How Long Should It Take to Get Your First Drink at a Restaurant?

How long should it take to get your first drink at a restaurant? At a full-service spot, plan on having your drink in hand within about 5 to 7 minutes of being seated, with under 10 minutes a realistic ceiling during a busy rush. The clock that matters runs from the moment the guest sits down, not from when the order is rung in, because the wait before a server greets the table is where most of the delay hides. Cocktails and wine-by-the-glass naturally run longer than a beer or a soda, but most well-run rooms keep their median first-drink time under 8 minutes across the board.

First-drink time is one of the few service metrics a guest feels in their body. They sit down, they look at the menu, and then they wait — and somewhere around the seven-minute mark, an empty table starts to feel ignored. Getting a drink down quickly buys you patience for everything that comes after it: a slow kitchen, a missing side, a check that takes a minute. It's the cheapest goodwill you can buy on the floor.

What counts as a 'good' first-drink time

There's no industry-certified number, but most full-service operators converge on a similar feel. The useful way to think about it is in tiers, measured from when the guest is seated to when the drink physically lands on the table.

These are rules of thumb, not laws. A high-volume neighborhood spot pouring mostly beer and wine can and should beat five minutes. A craft cocktail bar where every drink is stirred, strained, and garnished will run longer, and guests expect that. The point isn't to hit a universal number — it's to know your own median and your own worst tables.

The clock starts at seating, not at the order

This is the part most operators get wrong when they try to measure it. If you only time from order-fired to drink-delivered, you're measuring your bar's speed, which is usually the part that's already working. The delay that frustrates guests almost always lives upstream: the gap between sitting down and a server actually greeting the table and taking the drink order.

A common rule of thumb is to greet a new table within roughly the first 30 to 60 seconds, get the drink order in fast, and keep the table moving from there. If your greet is slow because a section got triple-sat, no amount of bar speed will save the first-drink number. Measure the whole arc, seated to drink-in-hand, or you'll fix the wrong station.

What actually drives the delay

• Greet lag: the table is seated but no server has acknowledged it yet — the single biggest hidden cost.
• Section overload: one server triple- or quadruple-sat during a wave, so every greet backs up.
• Bar bottleneck: a single bartender during peak, or a well that's out of a key spirit or garnish.
• Drink complexity: a round of stirred cocktails simply takes longer than four beers, and that's fine if guests know it's coming.
• Runner gaps: the drink is made and sitting on the rail, but nobody's free to run it.

Notice that only two of these are bar problems. The rest are floor-flow and staffing problems, which is exactly why first-drink time is such a good diagnostic — it tells you whether your bottleneck is the bar or the room.

How to measure it without standing there with a stopwatch

The manual way to get this number is to put a manager on the floor watching a handful of tables and writing down times. That works for an hour, on the tables you happen to watch, while everyone's on their best behavior because the boss is standing there. It doesn't scale to every table, every shift, every location, and the moment you stop watching, the number drifts back to whatever it really is.

You also can't fully get this from your POS. The POS knows when the drink was rung in and maybe when the check closed, but it has no idea when the guest sat down. So the most important segment — seated to first greet — is invisible to it. That's the segment where the wait actually lives.

Reading it off the cameras you already have

This is where vision-based measurement helps. The seat, the greet, and the drink reaching the table all happen in plain view of the cameras most restaurants already run. A vision-language model can mark when a party is seated, when a server first engages the table, and when the first drink lands — no POS event required for the part that matters. Pair that with the POS ring for the order itself, and you get the full timeline on every table, not a sampled handful. Most rooms can measure 80 to 90 percent of tables on their existing cameras from day one.

Practically, that means you stop guessing. You see the median first-drink time per section, per server, per shift, and per location, and you can tell at a glance whether a slow night was a bar problem or a greet problem. When several cameras cover one table, the readings are de-duped so the table isn't double-counted, and human reviewers validate the AI's calls in the first week so you trust the numbers before you manage to them.

What to do with the number once you have it

• Track the median and the 90th percentile, not the average — the slow tail is where guests get angry.
• Break it down by greet-time vs. bar-time so you fix the real bottleneck instead of yelling at the wrong station.
• Watch it by section during your peak wave to catch over-seating before it snowballs.
• Compare locations on the same scale so a strong room can teach a struggling one.

Measuring this by hand means a manager on the floor with a stopwatch and a clipboard, catching a few tables for an hour and hoping it's representative. Reading it off the cameras you already run measures it continuously, on every table, in a way nobody can game — and it does it across every location at once. That's the difference between a number you argue about and a number you can actually manage to.

FAQ