The short answer
Most black tea is brewed between 2 and 3 grams of leaf per 200 ml of water at a full boil for three to five minutes - that is the working frame for cut-tea and mug service - while orthodox whole-leaf styles sit closer to 3 grams per 150 ml, and delicate bud-heavy styles take slightly cooler water. The frame matters less than the discipline around it: pick the ratio on the lot's brew sheet, weigh the leaf, and repeat it identically across every shift and site. Ratio is the strongest lever a buyer controls after the lot itself, and the one that decides whether a kilogram of tea delivers the same cup in ten locations or ten different cups.
The reference ratios
| Line | Leaf and water | Temperature | Time |
|---|---|---|---|
| CTC black | 2-3 g / 200 ml | 100°C | 3-5 min |
| Assam | start from the CTC frame | 100°C | 3-5 min |
| Ceylon | start from the CTC frame | 100°C | 3 min |
| Keemun | 3 g / 150 ml | 95°C | 2-4 min |
| Lapsang souchong | 3 g / 150 ml | 95°C | 2-4 min |
| Jinjunmei | 3-5 g / 110 ml | 90-95°C | short, to taste |
| Dianhong | 3 g / 150 ml | 95-100°C | 2-4 min |
| Lemon black | as for the base black | 100°C | 3 min; also good cold brewed |
Every row is a trade reference frame, not a specification: cut teas and whole-leaf teas behave differently even inside the category, and lots move a little with the harvest. Tune each line by cupping on arrival, then freeze the numbers you validated into the station recipe. Where a cell says start from the CTC frame, that is a house starting point for lines whose sheets specify time and temperature rather than dose - set the dose by cupping and write it down.
Why the ratio outranks the timer
Of the three numbers a barista touches - dose, time, temperature - dose is the one that drifts most and matters most. Timers are cheap and keep time well; kettles hold temperature; but scoops lie. A scoop of dense CTC granules and the same scoop of fluffy whole-leaf dianhong differ in weight by a factor of two or more, so venues that dose by scoop brew two different teas under one recipe card. A pocket scale reading to a tenth of a gram costs less than a few servings of leaf and removes the entire problem. Dose also has the cleanest relationship to cost: it is the number that converts directly into servings per kilogram, which the strength and cup-cost note works through in full. Time and temperature fine-tune the cup inside the window on the sheet; dose decides whether the window exists at all.
Cut and whole-leaf teas also punish the same scoop differently, which is worth spelling out for buyers building a mixed program. CTC granules are small, dense and uniform, so a leveled spoonful repeats within a few tenths of a gram. Whole-leaf dianhong or jinjunmei is open, airy and irregular: the same spoonful can vary by half its own weight depending on how the leaf settles in the scoop. That is why the cut lines tolerate scoop-and-timer stations while the whole-leaf lines do not, and why a venue scaling up whole-leaf service buys scales before it buys more tea. The lesson generalizes: as leaf gets more expensive and more irregular, the weighing habit that protects the cup protects the leaf budget in the same motion.
Brewing for milk tea: strength on purpose
A milk tea base is judged in the finished drink, not in the cupping bowl, and the finished drink dilutes whatever you brew. Milk, ice and syrup each mute the tea, so a base that tastes correct black will vanish in the cup. The move is more leaf, not more minutes: raise the dose above the straight-tea frame and keep the steep inside the three-to-five-minute window, because past that point extra time mostly extracts harshness rather than body. Cut teas earn their place here - CTC-style rolling produces the small, dense particles that extract fast and strong, and Assam-style leaf brings the malty weight that stands up to milk. The desk cups bases against your actual recipe - milk, ice and all - before a lot ships, and the bubble tea base note walks through that selection in detail.
One recipe, every shift, every site
Consistency is a system, not a virtue. The working version has four parts. A written recipe per SKU: dose in grams, water volume, temperature, time, and what the finished liquor should look like - a photo taped inside the station beats a paragraph of prose. Scales and timers at every station where tea is brewed, no exceptions for small sites, because drift starts where the tools stop. A reference brew: at the start of the week, make one cup exactly to spec, keep it, and taste every shift's output against it; palates recalibrate themselves against the last cup they drank, and the reference breaks that loop. And a named owner: someone whose job includes the weekly check that the scale still reads true and the recipe card still matches what the hands are doing. Programs the desk supplies for multi-site operators get the same brew sheet per lot, so a new harvest changes the tea and not the procedure.
Put the numbers in the order
When you request samples, state the brewing reality you are buying for: vessel, dose, water, and whether the cup will meet milk or ice. A lot chosen for a 3-gram black mug service is a different purchase from the same category chosen for a strong iced base, and suppliers can only hit a cup they can see. We cup-check candidate lots at the reference frames and at your own parameters, issue per-lot brew sheets, and confirm the final numbers in writing with each program - the standard five-factor cupping at 3 g per 150 ml is the neutral yardstick both sides can point to when opinions differ. Start from the black tea category or send the drink spec through the quote form. For how the two classic milk-tea origins differ, see Keemun versus Dianhong, The same discipline holds on the green side of the menu, where temperature joins ratio as the number to write down.