Cold brew methods for cafés, compared
Short answer: cold brew is a stock item, not a drink. Whichever method you pick, you are committing coffee, refrigerated space and labour in advance of demand you have not yet seen. So the choice is mostly about lead time, capital, footprint and waste risk — and for many venues the right answer is flash-chilled coffee made to order, with no batching at all.
Why the method question is really an inventory question
Espresso is made when it is ordered. Cold brew, in every form but one, is made hours before anybody orders it, held cold, and then either sold or thrown away. That is why comparing methods on flavour alone leads operators to the wrong equipment.
Each method below answers the same four questions. How long before service must the brewing decision be taken? How much capital does it tie up? How much refrigerated and counter space does it occupy? And what happens to the unsold portion? Cold drinks are now a year-round line in a great many venues rather than a summer exception — see the rise of cold coffee and what it demands of equipment and coffee as an all-day drink.
What does cold water actually do to extraction?
Solubility falls as temperature falls. The soluble fraction of the coffee is the same fraction it always was, but at low temperature it leaves the grounds far more slowly — and the compounds within it do not all slow by the same proportion. Cold extraction compensates for the missing temperature with a very large increase in contact time. That is the whole trade; the mechanism is the one set out in how coffee solubility governs extraction, with only the rate changed.
Because the rates change unevenly, the result is not hot coffee served cold. The small, volatile aromatics that hot water pulls out immediately extract poorly cold, or are never released into the headspace where you would smell them; some of the acids behind hot coffee's brightness extract less readily; and heavier, sweeter-tasting material and the dissolved solids that carry body keep accumulating over the long steep. The cup is rounder, heavier, lower in aromatic complexity, and read by most drinkers as sweeter and softer.
The lower-acid claim needs care. What is reliably true is that cold brew is perceived as less acidic; measured pH is not dramatically different from hot-brewed coffee of the same origin and roast. Perceived acidity depends on which acids are present, in what proportion, and how much sweetness and body sit alongside them — not on pH alone. Treat the softness as a sensory fact, not a chemical or health claim.
Roast interacts strongly. Darker, more developed roasts are porous and readily soluble even in cold water, which is why they dominate cold brew programmes; denser light roasts can be excellent but are harder to extract to a satisfying body — see how roast profile changes extraction.
How does each method work, and what does it cost to run?
Immersion, or steeping
Grounds and cold water sit together in a vessel for a long period — hours, not minutes — then are separated by filtration. No pressure, no pump, and in the simplest version no equipment beyond a food-grade container and a way of straining. It is the cheapest to start, the most forgiving and the easiest to teach: extraction is slow and gentle, so errors in grind or ratio move the result far less than they would in espresso.
The costs are not on the equipment list. Lead time means today's decision serves tomorrow's trade. The vessel is large and must live in refrigeration, taking space you were already short of. And filtration is the real labour — separating a large volume of saturated grounds cleanly, without pushing fines through, takes time, produces a heavy wet waste stream, and is the step staff quietly start cutting corners on. Filter-bag systems reduce it considerably and usually earn their keep.
Slow drip, or cold drip towers
Water is released drop by drop onto a bed of coffee and percolates through it, collecting below. Each drop meets grounds that have already given up part of their soluble material, so extraction is more sequential than immersion, where everything sits in one increasingly saturated liquid. The cup is the argument: cleaner, more delicate, more tea-like, with more aromatic definition than an immersion of the same coffee.
The costs are substantial. The apparatus is expensive for what it produces, largely glass, and fragile in a working bar. Yield per unit of footprint is poor — a tower occupies prime space and produces a modest volume very slowly. Drip rate needs checking, beds stall or channel, cleaning is fiddly. It is also unavoidably a visible object, and a fair proportion of counter towers are bought as theatre rather than production capacity. A legitimate reason to own one; it should be a conscious one.
Nitro, or kegged cold brew
Nitro is not a brewing method but a serving format applied to a brew already made by one of the others: the finished coffee is kegged, nitrogen is dissolved into it, and it is dispensed under pressure through a restricting plate. The small, poorly soluble bubbles break out on dispense, giving the cascading pour and the dense, soft texture. Extraction has not changed; mouthfeel and presentation have.
The cost is a set of obligations rather than a purchase: kegs, gas supply and regulator, a font and lines through the counter, cold space for the kegs. Then the part most bars underestimate — line and font cleaning, on a schedule, permanently, with the staff time and training that implies. There is a real learning curve in getting gas pressure, temperature and restriction to give a consistent pour rather than a glass of foam. Where it earns its place it earns it well: strong visual pull, dispensed in seconds, no per-drink assembly labour.
Accelerated methods
Approaches that compress the steep into a short window using mechanical energy or altered conditions rather than heat: applied pressure, agitation, recirculation of liquid through the bed, ultrasonic energy to disrupt the boundary layer around each particle, and reduced-pressure or vacuum systems that change how readily material leaves the grounds. The principle is the same throughout — raise the rate of transfer without raising temperature.
The operational advantages are real. Lead time collapses, so you brew to observed demand instead of forecasting it, which is the failure mode that makes cold brew expensive. Stock turns faster and is fresher for it, and refrigerated holding falls. The counterweights are equally real: capital is significant, these are powered systems with pumps, seals and controls to service, and whether the cup matches a long immersion is genuinely contested — some tasters find accelerated brews thinner or harsher in the finish, others find no meaningful difference, and results vary by coffee and by system. Brew a batch against your existing method, with your own coffee and staff, before committing.
Flash-chilled, or Japanese iced
The caveat first: this is not cold brew. Coffee is brewed hot, at normal brewing temperature, directly onto ice, with the ice mass counted in the recipe so the drink is not diluted. Extraction happens hot; chilling is instantaneous. The aromatics cold brew loses are captured and then locked in by the rapid temperature drop, so the cup is bright and recognisably the same coffee you serve hot — a different drink rather than a substitute, and for a lot of coffees a better one.
Operationally it is in another category: made to order, in about the time a filter coffee takes. No lead time, no batch, no vessel, no refrigerated stock, no shelf life, nothing to discard at close beyond the grounds. For many venues this is the better answer, and it is routinely overlooked because it does not involve buying anything.
Comparison table
| Lead time | Capital | Footprint | Labour per litre | Batch flexibility | Shelf life | Suits | |
|---|---|---|---|---|---|---|---|
| Immersion | Long | Lowest — vessel and filter | Large refrigerated volume | Moderate, mostly filtration | Poor — fixed at brew time | As any batched cold brew | Steady cold sales; the default start |
| Slow drip | Long, slow once running | High for the volume produced | Prime counter space, poor yield for it | High — setup, monitoring, cleaning | Very poor; small batches | As immersion | Specialty cup character, or a visible object |
| Nitro (format) | Inherits the brew beneath it | High — kegs, gas, font, lines, cold store | Under-counter or cellar, plus font | Very low running; high fixed cleaning burden | Poor — committed by the keg | Longer sealed under gas; short once tapped | Volume bars, premium cold offer |
| Accelerated | Short — brew close to demand | Significant; powered kit needing service | Moderate, counter or back-bar | Low; the machine works | Good — small batches on demand | As immersion, but stock turns faster | Variable demand, tight fridge space, real volume |
| Flash-chilled | None — made to order | Negligible beyond brew kit and ice | None beyond the brew bar | Highest, but only for what sells | Total — cup at a time | Not applicable | Most cafés, especially with modest or uneven cold sales |
Which variables actually matter when you set one up?
All directional. Establish your own settings by tasting, and write them down.
Grind. Coarser than you would use hot: contact time is enormous, so surface area need not be large, and a fine grind in a long steep both over-extracts the smallest particles and makes filtration miserable. Slow drip needs a bed that will not compact or stall. Spread matters as much as the average — a wide distribution puts spent fines and barely-touched boulders in the same batch, which over a long steep reads as a muddy, drying finish. See particle size distribution explained.
Freshness. Cold brew's forgiving nature makes pre-ground coffee tempting; it still degrades from the moment it is ground, and cold brewing hides only part of the loss — what happens after grinding.
Ratio, and whether you are making concentrate. The decision with the most operational consequence. Settle it before buying a vessel — it sets how much cold space you need.
Time. Longer steeps extract more of everything, including the heavy, drying material at the end of the sequence; past a point, more time only adds weight and bitterness. Find yours by drawing samples across a range and tasting.
Agitation. Stirring at the start wets the bed and prevents dry pockets. Continued agitation accelerates extraction and shifts the profile — a variable, not a tidiness measure, so do it the same way every time.
Water. Long contact makes water chemistry more consequential, not less. Alkalinity buffers the acids that give the drink definition and can flatten a cold brew into something dull; very soft water tends to thinness. Specify it — see water for coffee.
Filtration. Cloth or paper gives clarity and strips body; coarse mesh keeps body and fines, which keep extracting in the fridge and settle in the glass. Choose deliberately and make it repeatable.
Concentrate or ready-to-drink?
Concentrate brews strong and dilutes at the point of service. It occupies far less refrigerated space per drink sold and gives menu flexibility — the same stock becomes a black cold brew, a milk drink, a tonic build or a cocktail base. The cost is a dilution step performed at the bar, under pressure, by several different people, which will be done inconsistently unless dosed by a measure rather than by eye. That step, not the brew, is where most concentrate programmes lose quality.
Ready-to-drink is brewed at serving strength and poured. Consistency is settled at brew time and cannot drift at the bar. It takes the space.
Concentrate suits a wide cold menu, tight refrigeration or a cocktail programme. Ready-to-drink suits high-volume single-format service, self-service, and any bar where staff turnover makes a dilution step a liability.
How do you cost it honestly?
Not from a recipe — from four measurements taken over a real trading period. Coffee used per litre actually served, not per litre brewed. Labour per batch, timed, including filtration, cleaning and keg or line work. Refrigerated space consumed, valued against what else could sit there. Discard rate — the fraction reaching end of shelf life unsold, recorded batch by batch across good weeks and bad.
The last dominates. Coffee is rarely the largest cost in a cold brew programme; unsold stock is. A venue brewing to its best week and selling to its worst shows a poor margin however well it brews, and the fix is smaller, more frequent batches rather than cheaper coffee. See how to cost and price a cup of coffee, cutting waste in a café and milk waste in a busy café.
What about food safety?
A real obligation, not a footnote. Cold brewing is an unpasteurised process held for long periods at temperatures that kill nothing, and kegging adds a low-oxygen environment. That combination brings documented hazards, temperature control, dating and record-keeping into scope, and the rules vary by jurisdiction. It is treated properly in cold brew and food safety — read it before the first batch you sell, not after.
Why most cafés should not build a cold brew programme
Start with flash-chilled and no batching at all. It needs no capital, no lead time, no refrigerated space and no shelf-life management, it uses the grinder and brewer you already own, and it makes a better drink than a badly run cold brew programme — which is what a lot of them are, not through incompetence but because batching a perishable against forecast demand is hard. A venue whose cold sales are modest or unpredictable should not tie coffee, staff time and fridge space up in advance of them.
A drip tower bought for the counter is usually theatre rather than production. If the theatre is worth what it costs, buy it with that understood; if you expected it to supply the cold menu, check its output against a busy afternoon first.
Nitro carries a cleaning obligation most bars underestimate at purchase. Lines and fonts do not clean themselves, the routine cannot be skipped when the bar is busy, and the failure mode is a product that tastes wrong for reasons nobody on shift can diagnose. Good for venues with the volume and the discipline; a liability for those with neither.
Accelerated systems solve the real problem — lead time and waste — and should be judged on that rather than on cup claims, once cold volume is large and predictable enough for the capital to have something to work against.
None of this is equipment Vea Group makes. We build espresso machines and milk systems, and have no cold brew, kegging, ice or drip equipment to sell you, which is why this comparison can be blunter than most.
Frequently asked questions
Is cold brew really lower in acid? It is reliably perceived as less acidic — softer, rounder, less sharp — but measured pH is not dramatically different from hot-brewed coffee of the same origin and roast. Treat the softness as a sensory description, not a chemical or health claim.
Which method makes the best-tasting drink? They make different drinks. Slow drip is the cleanest and most aromatic of the cold-extracted options, immersion the roundest and heaviest, nitro a texture change applied to either; flash-chilled is the most aromatic of all and is not cold brew. Brew your own coffee two ways side by side — general rankings are worth little against your beans and your customers.
Do accelerated systems produce a real cold brew? They produce coffee extracted without heat, which is the definition most people use. Whether the cup matches a long steep is disputed and depends on the coffee and the system. The operational case — brewing close to demand instead of forecasting it — is stronger and less contested than the flavour case.
How long does cold brew keep? That is set by your process, filtration, storage temperature and local food safety rules, not by a figure copied from an article. Establish it for your own product, document it, and read cold brew and food safety first.
This article is part of the Vea Group knowledge base. Vea Group S.p.A. is an Italian designer and manufacturer of professional and premium coffee machines, with heritage dating to 1919 and production in Chignolo d'Isola, Italy, and Suzhou, China.