How to measure over-pouring with a pour test

Over-pouring happens in small ways. A jiggerA small measuring cup for pouring exact amounts of spirits and other ingredients. is filled past the line, a pour from the bottle runs a little long, or a regular gets an extra splash. Each time, the drink holds more spirit than its price assumed. A few ml look like nothing, but you pour the same drink hundreds of times a week. So first see what it costs, then find the drink to test, run the test and fix the pours that miss.

What does over-pouring cost?

For the rest of this guide, take a Rye Highball that sells for $10, with a specA drink's exact recipe: each ingredient and how much of it. of 1½ oz (45 ml) of house rye. A 750 ml bottle costs $30, so each ml costs 4 cents ($30 ÷ 750). With $0.20 of soda and garnish, the drink costs $2.00 to make: a pour cost of 20%. Now pour ¼ oz (7.5 ml) over. The extra rye costs $0.30 (7.5 × $0.04), and the pour costThe cost of a drink's ingredients divided by its menu price, as a percentage. rises to 23%.

If you sell 50 a night, 6 nights a week, that is 300 a week. The extra rye then costs $90 a week (300 × $0.30). Over a year, that is $4,680 from the Rye Highball alone. In bottles, it is 2,250 ml a week, or 3 bottles at $30 each: the same $90.

Which drink should you test first?

Now turn it around. Say the week's stock count finds 2,250 ml of house rye missing after you account for sales, deliveries and recorded waste. This gap is the bottle's liquor variance. Start by checking every drink and batch that uses that bottle.

Suppose the Rye Highball is the only drink that uses it, and you sold 300 during the week. Dividing 2,250 ml by 300 gives 7.5 ml per drink. That suggests an average over-pour of ¼ oz, or 16.7% over the spec, but it doesn't prove that pouring is the cause.

Suspected over-pour per drink = gap in ml for the spirit ÷ drinks sold that use it

If several drinks use the same rye, you can't assign the entire difference to one cocktail. Use the spec amounts and sales records to calculate expected usage, check transfers and other losses, then test the pours most likely to explain the remaining gap.

If the stock count shows a gap of about the same percentage on every ingredient of one drink, that drink is one possible cause. But those ingredients may go into other drinks too, so use the pattern to decide what to test first, not as proof.

How to run a pour test

A pour test measures what each bartender really pours. It takes a few minutes per person, with kit from a bar or kitchen supply store.

What you need

Running the test

1. Each bartender pours their most-used measure, 1½ oz (45 ml), five times into the service glass, at busy-night pace. They pour as they do on shift: with a jigger, or as a free pour (straight from the bottle, timed by counting).
2. With ice, put the glass and ice on the scale, set it to zero, pour, and read it at once. Without ice, tip each pour into the cylinder.
3. Write down all five pours, then the average.
4. Repeat for other measures your specs use often, such as ¾ oz (22.5 ml).

Between pours, empty the glass or cylinder and set the scale back to zero. Use the same bottle fill level and pourer for each set of five, so the results are comparable.

While the scale is out, fill each jigger line with water and weigh it. A US 1½ oz line holds about 44 ml, or 44 g. A metric 45 ml line holds 45 g. A line that really holds 50 ml makes every careful pour 5 ml over.

If your specs are in ounces and your jiggers are metric, pick one rounded number for the whole bar: 30 ml to the ounce. Then ¾ oz is always the 22.5 ml line, for everyone.

Reading the results

Take Bartender A and the 1½ oz (45 ml) spec. A pours 47, 52, 49, 51 and 50 ml. That is 249 ml, an average of 49.8 ml (249 ÷ 5). So A is 4.8 ml over, or 10.7% (4.8 ÷ 45). From the shortest pour to the longest, the spread is 5 ml (52 − 47).

Over-pour % = (average measured pour − spec pour) ÷ spec pour × 100

The average shows how far off a pour is. The spread shows how steady it is.

• Steady but long, like A. A's extra 4.8 ml costs about 19 cents per Rye Highball (4.8 × $0.04 = $0.192). If A makes all 300 Rye Highballs in a week, that is $57.60. If A makes fewer, use the number A actually serves. Teach A to stop the pour a little sooner, then test again.
• Scattered, like Bartender B. B pours 40, 52, 44, 55 and 47 ml: 238 ml, an average of 47.6 ml. The average looks close, but the spread is 15 ml (40 to 55), so single pours miss on both sides. Show B the results, practice the measure together and test again. If the pours are still inconsistent, use a jigger for that measure during service.
• On spec, like Bartender C. C pours 45, 44, 46, 45 and 44 ml: 224 ml, an average of 44.8 ml and a spread of 2 ml.

Aim for the spec. No published standard says how far a pour may miss, so agree a limit with your team before the next test. Test on a schedule the team knows, and show each bartender their five numbers on the spot.

Free pour or jigger?

A jigger filled to the line gives close to the same pour every time, if the jigger is right. A free pour changes with the bartender, the pourer and how busy the night is.

Glassware can affect free-pouring accuracy. In a controlled study of bartenders, researchers found that pours into short, wide glasses were larger than pours into tall, narrow ones, particularly at their normal pace: 59.4 ml against 47.9 ml, with a 1.5 oz (44.3 ml) target.

A later study measuring drinks served in California bars found a different pattern: glass size mattered in some situations, but glass shape alone did not consistently predict how much alcohol was served. A shot held about 1.6 oz (48 ml) on average in a large shot glass, against 1.3 oz (38 ml) in a standard one.

The practical lesson is to test the glassware and pouring methods your bar actually uses rather than assuming a bartender will pour the same amount into every glass.

So the pour test uses the service glass and busy-night pace. If your team's free pour can't hold the spec after practice, bring in jiggers.

Where over-pours come from

Not every extra pour is accidental. A bartender may give a friend or regular a little more, and that won't necessarily show up in a supervised pour test. Stock counts can reveal unexplained usage, but you still need to investigate the cause before assuming anyone is intentionally over-serving.

Other over-pours come from how the bar is set up:

• A bartender learned to free pour with a different pourer.
• The jigger has no ¾ oz line, so every ¾ oz is a guess.
• A spec changed on the menu, but the change never reached the team.
• In a busy hour, a careful pour turns into a fast one.

How to stop over-pouring

Retrain each bartender on the measure that missed, the same day. Test again before the next stock count.

• Free pourers practice into a cylinder until their pour lands on the spec, then test again in the real glasses.
• With jiggers, fill to the line every time, over the tin. Keep a jigger at every station, with the lines your specs use.
• For a bottle that keeps showing a gap, fit a measured pourer. It stops at a set amount, such as 1½ oz, so nobody has to count.

Tell the team the cost in money, not percent: "$90 a week on the Rye Highball" is easy to remember. Work out your own figure with the over-pour calculator, from your bottle price, pour size and drinks sold a night.

After coaching, repeat the same five-pour test with the same glass and pourer, and record both results. At the next stock count, check whether the bottle's gap has fallen, allowing for changes in sales and any other recorded stock movements. Track each bottle's gap from one stock count to the next. A bottle that shows a similar gap every time is the next one to test.

Questions

Sources

1. Brian Wansink and Koert van Ittersum, Shape of glass and amount of alcohol poured: comparative study of effect of practice and concentration. Methods and table 2: 86 Philadelphia bartenders pouring a 44.3 ml shot into tall and short glasses, at their normal pace and after being told to take their time.
2. William C. Kerr, Deidre Patterson, Mary Albert Koenen and Thomas K. Greenfield, Large drinks are no mistake: glass size, not shape, affects alcoholic beverage drink pours. 480 drinks measured in 80 Northern California bars and restaurants, 2007. Table 1 (shot glass size) and the discussion (two focus groups of 16 bartenders).