The short answer
Strength and cost are the same dial. Every extra gram of leaf per serving raises the cup's cost and lowers the number of cups a kilogram delivers; every quiet trim of the dose saves leaf and spends customer goodwill. The discipline that keeps both honest is arithmetic run in the open: know your grams per serving, know your servings per kilogram, and know the strength the cup actually needs to pass your own taste check. When those three numbers are written down, cost conversations with a supplier become precise - the desk can select a lot whose natural strength lets the recipe hit the taste goal at the dose the budget allows. When they are not written down, every cost saving is a gamble served to customers one cup at a time.
The serving math
The whole calculation is one division. Servings per kilogram equals a thousand divided by grams per serving. At the 3-gram frame common in mug service, a kilogram yields roughly 330 servings; at the 2-gram end of the cut-tea frame, about 500; at a fixed pyramid bag dose of 2-3 grams, exactly what the pack says. Multi-infusion formats divide differently: a gongfu session of 7 grams through six or more rounds delivers its drinking cups at an effective gram-per-cup well below the session weight, which is why session formats look expensive per serving and are not; the gongfu note runs that arithmetic. Two refinements keep the division honest. First, use the real weighed dose, not the recipe card's aspiration - stations that scoop rather than weigh often run 20 to 30 percent over card, and every point of that is servings evaporating. Second, count waste: over-steeped pitchers dumped, undated cold brew discarded, and sessions retired early are all servings bought and never sold. The pyramid bag format exists partly because a locked dose makes this math exact.
Where strength is wasted
Waste hides in strength because extra leaf does not announce itself - the cup just reads a little darker, and nobody complains about free intensity. But past the taste goal, additional leaf and time buy bitterness, not value. The common leaks: doses running over card because scoops replaced scales; steeps stretched past the sheet's window to rescue a weak base, which extracts harshness instead of body; tea held hot on its leaf in an urn, darkening by the hour until it must be dumped and rebrewed; and bags sized for a large mug dropped into a small cup, where the same 2-3 grams over-deliver strength nobody wanted. Each leak has the same signature - cost drifting up while perceived quality stays flat - and the same audit: weigh what the station actually doses against the card, one afternoon, per SKU. Programs the desk supplies for hospitality get that audit in the onboarding checklist, because the gap between the card and the hand is where a quarter of the leaf budget can go; the black tea ratio note covers the station discipline that closes it.
The levers, ranked
When a recipe needs to hit a taste goal at a workable cost, four levers exist, and they are not equal. Leaf rate is the strongest and most expensive: grams move both flavor and cost linearly. Steep time is free within the window on the sheet - the difference between three and five minutes at the CTC frame is real body at zero leaf cost - and expensive past it, where it starts buying astringency instead. Water volume is a menu decision disguised as a brewing one: the same dose over 200 ml reads strong and over 300 ml reads balanced, so glass size is a strength decision. Temperature is the fine trim inside a family, worth a few percent, never a rescue. The professional sequence is to set volume by the menu, time by the sheet, temperature by the style, and only then adjust leaf to land the taste - which means the dose, the one lever with a price tag, moves last and least. Blending is the quiet fifth lever: a house blend can carry a signature profile at a steadier dose cost than any single lot, and the desk builds those against a reference cup rather than against a spreadsheet.
Writing strength into the order
None of this arithmetic works unless the supplier knows the cup being bought. The order conversation that gets a workable lot is specific: the vessel and volume it will be served in, the dose range the program can carry, the steep window the station can run, and what the cup must survive - milk, ice, or nothing. With that, samples are cupped against your parameters rather than a generic one, and the lot whose natural strength fits your dose at your window is the lot that costs less per cup in service even at the same price per kilogram. We coordinate samples from partner gardens and packing facilities, cup-check at the reference frames - the 3 g per 150 ml standard check, and the stronger 8.3 g per 110 ml full-strength pull from the standard five-factor cupping method - and confirm the validated dose and window in writing per program. Re-cup at each harvest change, because a lot's strength moves with the season even when its flavor profile holds. Start with the MOQ and lead time page for order mechanics, and send the cup spec through the quote form.
An illustrative program
Consider an illustrative mug-service program, numbers chosen for arithmetic rather than realism. It brews a CTC-style base at 3 grams per 200 ml; a station audit finds hands actually dosing 3.6 grams because the scoop runs heavy - a fifth of the leaf budget unaccounted for. Scales return the dose to card, and the recovered servings fund nothing more dramatic than accuracy. Next, the menu adds a large format at 300 ml; the same 3-gram dose now reads thin, and the temptation is more leaf. The sheet's window still has room: extending from three minutes toward five adds the body at zero dose cost, and only if the lot runs out of window does the gram move. Finally the program asks where a cost trim could come from; the honest answer is the cupping table, not the scale - a lot re-selected for higher natural strength at the same dose, validated against a reference cup, trims cost without touching the customer's experience. That sequence - audit the hand, use the window, then re-select the lot - is the whole method, and none of it required a single silent change to the cup.