.7, and it is the claim Chapter 23 will pick up again. The difference is that Meyer put an operating company behind it.

By the mid-2010s USHG operated a portfolio that included fine dining, casual restaurants, and a fast-casual chain that had been spun out. It was, by common consent, one of the most admired hospitality organizations in the United States, and it had spent three decades building the career-path and internal-promotion machinery that most independents never build at all.

The operating issue: a pay gap the tip system could not close

The problem USHG set out to address is a structural one and it exists in most full-service restaurants in America, including yours.

A tipped server in a busy, high-check dining room can earn substantially more per hour than a skilled line cook working the same service. This is not because the restaurant decided it should be so. It is because two different compensation systems are operating in the same building: the dining room is paid partly by guests, directly, in a variable amount tied to sales; the kitchen is paid entirely by the employer, at a fixed hourly rate, out of a labor budget that is already the second-largest number in the business.

Three consequences follow, and every operator reading this has seen all three:

1. The kitchen's wage is capped by the P&L; the dining room's is not. Raising cook wages costs the restaurant money one-for-one. Raising server earnings costs the restaurant nothing, because the guest pays it.
2. The pay gap makes cross-training and promotion across the wall nearly impossible. A cook who would make a superb server can take the job and get a raise. A server who would make a superb sous chef would take a large pay cut to do it, so they don't, and the BOH loses a candidate.
3. In many jurisdictions, the employer legally could not simply share tips with the kitchen. For much of the period in question, wage-and-hour rules in most of the country restricted tip pools to customarily tipped employees. An operator who wanted to close the gap by redistribution frequently could not. (The federal picture on tip pooling with back-of-house has since changed in specific circumstances — Chapter 20 covers the current framework, and it varies by jurisdiction.)

So USHG chose the remaining door. In October 2015 it announced that it would eliminate tipping across its restaurants under a program called Hospitality Included, beginning with The Modern in November 2015 and rolling out to the rest of the group over the following years.

The mechanics were straightforward and are worth stating precisely, because the mechanics are the lesson:

• Menu prices rose — reportedly on the order of 20–25% — to fund what tips had funded.
• Tipping was eliminated, not made optional. The check said so.
• All revenue now flowed through the restaurant, which meant the employer could allocate it, including to the kitchen.
• Cook wages rose, and a share of revenue could be directed to positions that had never had access to it.
• Career paths and internal progression became easier to build, because a promotion no longer meant crossing an invisible compensation boundary.

Read that list against §21.7 of this chapter and notice what it is: a retention program. A wage ladder, a promotion path, and a redistribution of the reward for reliability — funded, at scale, by an operator who could actually afford to try it.

What happened: the parts that worked and the parts that didn't

The program ran for roughly five years across a growing share of the group. In July 2020, as USHG prepared to reopen following the COVID-19 shutdowns, the company announced it would return to tipping.

Two things need to be said about that ending, and the order matters.

First, the ending was heavily conditioned by the shutdown. Restaurants reopening in the second half of 2020 were rehiring staff into an environment of enormous uncertainty, reduced capacity, and compressed volumes. A no-tipping model concentrates compensation risk on the employer at exactly the moment an employer has the least ability to carry it, and it asks returning staff to accept a different earnings structure at exactly the moment they are least willing to. It is genuinely difficult to separate "the model did not work" from "nothing worked in 2020."

Second, the difficulties were real and were reported throughout the program's life, well before the pandemic. They cluster into three categories, and all three are structural rather than cultural — which is the whole point of including this case.

The guest-perception problem. A menu price that includes service reads as expensive next to a menu price that does not, even when the total the guest pays is identical. This is a well-understood pricing phenomenon and Chapter 10 covers it: guests compare the number on the page, not the number on the check. An operator adopting the model unilaterally takes the full perception penalty while competitors two blocks away keep the lower posted price.

The server-earnings problem. In a strong, high-volume dining room, the best servers on a tipping model can earn more than a fixed hourly rate will pay them — and those servers are portable. Reports throughout the period described exactly the pattern you would predict: some experienced dining-room staff left for tipped positions elsewhere. Note carefully what this means. The program's cost fell disproportionately on the group of employees who were doing best under the old system, which is the hardest possible group to move, and they had somewhere to go.

The one-operator problem, which is the largest. Tipping in the United States is not a company policy. It is a market-wide norm — a set of expectations held by guests, by staff, and by every competitor for both. A single operator who changes it is not changing a policy; they are attempting to move a norm from inside it, while paying the full transition cost alone and receiving none of the benefit that would come from everyone moving together.

That is a genuinely different kind of problem from the ones in §21.7, and it is why the case belongs in this chapter.

What it shows

One. Culture investments are real investments, and the good ones are structural. Everything USHG did here was a §21.7 lever taken seriously: close the gap between what the reliable and the visible get paid, build a path that crosses the FOH/BOH wall, make the reward for skill legible. The company did not run a pizza party. It restructured compensation. That is the correct instinct, and it is the one this chapter argues for.

Two. Some of the constraints on culture are not cultural. This is the case's most useful lesson and the hardest one for an idealistic operator to absorb. The BOH/FOH pay gap is not caused by management indifference. It is produced by the interaction of a tipping norm, the wage-and-hour rules governing tip pools, and guest price perception — three forces that sit entirely outside any one building. You can be the best employer in the country and still be unable to solve it alone.

Three. A retention lever that shifts money between employee groups is categorically harder than one that adds money. Every lever in §21.7 is additive: the schedule costs \$2,964, the ladder costs \$5,880, family meal costs \$5,096. Nobody loses. A no-tipping conversion is redistributive — it raises the floor by lowering a ceiling — and redistributive changes generate opposition from the people who were winning, who are also frequently your most skilled staff. Neither kind is wrong. They are simply different problems with different failure modes, and an operator should know which one they are attempting before they announce it.

Four. Reversing a decision well is a management skill. USHG announced the change, ran it for five years, learned things, and reversed it publicly. It is worth noticing that §21.5 of this chapter warns against unbuilding a decision because someone complained — and that this is not that. A decision revisited on evidence, after a real trial, announced openly, is the good version. The distinction between "I changed my mind because the data moved" and "I changed my mind because you pushed" is the entire difference between authority and a negotiating position.

The lesson

The lesson is not "no-tipping doesn't work." Some operations run it successfully, particularly in markets and formats where the norm is weaker, where check averages are high enough to absorb the posted-price penalty, or where a service charge with clear disclosure achieves a similar allocation with less guest friction. (Service charges are a different instrument with different legal consequences — Chapter 20 §20.4 is where that distinction lives, and it is a large one.)

The lesson is about scope. Separate the culture problems you can solve inside your building from the ones you cannot, and spend your money on the first category.

Inside your building: the schedule, the pre-shift, the ladder, who gets called at 3:40, whether reliability buys choice or just more work, whether anybody ever asks a current employee what would make them leave. Every one of those is fully within a single operator's control, none of them requires the market to move, and §21.7 prices the whole bundle at 1.27% of revenue.

Outside your building: the tipping norm, the wage floor, guest price expectations, and what your competitors pay. These are worth having a position on. They are not worth betting your only restaurant on unilaterally, in year one, while you are also trying to hold a 60% prime cost.

An operator with one 68-seat restaurant and a \$70,461 labor gap does not get to reform American compensation norms. They get to post the schedule fourteen days out. Do the second thing. It is available on Monday and it is priced in this chapter.

Discussion questions

1. USHG's stated stakeholder order puts employees ahead of guests. Argue the commercial case for that ordering using nothing but the arithmetic in Chapter 1 and Chapter 21 — no appeal to values.

2. The chapter classifies retention levers as either additive (nobody loses) or redistributive (someone's ceiling comes down to raise someone's floor). Sort the seven levers in §21.7 into those two categories. Which of them are you most likely to underestimate the difficulty of, and why?

3. A no-tipping conversion asks the highest earners in the dining room to accept less. §21.7 argues that your best employees want standards enforced more than your weak ones do. Are those two claims in tension? Resolve them.

4. Suppose you could adopt one piece of the Hospitality Included model at Bellwether without adopting the rest. Which piece, what would it cost, and what would it break?

5. The case argues that a decision reversed on evidence after a real trial is categorically different from a decision unbuilt because someone complained. Write the three tests you would apply, before reversing any decision, to tell which one you are doing.

6. Bellwether's chef-owner reads Setting the Table and proposes eliminating tipping in year one to close the BOH/FOH pay gap. You are the FOH partner. Write your response in under 200 words. It must take the proposal seriously, name the three structural obstacles from this case, and end with a specific alternative that is available in month two.