Great Service Is an Operations Problem

Whenever a venue has a service problem, the reflex is the same: more training. Bring in a trainer, run a workshop, re-do the steps of service, print new cards for the team to carry in their aprons. I’ve watched owners spend real money on this and get nothing, and then conclude they hired the wrong people. Usually they didn’t. They tried to train their way out of an operations problem, and that rarely works.

Most bad service isn’t a knowledge failure. The staff usually know what good looks like. It’s a system failure. The operation is quietly making good service impossible, and no amount of coaching overrides a broken system.

The waiter isn’t the problem; the section is

Picture a server with a table waiting on water. She knows they need water. She’s not lazy and she’s not undertrained. She physically cannot get there, because she’s carrying too many tables, the station glasses are stored on the wrong side of the room, the ice well is empty because nobody set par for a busy night, and the POS terminal she needs is being hogged by someone sending a complicated order. Every one of those is an operational decision made by someone else, before she clocked in. She’s just the person standing where all of it becomes visible to the guest.

You can send that server on every hospitality course in the city and the table still waits for water on Saturday, because the problem was never in her head. It was in the section size, the layout, the par levels, and the terminal count. Fix those and service improves overnight with the same people you were about to blame.

This is the thing I try to get owners to see. When you watch service fail, you’re watching the last link in a chain. The failure is visible at the table, but it was manufactured upstream, in how the floor was designed, how the sections were cut, how the shift was rostered, how the kitchen and the pass are communicating, whether the mise for the front of house was actually done. Service is the output. The system is the cause.

Speed of service is a kitchen and design question

Take the most common complaint of all: food took too long. Guests experience it as a service failure: where’s my waiter, why is my food not here? But the waiter rarely controls the thing they’re being blamed for.

Ticket times are set by kitchen capacity, menu complexity, station balance, and how well the pass is being run. Is someone actually expediting, or are tickets just piling up and going out cold and out of sequence? If the kitchen is under-designed for the covers, or the menu has a dish that bottlenecks a station, or there’s no real expo calling the pass, food will be slow no matter how charming and attentive the floor team is. Then the reviews say “lovely staff, painfully slow,” and management responds by talking to the floor team, who were the one part of that sentence that was working.

Layout does the same thing quietly. A badly designed floor — service stations too far from tables, one printer for a room that needs three, a bottleneck at the entrance to the kitchen where everyone collides — imposes a tax on every single trip a server makes, all night, forever. You cannot train that tax away. It’s built into the building. The only fix is operational: move the station, add the terminal, re-cut the sections so a server’s tables are actually walkable in the time the service standard demands.

Rostering is a service standard

Here’s one almost nobody frames correctly: the roster is a service document. When a manager schedules for cost instead of demand, trimming a body to protect the labour line on a night that’s actually going to be busy, they have just decided, in advance, that service will be worse that night. They usually don’t experience it as that decision. They experience it as saving money. But they pre-loaded a service failure into the shift, and then the team wears the blame for it.

Good service needs the right number of the right people in the right positions at the right time. That’s a forecasting and scheduling discipline, not a personality trait. The warmest team in the world, understaffed on a full Friday, delivers bad service, because they spend the night triaging instead of hosting. Get the roster right and mediocre individuals deliver decent service. Get it wrong and excellent individuals deliver chaos.

Where training actually belongs

I’m not against training. Training matters, for the things that are genuinely knowledge and judgement: product knowledge, how to read a table, how to recover gracefully when something goes wrong, the culture of the room. That’s real, and you should invest in it.

But training only works on a foundation that lets people succeed. Train a team that’s set up to fail and you’ve just taught them, in high definition, the gap between what they know they should do and what the system allows them to do. That’s demoralising. It’s how you burn out good people. They know what good looks like, they came in wanting to deliver it, and every shift the operation stops them. The best of them tend to leave, which the venue often reads as a talent problem, and so it hires more people into the same broken machine.

So when I’m asked to fix service, I don’t start with the team. I start with the section sizes, the floor plan, the terminal count, the par levels, the roster against the real demand curve, the pass and how the kitchen talks to the floor, the menu and where it bottlenecks. I fix the system so that good service becomes the path of least resistance, the thing that happens naturally when people do their jobs, rather than the heroic thing they have to fight the operation to achieve.

Get that right and the training finally sticks, because now it’s teaching people to be excellent at a job the operation actually lets them do. Get it wrong and you’ll keep buying workshops and keep wondering why the water still doesn’t reach the table.