Hiring Restaurant Servers: What to Test and Train in 2026

When the usual way of hiring falls short?

Hiring by résumé and a quick chat falls short the moment you post the same opening twice in one quarter, because that tells you the filter is letting in people who cannot handle a shift.

With 986,000 job openings in U.S. accommodation and food services in October 2025 (BLS, JOLTS), the candidate walking through your door also has other offers in their pocket, and a ten-minute interview only measures whether they can talk about themselves. The clearest signal sits in your own payroll: count how many hires you signed in the last six months and how many of those people are still waiting tables today.

If half are already gone, the problem is not the labor market or the generation coming in; it is a method that hires blind and then blames the server for failing to guess the house standard.

Employee referrals: who they suit and what they cost

Employee referrals work best for a house that already has a stable core of good servers, because people recommend people like themselves and, with a healthy team, that multiplies what already works. The switching cost is low: a referral bonus paid when the new hire completes the probation period, never on the day they sign, together with a written rule so nobody brings in a cousin out of obligation.

For example, if you set that bonus to pay out at 90 days, the server who made the referral becomes the newcomer's sponsor and carries them through the first weeks, which is exactly when most people are lost. The limit shows up on teams with a bad climate, where the referral inherits the habits of whoever brought them in and the clique grows. There, fix the house first, and refer later.

Is a paid trial shift worth it?

Yes, a paid trial shift is the alternative that gives you the most information per dollar spent, because you watch the candidate carry plates and survive the service rush before anyone signs anything.

It fits full-service restaurants with a fast pace, where a weak server shows up in the average check and in reviews within days. Setting it up takes moderate effort: a short shift, paid at the house's normal rate, with a veteran server assigned to fill out an observation sheet with fixed criteria (how they greet, whether they write or memorize, how they react when the kitchen falls behind, whether they dare to suggest dessert).

For example, four hours on a Friday night tell you more than two interviews. But the trial demands discipline from the manager, and if you use it to plug a staffing gap without watching, you only got cheap labor and an annoyed candidate.

Internal bench: promoting the busser or the barback

An internal bench, which turns bussers, hosts or barbacks into servers, is the cheapest option over the long run and the slowest to get going. It suits a house with several locations or a large support team, where someone always knows the menu and the Thursday regulars.

The real cost lies in the plan: a written server standard and a mentor for each candidate, with a clear promotion criterion, because if the person with the most tenure moves up instead of the one who serves best, the bench rots. I got this wrong for years, promoting by seniority while my best busser left for another house that did make him a server.

The advantage nobody charges you for is that the promoted person has already proven they stay, and in this trade that is worth more than any experience claimed on paper.

Situational interviews instead of textbook questions

A situational interview replaces «what are your weaknesses?» with real dining-room scenes and tells you in twenty minutes how the candidate thinks under pressure. It is the alternative for the manager who cannot run a trial shift every week, or for the small house where the owner hires alone.

Switching costs little: you write five or six scenes of your own, such as the party of eight that arrives without a reservation or the steak sent back as overcooked, and you decide in advance which answer is good in THIS house. My view is firm: the experience on the résumé weighs less than the answer to the returned-plate scene, because that is where you see whether the candidate protects the guest or only protects themselves.

Its limit is just as clear, since it measures judgment and not physical stamina, so it works better paired with a short trial.

Rehiring the people who left on good terms

Rehiring a server who left on good terms is the fastest alternative of all, because that person arrives already knowing the menu and the chef's temper. It fits seasonal restaurants or those near universities, where many leave for school or a move and not over a fight, and the switching cost is close to zero: an honest exit interview and a list of former servers who agreed to be contacted, whom you message when an opening appears.

That is why the day someone resigns matters as much as the day they start, and the manager who lets people go with resentment shuts a door that would have filled a section within a week. The risk is nostalgia, rehiring someone who left for poor performance because «they already know the house». The rule I propose is simple: whoever you would hire again today without hesitation comes back, and nobody else.

The full system: how Masterestaurant sequences it

None of these alternatives is enough alone, and the combination Diego F. Parra recommends at Masterestaurant sequences the process like this: first the situational interview to filter judgment, then the paid trial shift to see the body at work, next a verified reference with a former direct supervisor and, finally, 21 days of guided training against a written standard. The order matters because each step costs more than the one before, and screening out early saves veteran hours.

What happens if you skip the training and leave the new hire alone from the first Friday? They make rookie mistakes at the highest-check table, take the complaint, get scolded in public by the manager and quit in the third week, so the opening reappears and the veteran who covered for them ends up more tired than before. That counterfactual is turnover seen from the inside.

When NOT to change how you hire?

Do not change your method if your floor team has stayed almost intact for more than a year, because the system you have, even if informal, is working and touching it to follow a trend can break whatever holds it together.

This happens in family-run houses where the owner knows every server and their family, or in small restaurants that hire once a year and fill the opening with a trusted acquaintance. There, adding formal trials and evaluation sheets piles on bureaucracy and returns nothing. Changing in the middle of high season is also a bad idea: a new process should debut in quiet weeks, when the manager has the headspace to observe.

What every house should do, even one that changes nothing, is write its service standard on a single sheet, because the day the trusted server leaves, that page will be all that remains of what they knew.

What changes when hiring has a system?

The most common mistake when hiring restaurant servers is believing the problem is finding people, when the real problem is that the people you find leave.

In October 2025 alone there were 462,000 voluntary quits in U.S. accommodation and food services (BLS, JOLTS), and each of them was, months earlier, a hire somebody celebrated. A selection system is judged by how many people are still in their section at 90 days, and once you see it that way you stop rewarding the manager who 'hires fast'. There is a tension in the trade worth settling head-on: do you hire attitude or experience?

The National Restaurant Association reports that 4 in 10 U.S. restaurant employees are teens or young adults, so much of your candidate pool will have no track record to show.

What changes when hiring has a system — in practice?

Hire for ATTITUDE and test experience in the trial shift, because attitude cannot be taught and the menu can; the newcomer who offers dessert unprompted is worth more than the veteran who waits for the table to wave.

Run the opposite scenario. What happens if you skip the trial shift to cover a Saturday? The new hire starts without the standard, makes mistakes on real tables, the server covering them loses tips, that coworker starts looking elsewhere, and two months later you have two openings instead of one. The three hours you saved cost you the person who knew how to work.

I got this wrong for years: I thought internal promotion was slow and the market would always have servers ready. In fact the busser who has spent months on your floor already knows the house, the kitchen respects them, and their learning curve sits in the menu and in selling, both trainable; the outside hire brings technique and none of the rest.

What changes when hiring has a system — key points?

The same rule applies to restaurant team development as a whole, including hiring kitchen staff: promote first, recruit outside second. Decision tree in 4 questions.

One: is your service standard written down? If not, write it before hiring anyone, since no test can be scored without it. Two: do you have a busser or runner with months in the house who wants to move up? If so, promote. Three: is your core team well rated and stable? If so, open referrals with a bonus paid at 90 days; if it is burned out, do not. Four: must the opening be filled in under a week? Still run the situational interview and the paid trial shift; the search can shrink, the filter never does.