From kitchen to line — why your recipe does not scale

A recurring conversation: a founder has a drink they have made dozens of times, everyone who tries it loves it, and they want to produce it. Then the first scaled batch tastes wrong, and nobody can say why.

Scaling a beverage formula is not multiplication. Here is what actually changes.

Flavor dosing is not linear in practice

In principle, ten times the batch takes ten times the flavor. In practice, several things interfere.

Measurement precision changes character. At kitchen scale you are measuring flavor in drops, and a drop either way is a large percentage error you have been unknowingly averaging out across attempts. At production scale the same relative error is invisible on the scale but enormous in the tank.

Extracts and essential oils are not fully soluble. At small volume, vigorous stirring keeps everything dispersed. At scale, poorly soluble aromatics can separate, ring the tank, or distribute unevenly. Solubilizers and proper emulsions exist for this reason, and they change the flavor slightly.

Your kitchen ingredients are not production ingredients. Grocery store vanilla extract, fresh citrus juice, whole spices — none of these are available at production scale with the consistency a formula needs. Substituting a commercial flavor for a kitchen ingredient is a reformulation, not a swap, and it is the single most common source of "it does not taste like mine."

Water is suddenly an ingredient

At home you use tap water and never think about it. In production, water is treated, and treatment changes the taste.

Municipal water varies by source and season in mineral content, chlorination, and pH. Production water is typically filtered, sometimes carbon-treated, sometimes reverse-osmosis. Each of those removes something. RO water in particular is aggressively neutral, and a formula developed against mineral-bearing tap water can taste hollow when made with it.

If your product is water-forward — sparkling water above all — this is not a detail. It is most of the formulation.

Process changes the product

Pasteurization or hot fill, if your product requires it, applies heat that drives off volatile aromatics and can cook delicate fruit notes. Formulas frequently need to be over-flavored going in so they land correctly coming out. You cannot know by how much without running it.

Proper carbonation changes perceived sweetness dramatically. A formula developed flat will nearly always taste too sweet once carbonated. Develop carbonated or accept that you will rebalance.

Time in tank, filtration, and pumping all apply shear and exposure that a kitchen batch never sees.

Shelf life did not exist in your kitchen

You drank your test batches within days. Your customers will drink the product months later.

Some flavor systems are stable. Others rearrange — citrus fades, spice notes come forward, vanilla rounds out. A formula optimized to taste perfect on day one may be optimized for a day nobody experiences. Formulating for month three is a genuinely different target, and it usually means a slightly different balance than the one that wins on the bench.

What the development process actually looks like

The sequence that works:

1. Bench replication. We reproduce your kitchen formula using production-grade ingredients. This is the step where the biggest surprises happen, and it is better to have them here than in a tank.
2. Adjust to match intent. Not to match your original literally — to match what you were going for. These diverge, and the conversation about which one you actually want is important.
3. Carbonate and re-taste. Rebalance sweetness and acid against real carbonation.
4. Small pilot batch. Real equipment, real process, real package. Everything that only shows up at scale shows up here.
5. Hold samples and taste over time. Weeks, then months. Adjust for where the formula is heading, not just where it starts.
6. Lock it. Write the specification down — Brix, acid, pH, carbonation volumes, ingredient statement — and stop changing it.

The part founders find hardest

Step 2. Almost every founder wants the production version to be identical to their kitchen version, and it usually cannot be, because some of what made the kitchen version good was fresh ingredients that do not exist at scale.

The productive framing is not "reproduce this exactly." It is "here is what I was trying to achieve, and here is what I care most about — get me as close as production allows and tell me honestly where the compromises are."

Founders who hold that position get a good product. Founders who insist on literal reproduction usually spend a lot of money discovering that fresh-squeezed does not come in a tote.

How long it takes

Four to twelve weeks for most projects, longer if the formula is unusual or the ingredient sourcing is difficult. It is the least predictable stage of a beverage launch and the one most worth doing properly, because everything downstream — packaging, labels, shelf life claims — is built on the assumption that the formula is final.