skininsider

What the skincare industry knows and does not tell you

Dispatch 6 August 2026Published independentlyNo affiliate links here
Formulation

Why a Lab Sample Changes at Production Scale

How a cosmetic moves from lab sample to production batch, what scale up proves, and what a brand should agree before release.

SectionFormulation
Reading10 min
Reviewed2026-10-09
EditorSkin Insider

A formula approved at the bench is a recipe proven at a few hundred grams, and the first production batch is that recipe remade on different equipment at a thousand times the volume. The brand is therefore signing off a process as much as a formula, including tolerances for acceptable change.

Scale up is the controlled route between a promising sample and sealed packs carrying a batch code. It matters because a cosmetic is not only a list of ingredients. Its final character depends on the vessel, mixing method, heating and cooling profile, filling line and chosen container. The formulator decides the recipe, the manufacturing team translates it to plant equipment, and the brand decides whether the agreed output is acceptable for release.

The bench sample is a formulation decision, not a production promise

The bench sample is usually made by a formulator or laboratory technician in small equipment. Ingredients are weighed in small quantities, phases are heated or mixed, and adjustments may be made quickly between versions. This is the sample a brand may assess for appearance, scent, thickness, spread and immediate feel on the skin. It proves that the proposed composition can produce a product with the intended broad characteristics under laboratory conditions.

It does not prove that a larger vessel will deliver the same result. A lab mixer can impart a different level of force from an industrial mixer. A small beaker may heat quickly and cool quickly. The person making it can observe and intervene throughout the process in a way that is not practical once material is being made at production volume. The sample also does not establish that the selected pump, tube, jar or bottle can fill consistently with the product.

The key decision at this stage belongs to the brand and formulator: which properties are essential and which are preferences. A pale cream, for example, may have a target appearance, but the decision should also state whether a slight variation in shade is acceptable. The same applies to scent strength, viscosity and spread. A sample should become a written product brief before it becomes a production instruction.

Related reading on this site explains how a brief becomes a formula and how ingredient percentages are chosen. Those pieces address the formulation decision; scale up asks whether that decision can be made repeatedly on the intended equipment.

Stability and compatibility work test time, storage and the pack

After a formula is selected, laboratory work commonly examines how it behaves over time and in contact with the proposed packaging. The work is arranged by the laboratory or manufacturer responsible for development. It may involve observing samples held under specified conditions and checking for changes such as separation, colour change, odour change, loss of viscosity or an altered appearance. Packaging compatibility work looks for interaction between the cosmetic and its container, closure, pump, liner or applicator.

This work can support the decision that the formula and pack are suitable to progress. It can identify obvious instability or a material that is not appropriate for a particular container. It also gives the formulator information needed to refine processing instructions before a larger batch is attempted. The role making the technical judgement is the person or team responsible for development and quality assessment, while the brand decides whether to fund further changes or choose another route.

It cannot prove every condition a finished product will meet after manufacture. It does not reproduce every mixing pattern in a large vessel, every pause before filling, or every variation in filling line operation. Nor can a test programme make a future batch identical to a bench sample. The site’s separate guide to what happens in a stability test explains those limits in more detail.

Compatibility is especially relevant at scale because the container is part of the finished product. A formula may be satisfactory in a laboratory container yet behave differently when pushed through a pump, exposed to a larger headspace, or held against a particular closure material for its intended life.

The pilot batch turns the recipe into an operating method

A pilot batch, sometimes called a scale up batch, is an intermediate manufacture intended to test the process before the first full run. It is made by the manufacturing team using equipment closer to, or the same as, the equipment planned for production. The manufacturing lead decides practical settings such as the mixing sequence, mixing time, temperature range, cooling approach and points at which checks are taken. The formulator may support this translation because the order and method of addition can affect the result.

The pilot proves more than the formula alone. It tests whether raw materials can be handled in the proposed vessel, whether the batch can be mixed uniformly, whether it can be transferred and whether it can be filled into the intended pack. It can reveal that an apparently minor instruction needs greater precision, such as when a thickener is introduced, how long an emulsion is mixed after cooling, or the temperature at which fragrance is added.

A pilot cannot guarantee that every subsequent full production batch will match it. It is still one run under particular conditions, using particular deliveries of raw materials and a particular equipment set up. Its value lies in converting observations into a controlled manufacturing instruction and agreed release checks.

StageUsually made byWhat it can establishWhat it cannot establish
Bench sampleFormulator or laboratory technicianInitial formula characteristicsPerformance on production equipment
Stability and compatibility workDevelopment and quality teamEvidence for progressing formula and packEvery manufacturing or storage condition
Pilot batchManufacturing team with technical supportPractical process and fill suitabilityPerfect repeatability of later batches
Full production batchManufacturing and quality teamWhether that batch meets release criteriaHow a product will be treated after sale
Retained sampleManufacturer or responsible supply partyA reference for the named batchA replacement for production records

Volume changes the physical conditions of the formula

At larger volume, the same written ingredient list can experience a different physical journey. Mixing shear is one example. Shear is the force created as material moves across a mixer or through a homogenising head. It can affect how droplets are reduced in an emulsion and how powders or gums disperse. A small lab mixer and a larger production mixer may not deliver the force in the same way, even where both are operated carefully.

The order in which phases are added also matters. Water, oils, emulsifiers, powders and fragrance may each need a defined point in the process. Adding a material too early, too late, too quickly or at an unsuitable temperature can change thickness, appearance or stability. The person who turns this into an instruction is making a manufacturing decision, not merely copying a recipe.

Heating and cooling are slower and less uniform in a large vessel. Material near a vessel wall, jacket or mixing zone can experience a different temperature history from material elsewhere. A batch may also sit for a period before it is filled. That hold time can matter where viscosity develops gradually, where air needs to disperse, or where a product is intended to be filled within a chosen temperature range.

These are reasons to record process settings, observations and checks. They are also why a brand should not treat a laboratory approval as a complete approval of the finished shelf product. The site’s guide to good manufacturing practice describes why controlled instructions and records matter at the factory stage.

Filling, headspace and the container finish the product

The product does not become a retail unit when mixing ends. It must move from the vessel through transfer equipment and a filling line into its final container. The filling team decides the set up within the documented process. Nozzle size and position, fill speed, product temperature and the way the product is cut off at the nozzle can affect a pack’s appearance. A thick product may form a peak. A fluid product may splash, trap air or leave residue around the closure area.

Headspace is the space left between product and closure. It is part of pack design, not unused detail. It can influence how a product looks in a clear pack, how easily a closure is fitted and how the pack copes with normal movement. The container itself can affect dispensing: a pump must draw the product, a tube must allow it to be squeezed out, and a jar must retain the intended appearance after filling.

The first production batch therefore tests the combined product and pack. It proves whether that particular batch, filled in that format, meets the agreed checks for release. It cannot prove that every unit will be handled gently in distribution or that every later batch will be identical without the same controls.

The relevant decision for a brand is whether it has approved the commercial pack as well as the formula. A product that looks or dispenses differently in its selected container is not merely a packaging issue. It may change the shopper’s use of the cosmetic and their perception of its quality.

Specifications and tolerances make acceptable variation visible

A specification is the written description of what a batch must meet before it is released. Depending on the product and arrangements, it can include appearance, colour, odour, viscosity, pH, fill quantity, microbiological requirements and packaging checks. A tolerance is the agreed range within which a measured or observed property may vary. Together, they replace an imprecise instruction such as “make it like the sample” with checks that different people can apply.

Some variation between a laboratory sample and a batch can occur without making the product unsuitable. Colour, scent strength, thickness and the way a product spreads can all shift. Natural variation in supplied materials, processing conditions and time can contribute. The useful question is not whether any difference exists. It is whether the difference falls inside a range agreed before production, and whether the range protects the product qualities the brand considers essential.

A certificate of analysis is a batch document reporting specified test results or confirmations against that batch’s requirements. Its exact contents depend on the product, testing plan and manufacturing arrangements. It can tell the brand what was checked, what result was recorded and whether the stated result met the relevant acceptance criterion. It is not a general guarantee about every characteristic a user may notice, nor does it replace review of manufacturing records and filled-pack checks.

If a result is outside specification, the batch should not simply be treated as acceptable because it is close to the target. The quality role must investigate under the manufacturer’s procedures. The outcome may be further assessment, a justified decision based on documented evidence, rework where appropriate and controlled, or rejection. The release decision belongs to the authorised quality process, with the brand’s contractual remedy decided separately in its manufacturing agreement.

Retained samples connect a sealed pack to its batch code

A retained sample is a sample of a specific batch kept as a reference after manufacture. It is normally held by the manufacturer or the party assigned that duty in the supply arrangement. It should be identifiable against the batch code and stored under the agreed conditions. Its practical purpose is comparison: if a question later arises about how a particular batch looked, smelt, dispensed or behaved, there is a controlled reference from that batch rather than only a memory of the original laboratory sample.

A retained sample can show what was kept from the batch at the time of retention. It can support an investigation alongside batch records, test results and packaging records. It cannot, by itself, explain every difference found in a pack after it has travelled through storage and use. Nor can it prove the condition of every unit made during a run. It is one part of a traceability system.

The party setting the retention arrangement should decide in advance what is retained, how many units or amount of bulk material are held, where it is held, how it is identified and for how long. Those details should match the product format and the responsibilities set out between brand and manufacturer. The batch code is the bridge between a retail pack and the records for its manufacture. The site’s batch code guide explains what a code can and cannot reveal to a shopper.

For a brand, the important point is that retention should be agreed before release, not improvised after a question arises. A retained sample is most useful when its identity, storage and associated documentation are already clear.

Questions to settle before committing to a production run

The decision to proceed should be based on written answers, not an assumption that a bench sample will reproduce itself. The brand makes the commercial commitment, while the formulator, manufacturing lead and quality release role each make distinct technical decisions. Ask the questions below before approving a run, and record the answers in the manufacturing and quality documentation.

  1. Will there be a pilot batch? Establish whether one is planned, what it is intended to test, who will review it and which party bears the agreed cost.
  2. What written specification applies? Ask for the release checks, target characteristics and tolerance ranges for appearance, colour, scent, viscosity, pH, fill and pack performance where relevant.
  3. Which process settings are fixed? Confirm mixing equipment, order of addition, temperatures, hold times, filling conditions and the action to take if an operating range is missed.
  4. What will the certificate of analysis report? Clarify which results, observations and acceptance criteria will be shown or made available for the batch.
  5. What is retained and for how long? Confirm the retained sample format, identification, storage conditions, holder and retention period.
  6. Who signs release? Identify the quality role responsible for determining whether the batch meets the agreed release requirements.
  7. What happens if the batch is out of specification? Agree the investigation route, decision authority, documentation and contractual remedy before material is made.

Limits of this guide

This guide concerns present day cosmetic development, manufacture, filling and batch release decisions. It applies to brands considering a production run and to readers trying to understand why a retail product can differ slightly from an approved lab sample. It does not assess individual factories, practitioners, companies or named products. It does not tell a reader which cosmetic to buy.

It also does not replace a product safety assessment, legal labelling review, stability programme, manufacturing agreement or quality system. Requirements and records can differ with the formula, packaging format, route to market and responsibilities assigned between parties. A batch code and retained sample support traceability, but they do not disclose a complete production history from the pack alone.

This is not a guide to reporting a reaction, seeking redress, evaluating clinical treatments or checking the credentials of an individual. It stays at the shelf and manufacturing chain: the product, the pack, the documented process and the decisions made before release.

Questions readers ask

Does an approved lab sample guarantee the production product will match it?

No. A lab sample shows that the selected formula can achieve the intended characteristics in laboratory conditions. Production uses different equipment, larger volumes, different heating and cooling behaviour, transfer steps and a filling line. A pilot batch and written tolerances are the practical way to decide what variation is acceptable before the full run is released.

What is the main purpose of a pilot batch?

A pilot batch tests the manufacturing process, not just the ingredient list. It can show whether the proposed mixing sequence, temperature profile, hold time, transfer method and filling set up produce an acceptable product in the chosen pack. It also helps turn laboratory observations into a repeatable manufacturing instruction.

Why can a cream become thicker or thinner at production scale?

Thickness can change because mixing force, ingredient dispersion, temperature history, cooling rate and time before filling change with vessel size. Some formulas also develop viscosity over time. A specification should state how thickness is assessed and the acceptable range, rather than relying only on a visual comparison with the bench sample.

What does a cosmetic batch certificate of analysis tell a brand?

It records results or confirmations for the checks included in that batch’s specification. It can show whether reported results met stated acceptance criteria. Its contents vary by product and manufacturing arrangement. It does not by itself describe every production event or guarantee every sensory detail that a user may notice after opening the pack.

Can packaging change how the finished cosmetic performs?

Yes. The container, closure, pump, nozzle and headspace affect filling, dispensing and appearance. A product may be suitable in one format but difficult to pump, squeeze or fill neatly in another. Compatibility work and production filling checks examine the formula and pack together, rather than treating the container as an afterthought.

What should happen when a production batch is outside specification?

The batch should be held for review under the manufacturer’s quality procedures. The quality role investigates the result against the documented specification and production records. The possible route depends on the evidence and agreed procedures, but release should follow a documented decision rather than an informal judgement that the difference seems minor.

Why keep a retained sample when the product already has a batch code?

The batch code identifies the production batch, while the retained sample provides a controlled physical reference from that batch. Used with manufacturing and test records, it can help compare a later question with the product retained at the time of manufacture. It cannot, on its own, establish the condition of every retail unit.

Who should sign off a first production batch?

The sign off should follow the responsibilities set out in the manufacturing and quality arrangements. The manufacturing team records how the batch was made, quality reviews the agreed release checks, and the brand decides whether to accept the commercial output within its agreement. These are separate decisions and should be identified before production begins.

Primary sources

Cited because they are public, stable and checkable. Read them rather than taking our word for any of this.

This is journalism about an industry. It is not medical, legal or regulatory advice. For a skin condition, see a clinician.

This article contains no commercial links. No affiliate links, no sponsored placements and no paid mentions appear anywhere on this site. No brand, manufacturer, retailer or agency is named in our editorial, and no company that pays us can appear in it. Published independently.

The dispatch

One email a fortnight. A new translation, a mechanism explained, and what changed in the rules. No product recommendations, because we do not make any. No affiliate links, because we do not take commission. Unsubscribe in one click.

We use your address to send the dispatch and nothing else. See privacy. Sponsorship rates and limits are published on the newsletter page.