A couple of years ago, my parents and I ate dinner at a new coffee shop in my hometown of Kuching, Sarawak (population 540k; nothing ever seems to happen when you’re living there).
A typical Malaysian coffee shop works like this: an operator buys a store lot and fits it out with gas pipes and sinks and low walls to cater to food stalls. They then rent these stalls out to other food operators, who set up shop to sell their own dishes. A typical Malaysian or Singaporean coffee shop will thus have multiple food options run by different hawkers; the coffee shop owner will usually keep the drinks stall to themselves — drinks being, incidentally, the highest margin thing you can sell in the F&B business.
Anyway, my parents and I sat down at this new coffee shop and were immediately accosted by one of a small army of drinks waiters. “Minum?” the waiter asked, in Malay, brandishing an Android tablet.
We placed our orders and then I stared, perplexed, at the sheer number of wait staff in the coffee shop. A group of about five was clustered around the drinks stall; others were making their rounds — wiping down tables, or hovering near patrons who had just sat down, plasticky Android tablet in hand, waiting for new drink orders.
“What’s going on?” I said. By this point I understood the cost structure for a typical F&B business, and I knew that such a large wait staff was unusual for a coffee shop of this size. Sure, while a coffee shop operator was somewhat of a landlord (they collected rent from each stall operator, guaranteeing a somewhat stable revenue stream) — surely those rents, combined with the revenues from the drinks stall, weren’t so good so as to be able to justify such a large labour force?
“Oh,” my dad said, waving his hand dismissively at the small army, “Everyone knows this coffee shop is owned by loan sharks. It’s probably a money laundering operation.”
“Ahh,” I said, and then I realised that we had paid for the drink orders in cash, and that the wait staff were probably also paid in cash. High velocity cash businesses made for an ideal front if your goal is to launder money — “Ok” I said, “I guess that explains a lot.”
F&B Incentives Explain Behaviour
In our last (members-only) piece, we discussed the incentives that underpinned the F&B industry, as a way to explore Charlie Munger’s “all my life I’ve underestimated the power of incentives” aphorism. I walked you through the process that my old boss and I used to peel back the nature of the incentives in the restaurant business, and then we discussed how the cost structure of those businesses seem to explain a surprising number of observable properties about the F&B world — things like the difficulty of hiring, high staff turnover numbers, and the behaviours we got when we sold software to cafes and restaurants.
The flip side of that idea, though, is that whenever you find behaviour that deviates from a set of mainstream incentive-motivated behaviours, you should probably take a closer look. The story about the maybe-money-laundering-coffee shop above is one such example; anyone familiar with F&B business economics would have squinted suspiciously at the large number of wait staff hovering around (which meant disproportionately high labour costs). Either the coffee shop operator was a novice, or the business was a money laundering operation — I didn’t stay home long enough to find out.
But that is just one story. There are others. Here are two.
USHG and Enlightened Hospitality
Kevin Kwok has an essay titled Aligning Business Models to Markets that discusses Danny Meyer’s Union Square Hospitality Group (USHG). Meyer is most famous for running a number of incredibly successful restaurants, amongst them Union Square Cafe and the Michelin-starred The Modern. He also created Shake Shack, which is now a multi-billion dollar publicly traded company. In his autobiography, Setting the Table, Meyer opens with:
Over the course of the past twenty-one years I’ve opened and operated five white-tablecloth restaurants; an urban barbecue joint; a feel-good jazz club; a neo-roadside stand selling frozen custard, burgers, and hot dogs; three modern museum cafés; and an off-premises, restaurant-quality catering company. So far, I haven’t had the experience of closing any of them, and I pray I never will.
Meyer’s book makes a compelling, if simple argument: do hospitality well, get service really, really right, and you’ll have the secret to his success. But the truth is more complex. Kwok points out that providing excellent service in the F&B business works against the structure of the industry. For a variety of incentive-related reasons (which we discussed in the previous piece), talent is extremely difficult to hold on to in the restaurant business, much less train to provide good service. This in turn means that the typical F&B business will struggle to provide good service at scale.
So what about Meyer’s business allows him to do this?
USHG is a constellation of very different restaurants and chains. At one end it has Michelin star fine dining restaurants like The Modern and Gramercy Tavern. While at the other end it has the large chain Shake Shack. And many restaurants in between those two ends of the spectrum of pricing and scale.
Unlike many restaurant groups, this variety means Union Square Hospitality Group can hire people early in their careers — and plan for them to advance their careers from within USHG.
You can start working at Shake Shack, and then move on to managing their own Shake Shack or working in one of USHG’s more upscale restaurants. This is true both on the business or chef sides of the business.
If you do well you could go on to run a restaurant in USHG’s portfolio. Or if you wanted to open your own restaurant, you could open one with Danny Meyer as part of USHG — or start your own restaurant and have USHG as an early investor. In fact, another possibility is what the three Michelin star restaurant 11 Madison Park did. It was a USHG restaurant that they sold to its general manager and head chef, who’d both worked at USHG for years.
This is certainly a structural advantage, but it isn’t the only advantage that USHG enjoyed. From his first restaurant onwards, Meyer regarded quality service as a top priority, and he sought methods to spread a certain hospitality culture that was uniquely his.
To give you a taste of that culture, here are two anecdotes about Meyer’s restaurants from the book The Culture Code:
A young woman, recently moved to New York from the Midwest, took her parents out to dinner at 11 Madison Park to celebrate her new start in the big city—and to allay her parents’ fears about the difficulties of living in New York. Toward the end of dinner, as they looked over the dessert menu, the father pointed to a forty-two-dollar glass of dessert wine called Château d’Yquem and commented on how insanely expensive New York was. The waiter overheard the father’s comment and, moments later, reappeared carrying a bottle of Château d’Yquem and three glasses. The waiter said, “We are so grateful you came tonight. I heard you talking about the Château d’Yquem. This is one of the rarest and best dessert wines in the entire world, and we would love to offer you each a taste with our compliments.” A small explosion of surprise and delight ensued.
Then there was the time a dining companion of Nebraska senator Bob Kerrey found a beetle in his salad at Gramercy Tavern. The next day Kerrey and his friends were eating at another of Meyer’s restaurants. After they were seated, a salad arrived garnished with a small piece of paper on which the word Ringo was written. The waiter said, “Danny wanted to make sure you knew that Gramercy Tavern wasn’t the only one of his restaurants that’s willing to garnish your salad with a Beatle.”
Think about the systems and the service culture necessary for these two stories to have happened. The waiter in the first anecdote had to overhear the conversation between father and daughter, understand the dynamics at the table, and then had to be empowered to gift them the free bottle of wine. The wait staff in the second anecdote had to rely on an understanding of the customer across restaurants in the same F&B group. Neither were accidental.
Much of what we know of USHG’s early practices comes from NYU doctoral student Susan Salgado, who embedded herself in the organisation for six months during her dissertation. She concluded the following (in what has become a somewhat famous piece of work on organisational behaviour):
The results indicate that Union Square Cafe achieves its differentiation strategy of ‘enlightened hospitality’ through a synergistic set of human resource management practices involving three key practices: selection of employees based on emotional capabilities, respectful treatment of employees, and management through a simple set of rules that stimulate complex and intricate behaviours benefiting customers (emphasis mine).
To hear Salgado tell it, Meyer’s results come from a three legged stool — a set of self-reinforcing systems that could produce a culture with ‘enlightened hospitality’ as its output. If you believe Kwok’s observation that USHG’s business structure allows it to attract (and reward) top talent, then Salgado’s thesis explains how Meyer took advantage of that influx of talent: he reinforced it with a group identity that kept employees feeling nurtured and engaged, taught catch-phrases and heuristics to guide service behaviour, and designed a hiring process that weeded out customer-facing staff with the wrong set of emotional capabilities.
(This last bit is an extremely crude way of putting it — the onboarding process for a new service employee in USHG was actually designed to create familiarity with multiple co-workers in a single week, allowing the existing group culture to either accept — and rapidly nurture! — or reject problematic new hires).
I’ve covered Salgado’s work in another blog post, where I wrote:
After her MBA, Salgado offered to work full time for Meyer at USH. Her solution to this problem was to formalise all of Danny’s practices as a series of management training classes. There, they simulated difficult situations that managers would have to deal with on a day-to-day basis: how to reprimand an employee who kept making mistakes, how to pre-emptively notice a customer growing frustrated, and how to mediate the inevitable disputes between front-of-house staff and back-of-house kitchen crew. Formalising Danny’s intuitive practices helped USH scale their operations to an empire of incredibly successful restaurants.
And this was how Meyer systematised and scaled his hospitality culture.
To reiterate: whenever you find behaviour that deviates from a set of mainstream incentive-motivated behaviours, you should probably take a closer look. In this case, what you’ll find is a set of interlocking systems that together creates a rather different set of incentives — a set of incentives that, if Meyer would have you believe it — resulted in a remarkable restaurant business.
Singapore’s Food Court Operators
In the previous essay on F&B incentives, I wrote the following bit about the restauranteur worldview:
(...) No, given their business model, what most restauranteurs wanted were new (or repeat) customers. I mentioned earlier that when staff costs exceeded 30% of the previous night’s revenue, you know that you’ve either hired too many people or you’ve collected too little cash. And then I said that you have one of two choices: fire people, or increase revenue. We’ve talked about the former, so let’s now turn our attention to the latter.
Broadly speaking, in the restaurant business there are two ways to increase revenue: 1) increase order size (that is, revenue per customer), and 2) increase number of table turns per serving period (or sales velocity, if the business has no tables to speak of). The average restauranteur is highly motivated by both.
This explained a number of things that we eventually noticed. For instance, when we told restauranteurs “unlike a waiter, a self-service kiosk would always upsell dessert, and we know it increases average order size by $3-$5”, their eyes would light up; it hit on two pain points core to the F&B industry (that you cannot train properly with high employee churn, and you want to increase revenue per customer). And while the average restauranteur would not be motivated by inventory cost savings, we discovered that F&B stores with a central kitchen (e.g. bakery stalls in malls, or shops that sold steamed Chinese baos, say, who had to procure them pre-made from a central kitchen) were very receptive to our pitch that “the inventory system can warn your central kitchen that so-and-so outlet is running out of items to sell, which means you can resupply them, and the outlet can stay open for longer and not close prematurely.”
In other words: use our software to increase daily revenue / maximise inventory velocity! It was like music to their ears.
This pattern of behaviour held with nearly all of the F&B operators we sold Point of Sales systems to. It held across tiny cafes with zero seating, it held with new eateries at the heart of Singapore’s central business district, and it held with mid-sized Chinese restaurants, the type with a hundred or so tables spread across two locations.
But then we started talking to food court operators in Singapore, and our pitch no longer worked. It was the oddest thing. We would give our usual spiel about attracting more customers, or increasing order size, and they would nod and say “interesting interesting” but it was clear that nothing we said was of particular interest to them.
Food court chains were an interesting evolution of the coffee shop model that I’d described at the top of this essay. Beginning in the 80s, a number of coffee shop operators started expanding throughout Singapore — with the largest eventually owning 56 or so shops. These operators then pivoted into air-conditioned ‘food courts’ in the mid-to-late 90s, as Singaporean consumption shifted to glitzy malls across the island state. During my time with my old company, I never did figure out why the operators of these larger chains spurned most of our sales pitches — I figured that there was probably something in their business model that we weren’t seeing.
A couple of years later, long after I had left the company, my old boss told me he had figured it out. “These guys got to where they were by focusing on process efficiencies.” he said. “They are ridiculous process people — they run everything according to playbook, so even the cooks they hire have to follow exacting SOPs (standard operating procedures). They don’t think about their businesses the same way the rest of our F&B customers do.”
I had caught my old boss’s implied point about the training SOPs immediately — it meant that these businesses had found a solution for the F&B labour problem, by standardising around dead simple playbooks, tuned for unskilled labour. But it also hinted at other things — anyone who has eaten in Singapore’s food courts would likely have observed that the tables, chairs, cutlery and plates were standardised across all the outlets, and often branded with the chain’s logo. My old boss describing these businessmen as ‘extremely process oriented folk’ meant that the chain operators were running tight ships, and at scale.
“So you’re saying that they’re not motivated by increased business?”
“Well, they seem to be more focused on costs.” my old boss replied. This was remarkable to me: it stood in stark contrast to the typical F&B entrepreneur, who wouldn’t be particularly motivated by cost savings (again, for reasons that we explored in the previous piece). What my old boss left unsaid: Singapore’s food courts were all located in high traffic areas, like in malls, under office developments, and near train stations. Increasing table turns probably wasn’t the highest priority for these operators. But by then I was already working through some of the other implications of this observation.
I quickly learnt that Singaporean food court operators had a number of business model differences with the coffee shops that they had evolved from. The operators ran less like landlords and more like quasi-employers — for starters, they withheld all the cash paid by customers to stall-holders. The operators would then pool that money into a centralised fund, from which they would deduct administrative charges, fixed or variable rental costs, stall-holder’s supplier costs, stall-holder employee salaries, before then paying what was left of the proceeds to the stall-holders. That pool was treated as working capital — and would occasionally be used to invest for returns, pay for operating expenses, or fund capital expenditure. Some food court operators charged a variable amount as rent (20% of monthly revenue, for instance); others charged a fixed amount.
The food court operators also had an interesting dynamic with mall owners. Most restaurants suffer from something business maestro John Malone calls ‘wholesale transfer pricing power’ — that is, they lived at the mercy of their landlords. A great food court with a good brand, on the other hand, could attract foot traffic to a mall (and vice versa) so there was more of a symbiotic relationship between a food court and a mall than between a restaurant and their landlord than you might expect.
And in fact some of the food court operators had set up real estate investment arms in order to stake out new sites for expansion. Starting in the late 2000s, Singaporean food court operator Koufu began doing property investments out of a subsidiary, eventually winning mall management contracts for Punggol Plaza and Fernvale Point. Other competitors bought stakes in property development companies; the competition shifted from getting early dibs on mall sites to denying access to mall sites to the competition.
These location plays weren’t by accident. So long as they picked good locations with high footfall, these food court operators could guarantee a certain level of attraction to potential stall-holders, a certain level of table turns, and therefore a certain level of returns.
So why were these operators so concerned with costs? Well, at this point, the answer might be obvious: if you consistently picked good locations, you could take high table turns for granted, which in turn meant that you would have to squeeze most of your profit improvements out of cost reductions. So it seemed like my old boss was right — the operators that won did so thanks to process efficiencies. (The ones that survived also had methods to deal with the cyclical nature of the business, but that’s a separate story). In sum: these businesspeople responded to incentives that emerged from scale.
And what an array of cost reduction options they had! Unlike small coffee shops and individual F&B outlets, food court operators could demand lower ingredient costs on behalf of their stall-holders. Operators would also run centralised cleaners to clear tables and wash utensils, which meant a single SOP for onboarding and training cleaners, which they could then deploy across all of their properties. And, as I’ve mentioned earlier, the operators also standardised on common furniture and cutlery, with some going so far as to demand food stall operators wear standard uniforms. And I learnt that nearly all the major food court operators in Singapore ran central kitchens today, in order to wring out further process efficiencies for soup bases and ingredient prep and so on.
No wonder our sales pitches fell flat with this demographic: their incentive structures were totally different.
“What are you going to do?” I asked my old boss, near the end of our meeting.
“I don’t know,” he said. “They’re a tough market to crack. But we’ll figure something out.”
And so it goes.
So what have we covered? In our previous piece, we discussed what it was like when you peeled back the incentives for one particular industry, and then worked out all the second and third order implications that unfolded from those incentives. In that essay I was mostly writing from the perspective of a vendor who had to figure out what motivated these businessmen, in order to go sell to them. But it turns out that you can do one other thing once you’ve grokked the incentives for a particular industry — you can look for aberrations in typical behaviour, and then treat those aberrations as signals to dig into the incentives that motivated these companies to do as they do.
Occasionally, what you’ll find at the other end of that variant behaviour is a money laundering operation. Other times you’ll get a wonderfully designed set of interlocking incentives. But in just about every case I’ve seen, you’re guaranteed to find a very interesting organisation indeed.
A major source for this piece was Koufu: Fortune of an Entrepreneurial Singapore Hawker, but I ran a number of facts by F&B friends and acquaintances. As always, all errors are mine; reach out if you see anything you’d like corrected.