How Custom Supplement Formulation Works: From Brief to Bench to Pilot

Search “how supplement formulation works” and you get a clean five-step list: concept, ingredients, formulate, sample, produce. Sometimes a week count is bolted on: four weeks, eight weeks, twelve. The list is a fair map of the stages. What it leaves out is the terrain: what you have to bring before step one, what actually happens across two or three rounds at the bench, and the one seam where a formula that’s perfect in a beaker quietly comes apart. The jump from the bench to a production line.

That seam is where launches slip. Not because the formula was wrong, but because a sample mixed by hand in a 200-milliliter beaker is a sample at beaker scale, and a production tank is a different physical world. Actives that dissolved fine can drop out. An emulsion that held can break. A flavor that read clean can drift. The formula doesn’t automatically survive the trip, and the pilot run exists precisely to find out before your first real batch does.

This is the custom path from the inside: the brief the bench actually needs, what each sample round is for, why the pilot run is not optional, and how reformulating a product you already sell differs from a clean-sheet build. One thing up front: there is no honest single timeline for a custom build, and any guide that hands you a fixed number of weeks is guessing before it has seen your actives. A short list of hard variables sets the clock, and we name them below.

Short answer. Custom formulation runs from a brief, through the bench, to a pilot. The brief is your target and constraints, not a finished recipe. The bench builds and iterates the formula across two to three rounds. The pilot makes it on real production equipment, often in the hundreds of units when compatible materials are on hand, to prove it survives the jump from beaker to line. The total timeline is set by how hard the flavor, stability, and sourcing turn out to be, which is why no honest number exists before feasibility runs.

Best for: Founders building a first product, and brands reformulating or extending a line, who want to know how the custom path actually runs.

Key decision: Whether the product itself has to be your differentiator (a custom, owned formula) or your brand can carry a proven stock base at lower cost and higher speed.

Apollo path: Apollo Future Labs runs the bench, the pilot, and the line from one floor in Livermore, California. A manufacturing quote returns the path, the real minimum, and a timeline built on your actual formula, not a template.

Educational overview: not legal, regulatory, or medical advice. Requirements change and vary by jurisdiction and sales channel. Last reviewed July 2026.

What a supplement formulation brief actually needs

The bench cannot start on “a greens powder, but better.” It starts on a brief: a set of inputs a formulator can act on. The more of these you bring, and the more honest they are, the fewer rounds you burn getting to a formula you’ll sign off. A vague brief doesn’t save you time; it spends your first sample round discovering what you could have said in an email.

You do not need a finished recipe. That’s the formulator’s job. You need a clear target and real constraints. Here is what actually moves the work forward:

• The goal and the positioning. What the product is for and who it’s for, in one or two lines. “A bedtime sleep shot for shift workers” tells a formulator more than a list of trendy ingredients, because it sets the format, the dose, the taste target, and the claims lane at once.
• The actives and rough doses, and which are must-haves versus nice-to-haves. If a specific dose is the whole pitch, say so; if you want “a clinically studied amount of magnesium” and are flexible on the form, say that instead. Flag any branded or patented actives you intend to use, because they carry their own specifications, sourcing, and sometimes label conditions.
• The format. Liquid shot, syrup, capsule, powder, gummy, or “open to what fits.” Format decides half the formulation problem, and the right format is sometimes not the one you arrived with.
• The claims direction. The structure/function story you want to tell drives what has to be in the formula and at what level. You don’t need final label copy, but the direction shapes the build, and what you can and can’t say is a question for your counsel, not the bench.
• The hard constraints. A target cost-per-unit ceiling, the sales channel, the label real estate you have to work within, allergen and vegan and clean-label rules, and a target shelf life. Constraints are not the enemy of a good formula; they’re what keeps the bench from designing something you can’t sell.
• A benchmark product. A competitor you admire, a sample you like the taste of, a texture you’re chasing. “Make it feel like this” removes more ambiguity than a paragraph of adjectives.

Bring most of that and the bench opens on round one instead of on a discovery call. It also lets a good shop tell you something useful before anyone mixes anything: whether what you’re describing can actually be made the way you’re picturing it.

Feasibility first: can it be made, and should it be made this way?

Before a single sample is mixed, a disciplined shop runs a feasibility screen. The step-by-step guides skip this, which is why their timelines are fiction; feasibility is where the real ones get set. It’s a paper-and-experience pass over the brief that answers one blunt question: is this makable as described, and is this the right way to make it?

Feasibility looks at the things that quietly decide a formula’s fate:

• Compatibility. Do the actives get along with each other, with the format, and with the container? Some ingredients react, some compete for the same solubility budget, some degrade in the presence of others over shelf life. A mineral that’s stable in a dry capsule can misbehave in an acidic liquid.
• Does the dose physically fit the format? A four-gram active does not go into two small capsules. A poorly soluble ingredient fights a clear liquid and may force a suspension, a different format, or a different form of the ingredient.
• The liquid variables: solubility limits, pH, and preservation. In a liquid these problems show up immediately and get worse with concentration, which is the whole subject of emulsions and suspensions (a big reason a liquid-first floor tends to catch feasibility problems early, before they cost you a round).
• Taste load and masking difficulty. Bitter aminos, earthy botanicals, and mineral off-notes tell you up front that flavor is going to be the hard part and the formula needs room for it.
• Regulatory flags. Whether an ingredient is a new dietary ingredient that your counsel needs to evaluate, or carries its own constraints, is far cheaper to surface now than mid-project. Apollo flags it; your counsel decides it.

The value of feasibility is a fast, honest answer: sometimes “yes, straightforward,” sometimes “yes, but not like that,” and occasionally “not in this format.” Hearing “not like that” before you’ve paid for three bench rounds is the screen doing its job. It’s the difference between a formulator who wants the project to succeed and one who’ll happily bill you for rounds that had no path to converging.

The bench: two to three sample rounds, and what each one does

This is the part everyone pictures and almost no one describes accurately. “Multiple rounds of revision are common” is true and useless. Here is what the rounds are actually for. Expect two to three of them to dial in a formula, and more when flavor and masking are hard.

A round is a loop: the bench makes a small sample, you and the team evaluate it against the target, and it gets revised. What separates a good process from a slow one is that each round has a distinct job, so you’re not re-litigating the same thing three times.

• Round one: proof of concept. Does the formula come together, hold together, and land in the neighborhood of the target? This round is about structure, not polish. It’s often rough on taste on purpose; the question is whether the thing is buildable and stable enough to be worth refining.
• Round two: dial it in. Taste, texture and mouthfeel, color, clarity or suspension, and the first real read on stability. This is where flavor masking does most of its work, turning a functional formula into something a customer will actually take twice. Concentrated liquids and shots are the hard case, because you’re masking a heavy active load in a small volume; a functional shot is where this gets genuinely difficult and where extra rounds earn their keep.
• Round three: lock it. Confirm the sample is repeatable (that the bench can make it again and get the same result) and document the specification. What you sign off here becomes the master formula: the exact ingredients, grades, amounts, and order of operations that the pilot and every future batch are held to.

Why two to three and not one? Because the qualities you care about trade against each other. A first sample that tastes great but separates in a week isn’t done. A rock-stable sample that tastes like the raw active isn’t done either. Rounds exist to resolve those tensions deliberately instead of hoping they resolve themselves at scale. When a shop promises a finished custom formula in a single pass, it’s either working from a stock base it already owns or setting you up to discover the problems later, on your dime.

What you walk away with is a documented formula and spec, and if this is a true custom build, that formula is yours to own, per your agreement, rather than a shared base the manufacturer also sells to the brand next to you on the shelf. That ownership is the entire point of going custom, and it runs through Apollo’s custom formulation and R&D work.

Why a bench sample can fail at production scale, and what the pilot proves

Here is the seam the search results gloss over, and the single most important thing to understand about custom formulation: a sample that is perfect in a beaker is a sample at beaker scale. Production is a different physical environment, and the formula has to be re-proven there. This is not a formality. It’s the step that decides whether your launch date holds.

What changes between the bench and the line:

• Mixing dynamics. A lab stir bar and a production mixer don’t apply shear the same way. An emulsion or suspension that came together by hand can break, separate, or settle in a larger vessel, or it may need a different order of addition and a different mix time to hold.
• Temperature and time. A bigger batch heats and cools more slowly, and it sits through a longer process. A heat-sensitive active or a delicate flavor that was fine during a two-minute bench mix can drift over a production cycle.
• Aeration and foaming. Faster, more powerful mixing can whip air into a liquid that a gentle bench stir left alone, changing appearance, throwing off fill accuracy, and sometimes destabilizing the product.
• Fill behavior. A viscous, foamy, or particle-laden liquid that pipettes cleanly by hand behaves differently moving through a fill head at line speed. Hitting a small target amount repeatably is its own discipline, and it’s sensitive to exactly the properties formulation controls.
• Flavor at volume. Sweetness and masking that read perfectly in a small bench batch can read differently at production scale, which is why formulators expect to re-taste and adjust when they scale, not before.
• The real timeline. A sample you taste the day you make it hasn’t been tested by a week on a warm truck or a month on a shelf. Settling, separation, color shift, and potency loss show up over time, not on day one.

The pilot run is the answer to all of it. It makes the formula on real (or representative) production equipment, at a real but contained size: often in the hundreds of units when compatible materials are on hand. That’s small enough to be affordable and large enough to be honest. The pilot is where the translation problems surface while they’re cheap to fix, and it’s where you confirm the formula still meets its spec after it has been made the way it will actually be made, at the speed it will actually be made.

Skipping the pilot doesn’t save time. It relocates the risk to your first committed production run, where a broken emulsion or an off-spec batch costs materials, a slot on the schedule, and a launch window: the exact things a pilot is designed to protect. The run size itself is a separate conversation, because a pilot’s “minimum” is really two numbers, which Component Minimums vs. Finished Runs pulls apart. And what changes as you go from that pilot to full production volume is its own subject: From Pilot Batch to Full Production covers the scale-up mechanics in detail.

Custom formula vs. a stock base: what you’re actually buying

Custom is not automatically the right call. It’s one of two honest routes, and the trade between them is clean.

A stock or lightly customized base is fast and cheap to start. There’s no bench to pay for; a private-label fast start can reach finished goods in about four to six weeks when the base and components are ready, at the lowest first commitment of any path. What you give up is ownership and differentiation: the base isn’t yours, and the same one can sit behind other brands. For a lot of products that’s exactly the right trade, because the brand was always going to be the differentiator and the formula just has to be good and consistent.

A custom formula costs more and takes longer up front (you’re paying for the bench rounds and the pilot), and in exchange you get a formula that’s yours, owned as your intellectual property per your agreement, and a product that is genuinely distinct rather than a relabeled catalog item. When the product itself is the pitch, that’s the whole game, and the cost of development is the price of admission. If you’re weighing the two routes head to head, White Label, Private Label, or Custom Formula: Which Path Fits Your Supplement lays the decision out side by side; this article is what the custom route looks like once you’ve chosen it.

Now the timeline question, honestly. There is no fixed number of weeks for a custom build, and a shop that quotes you one before running feasibility is quoting a template, not your product. What sets the clock is a short list of hard variables:

• How hard the flavor is to mask. Easy actives converge in a couple of rounds. Bitter or earthy ones add rounds.
• Whether the formula needs stability data before you can set a shelf life. A defensible expiration date can require real time on test, and you can’t shortcut a claim you haven’t substantiated.
• Material lead time. A common ingredient ships quickly; a specialty or custom active can carry a meaningful lead time of its own, independent of the bench.
• Regulatory checks. An ingredient that triggers a review your counsel has to work through moves the date, and it should.

For liquids specifically, what drives that development cost and time gets its own breakdown in the liquid cost drivers. The useful takeaway: a real timeline is an output of feasibility, not an input to the sales pitch.

Reformulating a supplement vs. a clean-sheet build

Not every custom project starts from nothing. A lot of them start from a product that already sells and a specific reason to change it, and that’s a different job from inventing one, with different risks.

A clean-sheet build starts from a target you’re defining and a blank formula. A reformulation starts from a product that already exists and a constraint you can’t move. The reason for the change narrows the work, which usually makes it faster, but it also fences you in. Common reasons brands reformulate:

• Cost-down: an equivalent active or a process change that protects the margin without changing the label claim or the experience.
• An ingredient you have to remove: a supplier discontinued it, a retailer or marketplace banned it, or a clean-label push means an additive has to go.
• A supplier change: the same ingredient from a new source can behave differently enough to require re-work and re-testing.
• A taste or stability complaint: the product sells but separates, clumps, browns, or drifts off flavor before its date, and the formula has to be fixed without changing what customers already like.
• A claim or label change: a new positioning or a dose change ripples back into the formula.

What’s easier about a reformulation is that “right” is already defined: you have a benchmark, the current product, to match or beat. What’s harder is that you’re boxed in by what can’t change: the label claim, the price point, the attributes customers already expect. You’re threading a needle, not painting on a blank canvas.

And there’s a trap almost everyone forgets. Change the formula, the process, the packaging, or the manufacturing site, and you may re-open stability testing and have to re-establish the shelf life. A swap that feels small on paper (a new flavor system, a different bottle, a new supplier for one active) can reset the shelf-life clock and add time you didn’t budget. Surface those triggers before you commit, not after.

One more distinction worth drawing: if you already own a finished recipe and simply need it made or scaled (not developed), that isn’t custom formulation at all. It’s bring your own formula, where the work is spec review, sourcing, and a pilot rather than bench development. The first thing to confirm on that path is that you actually own the recipe; formulas built under a prior private-label deal often stay with the manufacturer that made them, and ownership is a question for your counsel.

The formulation brief checklist

Bring this to a first conversation and the bench can start on round one. Copy it into an email; the blanks you can’t fill are the questions worth talking through before anyone quotes you a timeline.

The target

• What the product is, and who it’s for, in one or two lines.
• The structure/function story you want to tell (direction, not final copy).
• A benchmark product: a competitor, a taste, or a texture you’re chasing.

The actives

• The key ingredients, with rough doses.
• Which are must-haves and which are flexible.
• Any branded or patented actives you intend to use.

The format and pack

• Preferred format, or “open to what fits.”
• Any packaging you already have in mind (and whether it’s stock or custom).

The constraints

• A target cost-per-unit ceiling.
• Sales channel (retail, DTC, marketplace, practitioner).
• Allergen, vegan, and clean-label rules you have to meet.
• Target shelf life.

The flags

• Any ingredient you suspect is novel or restricted (for your counsel to check).
• Anything you must avoid: an allergen, an excipient, a supplier.

The reference

• The run size you’re aiming for, so the pilot and the minimums can be scoped.

You won’t have every line, and you don’t need to. A brief that’s honest about what’s fixed and what’s open is worth more than one that’s confidently wrong. The gaps are where the feasibility conversation starts.