Most Restaurant Quiz Questions Are Useless — Here's How to Tell the Difference
Hey Team!
I've been looking at restaurant training quizzes for a long time. And I mean really looking at them, the way you do when you're trying to figure out why a $300 quiz software isn't moving the needle on table turn times or comp rates or guest complaints. What I keep finding is the same thing, over and over. Most of the questions are useless.
Not slightly off. Not 80% there. Useless. The kind of question that a server could get wrong every single time and you'd never see it in your numbers. The kind that your best employee might miss because she was thinking about the table she just sat, not memorizing which of the four salad dressings came from a supplier in Vermont.
Here's how I think about it now. I use two questions to judge any quiz question I read. First: would a guest ever actually ask this? Second: does getting it wrong cost you money or put someone's safety at risk? If the answer to both is no, cut it. Don't revise it. Don't keep it as a "bonus" question. Cut it.
Let me show you what I mean with some real examples of the junk that ends up in these things.
"Which fork goes with the entrée?" Nobody at your family restaurant on a Tuesday night is asking about fork placement. And if you're at a place where fork knowledge actually matters, the question still doesn't pass the test, because the server either knows the table setup from doing it a hundred times or they learn it in thirty seconds of shadowing. A quiz doesn't move that needle. The answer lives in muscle memory, not recall.
"How many items are currently in the Starters section of the menu?" This one kills me. I've seen it. It's a trivia question dressed up as training. A guest is never going to ask how many items are in a section. And even if they did, the answer changes every time you 86 something. Memorizing a number tells you nothing about whether that server can describe what those items taste like.
"Which of the following ingredients is NOT in the Southwest Chicken Bowl?" Think about that construction for a second. You're training someone to know what's absent from a dish. But the guest doesn't ask what's not in it (unless they have an allergy, and we'll get to that). They ask what IS in it. They ask how it's prepared. The "which is NOT" format is a logic puzzle, not a service skill. It tests whether someone can eliminate options, not whether they can sell the bowl.
And then my personal favorite: anything that tests brand trivia. The name of the flour supplier. The year the recipe was developed. The city where the original location opened. I've seen whole sections of quizzes that read like the back of a cereal box. That's brand storytelling, and there's a place for it, but it's not a passing quiz question.
Now here's what a real question looks like.
"A guest tells you they're allergic to tree nuts. What's in the house vinaigrette?" That's a real question. A guest will ask it. Getting it wrong can land someone in the hospital. That question belongs in every quiz, every training cycle, and it should be taken seriously by every person on the floor.
"The green chile chicken dip comes out hot. A guest asks if it's spicy. What do you tell them?" That's real. You've got to be able to describe heat level accurately or you're going to have a guest sending the dish back. That costs you food. It costs you a comp. It costs you trust at the table.
"What comes on the side with the pulled pork sandwich?" Again, real. A guest is absolutely going to ask. And if your server says "I think it's fries?" when it's actually coleslaw and a pickle, that's a moment of doubt at the table. Doubt kills tips. Doubt makes a guest feel like nobody there actually knows their food.
The difference between the bad questions and the good ones comes down to whether the knowledge being tested actually shows up at the table. I don't care how many questions are on your quiz. I care whether passing it makes your staff more capable of having a real conversation with a guest about what they're about to eat or drink.
The pattern I see in bad quizzes is that they were built to prove someone did the training. They're defensive documentation, not development tools. You administer them, you get scores back, you file them away, and nothing changes because nothing was learned that would change anything.
Good training questions come from thinking about the last five complaints you got and the last five things a guest asked that your server couldn't answer. That's your quiz. Everything else is noise.
I built ShiftTrained because I spent too many years watching operators waste time on questions that didn't matter while their staff went into service not knowing what was in the sauce. The platform generates questions from your actual menu, which pushes the output toward the stuff that actually matters: ingredients, descriptions, heat levels, allergens, what comes with what. It's not perfect, but it's starting from the right place.
The next time you look at a training quiz, run every question through those two tests. Would a guest ask this? Does getting it wrong cost something real? You'll probably cut 40% of what's there. What's left will actually teach your people something.
Have a great day! — Terry Psaltakis
Your AI Restaurant Guy
For more on putting this into practice, see how ShiftTrained approaches cocktail training.
