First Round All posts

Promotions · Operations

The drink-special timing that actually moves orders

A special is a prompt, not a price. The three moments in a night when one actually changes an order, the break-even arithmetic before you discount anything, and why a window beats a whole night.

Most drink specials do not fail because the price was wrong. They fail because they were announced to nobody, at no particular moment, to people who had already decided what they were drinking.

A special is not a price. It is a prompt — an interruption that arrives at the moment somebody is deciding whether to order again or go home. Get the moment right and a modest discount moves real volume. Get it wrong and you have handed margin to people who were going to order anyway.

Start with the arithmetic, because it disqualifies about half of what people try.

Before you discount anything: the break-even

A discount only works if the extra volume covers the margin you gave up. That is calculable, and you should calculate it before you write anything on a board.

Say a pint sells at $8 and costs you $2. Your gross profit is $6 a unit.

Drop it to $6, and your gross profit is now $4 a unit.

$6 ÷ $4 = 1.5. You need to sell 50% more pints during the special just to make the same money as doing nothing.

That is the formula for anything you are thinking of discounting:

Break-even volume multiplier = current gross profit per unit ÷ discounted gross profit per unit

Run it on your own numbers before every special. It produces some uncomfortable results — a $8 drink discounted to $5 needs twice the volume; a 50%-off cocktail on a 70% margin often needs more than triple. And crucially, that uplift has to be new volume. Every unit sold to somebody who would have paid full price makes the number worse, not better.

This is why the timing question is the real question. The only way a discount clears its break-even is by landing on people who were about to order nothing.

The three moments that matter

A night is not a flat surface. There are three points where an order is genuinely undecided, and a prompt at any of them changes behaviour far more than the same prompt at any other time.

1. The arrival window — the first ten minutes

People decide their whole night's drink in the first order. Whatever someone starts on, most will stay on it. That makes the arrival window the highest-leverage moment in the entire evening, and the one nearly every bar wastes.

An arrival special is not really a discount, it is an anchor. Put the thing you want people drinking all night in front of them before they default to what they always order. If you would rather sell a $12 cocktail than a $6 domestic, the moment to say so is while they are still standing up.

This is also the cheapest one to run, because you are not discounting — you are recommending. The "special" can just be a staff member saying "we're doing the ginger old fashioned tonight, it's what everyone's on."

2. The trough — between drink one and drink two

This is the leak. Somebody finishes a drink, nobody says anything, and within a few minutes they are deciding whether this is a one-drink night. Every bar has watched a table leave with empty glasses in front of them.

The trough is where a timed special earns most of its money, because it is the only moment where the alternative to ordering is genuinely nothing. Volume you generate here is close to pure new volume, which is exactly what the break-even formula demands.

Two ways to work it, and the second is better:

3. The stay-or-go moment

Late on, the question is not what to drink, it is whether to have one more. A discount is a weak lever here — people leave because the night has run out of reasons, not because drinks are expensive.

What works is a reason to stay ten more minutes: something about to happen. A last round of a game, a final draw, the end of a set. If you have nothing happening, a small "one for the road" price on a single named drink is a reasonable fallback, but it is the weakest of the three moments and you should not build your promotion around it.

Six rules that decide whether it lands

A window, not a night. "Half-price wings, 8:00–8:30" beats "wings special tonight" on every axis: it forces a decision now, it is announceable as an event, it caps your exposure to a known number of units, and it concentrates the kitchen hit into a period you staffed for. An all-night special is just a lower menu price, and your regulars will simply pay less for the same order.

Announce it out loud. Do not display it. A chalkboard is read once, on the way in, by somebody who has already committed to being there. It cannot reach the person sitting with an empty glass at 8:40. A spoken prompt reaches the whole room simultaneously, at a moment you chose. If the only place your special exists is a board, you are running a price change and calling it a promotion.

Attach it to something already happening. A room's attention is collectively available only at breaks — between rounds of a game, at a set change, at halftime. Announcing into a break gets heard; announcing into a room mid-conversation does not. If nothing is happening on your Tuesday, this is an argument for having something happen, not for announcing louder.

Discount the adjacent item, never the bestseller. Cutting the price of your top-selling beer is the purest way to lose money: nearly all of that volume was already yours. Discount the thing next to it — the higher-margin cocktail you want to grow, or the appetiser that adds forty minutes of dwell time. Food specials are usually the better lever anyway, because food keeps people seated and seated people keep drinking.

One thing, not three. A special with three options is a menu, and a menu requires deliberation. At the trough, deliberation loses. Name one drink, one price, one deadline.

Prefer a bundle to a discount. "$12 for a beer and wings" protects margin far better than taking 30% off both, because you control the combination and the customer compares it to two full prices rather than one. Bundles also raise the check rather than lowering it.

Getting staff to actually say it

A special that lives on a schedule and not in a person's mouth does not exist. Before service, the team should know four things, and they should be written somewhere they can see:

  1. What the offer is — one line.
  2. When the window opens and closes.
  3. Who says it, and at what moment.
  4. The exact sentence. Not "push the wings" — an actual script: "Kitchen's doing half-price wings for the next twenty minutes if you want to grab some before it ends."

That last one is the difference between a special that gets mentioned and one that gets sold. A specific sentence gets said; a general instruction gets forgotten by the second table.

Measuring it without fooling yourself

Compare the same weekday, the same hour, week over week — and count units, not revenue. Revenue during a discount can rise while gross profit falls, which is the exact trap the break-even formula exists to catch.

Three questions after four weeks:

Four weeks is enough. If a timed special has not moved units by then, move the window before you move the price — the moment is nearly always the problem, and the price nearly always is not.

The awkward part of all this is that the three moments are things you have to notice, live, while also running a bar. Knowing the trough arrived at 8:40 is easy in hindsight and hard at the time.

That is where having something structured happening in the room helps: a game night has scheduled breaks, so the trough is not something you have to spot — it is on the run sheet, and the microphone is already in someone's hand. First Round puts the drink special on the screen at the break and prompts whoever is hosting to say it, which is a small thing that mainly solves the "we forgot to announce it" problem.

But none of the reasoning above depends on that. Run the break-even, pick a window, write the sentence, and count units.