Production identity
Category: batched cocktail · Style: Batch production · Yield: 10 controlled services before ordinary handling loss · Portion: 1/10 of verified batch by measured service portion
Glass / vessel: House-appropriate glassware · Ice: As specified by service method · Garnish: As required by house spec
Formula — 1× target batch
Scaling table
Scaling is arithmetic, not permission to ignore service behavior. Verify the 1× formula first, then make a controlled intermediate batch before a 50× event batch.
Method
- Verify every bottle/juice/syrup against the current house recipe and ingredient condition.
- Measure the complete batch into a sanitized food-safe vessel; add the specified water only if included.
- Mix thoroughly, chill promptly, label and hold under the venue’s approved procedure.
- Before service, make one finished test portion in the service glass and compare it to the single-drink sensory standard.
- Portion with a measured serving tool and do not top up a weakening batch by guesswork.
Working numbers
Measured pre-dilution liquid represented in ounce-based ingredients: 37.5 oz where applicable. Dilution control: Pre-dilution water in this training formula: 7.5 oz total. Calibrate against the venue’s ice, temperature and service method before locking the spec.
Storage: Keep cold, closed, labeled and segregated under the venue’s approved procedure. Fresh-citrus batches are same-shift quality products unless the venue has a validated alternative procedure. Shelf-life rule: Use the more conservative of the venue’s safety rule and sensory-quality limit. Do not carry a fresh-citrus batch forward merely because it is still cold.
Labeling: Manhattan — 10-Serve Pre-Diluted Batch · 10 serves · recipe version · production date/time · discard/review point · preparer.
Critical controls
- Scale from the current single-drink recipe, not memory.
- Use actual bottle ABVs for any batch-strength calculation.
- Homogenize before the first pour and re-mix when separation is possible.
- Taste a finished service portion before opening the batch to guests.
Production verification
Make the formula exactly as written before adjusting it. Record actual packaged or finished yield, starting ingredient temperature where it materially affects the result, and the service portion. Compare the finished sample against the sensory target: Boozy, herbal, lightly sweet. If the formula includes Rye whiskey, Sweet vermouth, Aromatic bitters, Cold potable water for controlled pre-dilution, verify that each product is the intended house specification; a substitution can change sugar, acidity, bitterness, aroma, alcohol strength or texture.
Sensory standard
The target is Boozy, herbal, lightly sweet. Evaluate the beverage at its intended serving temperature and dilution. The standard should be descriptive enough that two trained bartenders can compare results and identify which axis—sweetness, acidity, bitterness, alcohol, aroma, body, carbonation, temperature or dilution—is outside the house target.
Supporting professional reading
From the Keep ’Em Coming Back books
Behind the Bar
This page is a practical extension of Volume 4. The book explains the reasoning; this resource helps put it to work.
Whiskey · Spirit-Forward
Manhattan is a boozy, herbal, lightly sweet spirit-forward recipe. Use this page as a working specification, then test the exact brands, ice, glass, price, and service conditions before adopting it as a house drink.
Single-serve spec
• 2 oz (60 ml) Rye whiskey
• 1 oz (30 ml) Sweet vermouth
• 2 dash Aromatic bitters
Method
• Add the liquid ingredients to a mixing glass with ice.
• Stir until cold and properly diluted.
• Strain into a coupe or nick & nora, garnish, and serve.
Shared testing and documentation controls: Bar recipe operating standard →
Recipe-specific working guide
How Manhattan works in a real bar
Manhattan is a boozy, herbal, lightly sweet spirit-forward built around whiskey. This rewrite treats the listed specification as a working service standard: every measure, preparation choice, product, temperature, and finishing step must be tested together before the venue adopts it.
Ingredient functions
• 2 oz Rye whiskey is the structural lead. The house test should record both brand and proof; a substitute with a different flavor concentration can make the same written specification taste hotter, softer, sweeter, or more aromatic.
• 1 oz Sweet vermouth is the flavor bridge between the base and the balancing ingredients. Measure carefully: small overpours can dominate the recipe or leave the finish sticky.
• 2 dash Aromatic bitters seasons the drink in a small but powerful dose. Define the dash, pinch, or dropper used by the house; vague seasoning measures create more variation than their volume implies.
Technique and failure points
Manhattan depends on controlled stirring to chill and dilute while retaining a dense, clear texture. Combine the measured ingredients over firm ice, stir with an even path, and strain into coupe or nick & nora, garnish, and serve as soon as the house dilution point is reached.
The critical control is stopping at the same point every time. Too little stirring leaves the drink hot and sharp; too much makes the finish thin. Glass temperature, ice quality, and batch temperature all change the number of turns required.
• Inconsistent ice quality changes temperature and dilution even when every liquid measure is correct.
• Changing the glass, fill line, or garnish without retesting alters yield and guest perception.
Service-speed and batching plan
With 3 measured ingredients and a moderate service classification, Manhattan should be treated as a defined station procedure rather than a memory test.
Failure laboratory
Version control
Do not silently change the working formula. Record the recipe version, date, reason for change, test result and person approving the update. If a brand or product specification changes, rerun the sensory and cost check before treating the substitution as equivalent. A formula is professional only when the team can reproduce the same intended result and explain why a revision was made.