Angelica and Astragalus Chicken Soup OEM/ODM | Custom Recipe, Private Label Bowl and Shelf-Life Validation

A chicken soup in a self-heating bowl is not a soup recipe with a heater attached. The bowl, the fill weight and the retort schedule are three parts of one specification, and changing any of them re-opens the shelf life.

That is why this product is usually lost at the second order rather than the first. The first run is watched closely. The second run is scheduled like any other SKU, the bowl format has drifted by a millimetre, and the validated process no longer describes what is actually in the retort.

The first three months of a soup launch are usually paid for by novelty. The next three are paid for by logistics, and that is where the arithmetic stops working for brands that did not cost the format before they printed the carton.

Four costs do the damage, and none of them appears on a factory invoice. They appear on a freight bill, a credit note, a rework line and a slotting fee.

A 258 g bowl is a heavy unit for its retail price. Water is the cheapest ingredient and the most expensive thing to move, so a soup shipped across an ocean carries a freight cost per unit that a powder or a concentrate does not. The brands that survive this either price for it from the first listing or move production closer to the market.

A self-heating bowl has an outer shell, an inner container and a heater between them. That structure resists vertical load well and resists a corner impact badly. Damage rates are set by the carton, the divider and the pallet pattern, not by the bowl alone, which means a packaging change made to save cost usually shows up as returns three weeks later.

A kitchen recipe is written for four litres. A production batch is written for a retort basket. The translation between them changes the ratio of solids to broth, the size of the chicken pieces and the amount of free liquid, and a soup that tastes right in a test kitchen can arrive thin at production scale.

Retailers rarely take one SKU. They take a flavour set, which means the second and third recipes have to be developed while the first is still being validated. Brands that planned one product and were asked for three discover the cost of development at the worst possible moment.

It needs a format decision made before the recipe is locked, and a process schedule written per format rather than per recipe. Those two decisions cost almost nothing at the start and are the most expensive things to change afterwards. Everything else in this article follows from them.

This is a black chicken soup with angelica root, astragalus root and a supporting set of ingredients that Chinese households have cooked with for generations. The order matters because the roots need time that the aromatics do not.

Black chicken, ginger, red dates, longan and wolfberry form the base. Black chicken is a small-framed bird with darker skin and less fat than a commercial broiler, which is why the broth is lighter and why the meat has to be cooked gently to avoid turning dry. Ginger goes in early for aroma, the dates and longan later because their sugars scorch against a hot base.

Angelica root and astragalus root are the two named botanicals, with yellow essence (Polygonatum) and black beans completing the set. These ingredients sit inside what Chinese food culture calls food-medicine homology, meaning materials recognised in the tradition as both food and medicine. Naming them correctly and sourcing them by species and grade is the difference between a repeatable product and a stew.

The tradition gives a brand a story and does not give it a health claim. What can be written is what is in the bowl and how it has been eaten: a chicken broth with named roots, cooked in a household way, now shelf-stable and self-heating. What cannot be written is a statement about what the bowl does to a body, and that boundary holds in every market we export to.

That constraint is not a limitation on the product. It is a limitation on the copy, and the brands that accept it early spend their marketing budget on the occasion rather than on a claim a regulator can strike.

The table below pairs the failures we see most often in shelf-stable soup with the specific control that prevents each one. Every row corresponds to a document or a parameter, not to an intention.

The first two rows are the ones that decide whether the product is any good. Separation and dry meat are both consequences of a mismatch between the fill and the process, and neither can be corrected after the bowl is sealed. A thicker broth is not a fix; it is a different recipe.

The last two rows are commercial rather than technical, and they are the ones that cost a listing. A batch that tastes slightly different from the approved sample is a quality argument. A label with the wrong net content is a compliance event, and it happens most often when a brand changes market without changing the artwork.

Only the label row. Everything above it is locked at the point the schedule is validated, which is why the specification has to be right before the first production run rather than after the first complaint. A brand that plans to fix the recipe after launch is planning to re-validate the process, and that is a project, not a tweak.

The retort is the part buyers ask about and the part that is easiest to describe wrongly. A temperature on its own says nothing about whether a bowl is commercially sterile, because the time at that temperature, the load in the basket and the behaviour of the coldest point all decide the result.

In a bowl with solid pieces and free broth, heat moves through the liquid by convection and through the solids by conduction. The slowest-heating point is therefore not the geometric centre of the bowl. It is set by the product and the format, and it is located experimentally before a schedule is written.

The schedule fixes a temperature, a holding time and a required accumulated lethality for that product in that format. The operator's job is to reproduce the schedule and record the evidence. The schedule itself is established by qualified persons with expert knowledge of thermal processing requirements, which is the standard 21 CFR 113.83 sets for low-acid foods in hermetically sealed containers. It is not adjusted on the floor.

The target has a name. Commercial sterility is defined in the Codex code of practice for low and acidified low acid canned foods, CAC/RCP 23-1979. The wording is "the condition achieved by application of heat, sufficient, alone or in combination with other appropriate treatments, to render the food free from microorganisms". That definition describes the whole process, not the temperature alone, which is why a chart showing 121 °C proves nothing until the time at the cold point is included.

The bowl has to survive the pressure and temperature of the cycle without losing its seal, and the headspace has to be set so the container does not deform beyond its specification. A bowl that passes the seal check before retort and fails after it has a packaging problem, not a process problem, and the two are often confused in a complaint.

Headspace is the quiet variable here. Too little and the bowl deforms under pressure as the contents expand; too much and the seal takes a load it was not designed for when the bowl cools. The setting is established during the packaging trial with the real fill, and it is one of the reasons a fill-weight change cannot be treated as a numbers-only edit.

The declared shelf life for this SKU is twelve months at ambient temperature, stored below 25 °C. That figure comes from a storage trial that holds samples of the actual product in the actual pack and checks them at intervals, together with a validated process that establishes the starting microbial state. One without the other proves nothing: a validated process on a poorly sealed bowl fails early, and a perfect seal cannot rescue an under-processed fill.

Customisation in this category runs along three axes, and buyers usually ask about all three in the same email. Separating them makes the answer shorter, because each axis has a different cost and a different approval path.

The broth base, the solids ratio, the piece size and the herb set can all be changed. What cannot be changed freely is the combination of recipe and format that has already been validated; a new recipe in an existing bowl needs a new schedule, and that is the cost most often left out of a quotation.

Bowl format, lid, sleeve, carton and the heater configuration are all decisions. The bowl is the expensive one, because a new shell means new tooling and a new thermal validation. The sleeve and carton are cheap and can change between runs.

Ingredient naming, net content, nutrition panel, allergen statement and language are set per destination market. We prepare the technical content and check the label against the destination rules; the brand's regulatory contact signs off. This is the axis where a mistake is cheapest to fix before print and impossible to fix after.

The boundary is validation, not willingness. A change that touches the thermal process re-opens the schedule, and a change that touches the bowl re-opens both the schedule and the packaging trial. We say which one applies before quoting, because a buyer who discovers it after the purchase order has been issued has lost a month.

The two routes, ODM and OEM, run on the same line with the same parameters and the same batch file. The difference is who owns the recipe, not which factory makes the bowl, and buyers who have been burned by subcontracted OEM orders tend to ask about this directly.

A brand that develops a soup and watches a competitor list something similar has a legitimate complaint and usually no contractual answer. Exclusivity is the answer, and it only works when it is specific.

A developed recipe can be exclusive to one customer in one market and one category for a defined period. The terms are written into the development agreement before the first sample is sent, not agreed verbally after the product succeeds. They name the market, the category and the period, because an exclusivity clause that says "our market" is not enforceable anywhere.

A catalogue formula cannot be made exclusive after the fact, and no supplier should pretend otherwise. Neither can a recipe built entirely from public ingredients be protected against a competitor who asks a different factory for the same soup. What exclusivity buys is time and a documented position, not a patent.

There is also a limit we impose on ourselves. We will not sign exclusivity for a market we cannot actually police, because a promise that cannot be checked is worse than no promise at all.

Every bowl carries a batch code, and every batch code points to a file. The file is the product's memory, and it is what a retailer's auditor, a customs broker or a customer complaint will eventually ask for.

The retort is the critical control point on this line. The basket loading pattern, the position of the thermocouples and the circulation of the heating medium all decide whether the coldest point reaches the scheduled lethality. The loading pattern is therefore a controlled document, not a suggestion to the operator.

The batch file holds the incoming lot sheets for the chicken and each botanical, the fill and seal records, the retort chart with the accumulated lethality, the sensory panel result and the release signature. The operator signs the run record, quality signs the release, and the file is stored under the batch code for the period the destination market requires.

Two details make the file useful rather than decorative. The first is that the retort chart is the original instrument record, not a transcription, because a transcription is only as good as the person who copied it. The second is that the processing records are reviewed for completeness before the batch is released rather than after a complaint, and the release decision is dated and attributable.

The file is also a traceability document. Article 18 of Regulation (EC) No 178/2002 requires that "food and feed business operators shall be able to identify any person from whom they have been supplied with a food". One batch code connects a bowl on a shelf to the supplier lot of the chicken and to the customer the pallet went to, with no gap in between.

A retort basket was loaded with one extra row of bowls. The pattern was within the weight limit, so nothing on the loading sheet flagged it. What changed was the gap between bowls, and with it the flow of the heating medium through the basket.

The deviation was caught at release, not by the operator. The thermocouple record for that cycle showed the coldest point rising more slowly than the validated profile, and the accumulated lethality came in below the value written on the process schedule. The batch was held in the warehouse rather than released, re-processed where the product quality allowed it, and the portion already dispatched under that batch code was replaced.

What changed afterwards was a document and a habit. The validated loading pattern was laminated and fixed at the retort mouth. Every basket is photographed before the door closes, and the photograph is filed with the cycle record. Loading density is written into the process schedule as a parameter rather than described as "full but not packed". The thermocouple positions are marked on the basket itself, so a new operator cannot place them from memory.

A bowl change is the most expensive request in this category, and the cost is not the bowl. It is the validation that follows the bowl, and the fact that the validation is per format rather than per recipe.

A new shell means new tooling, a new fill target and a new thermal profile. The heater configuration is part of that decision: the amount of heat released, the rate at which it is released and the venting of the steam all interact with the bowl geometry. Our 2021 design patent on the rotating self-heating bowl shell came out of exactly this problem, because the locking structure decides how evenly the heat reaches the contents.

The heater itself is a chemical reaction, not an electrical element. Water is added to a compartment containing calcium oxide, the reaction releases heat, and the steam it produces is what warms the food through the inner wall. That means the bowl has to vent in a controlled way, the inner container has to stay sealed while the outer one gets hot, and the outer surface has to be safe to hold at the end of the cycle. Each of those is a bowl design decision, and each one changes when the shell changes.

Net content and solids ratio are specified together. Raising the solids for a premium claim reduces the free liquid, which changes how heat moves through the bowl and can move the cold point. It also changes the fill height, the headspace and the sensory balance, so a solids change is a recipe change with a process consequence.

Every format gets its own loading pattern and its own thermocouple map. That is the reason a new bowl cannot simply be slotted into an existing production window: the pattern has to be validated on the real bowl with the real fill before the first commercial run.

The table is the honest version of a quotation conversation. The first row is a project. The last row is a purchase order. Buyers who know which one they are asking for get a useful answer on the first call.

We decline three categories. We do not take on medicinal products, meaning anything that has to be registered as a drug in the destination market. We do not take on concepts that only sell if the pack makes a disease claim. And we do not take on formulations built around an ingredient whose identity, origin or legal status cannot be documented, because that material cannot be defended at a border or in a retailer's audit.

Product formats are a different question. A format this line does not currently run is scoped as a project, with its own tooling and validation, and it is discussed before a quotation rather than ruled out in an email.

Four items make the first sample useful. They are the destination market, the bowl format and net content you intend to sell, the flavour direction with a reference product if one exists, and the price band the unit has to sit in. With those, development can start against a specification rather than a wish.

Send the reference product even if it is a competitor's, and say what you want different. A physical reference settles questions about broth darkness, solids visibility and herb intensity that no brief will answer in the same number of words.

The bowl platform behind this SKU took a SIAL Innovation Award at the Shenzhen show in September 2025, which is a useful signal about where the category is going and not a substitute for a specification. ZeaGrove (Guangzhou) Health Industry Co., Ltd runs this line in Guangzhou and exports most of its volume; the next step is a process schedule scoped to your format, not a price list.

The schedule runs at 121 °C, with the holding time set by the validated process for this product and this bowl rather than by a single figure we publish. What we can tell you is what the batch record shows: the retort chart, the coldest-point thermocouple data and the accumulated lethality against the value on the schedule. Those are the numbers an auditor will ask for, and they are specific to your bowl rather than to a generic soup specification.

By a storage trial on the actual product in the actual bowl, checked at intervals for microbiology, seal integrity and sensory change, together with a validated thermal process that sets the starting state. The trial is run per recipe and per format. A new bowl or a change in solids means the trial is repeated, because the shelf life belongs to the combination, not to the soup alone. That is also why a shelf-life claim cannot be borrowed from a competitor's pack.

Yes, for a developed recipe in a named market and category for a defined period, written into the development agreement before sampling begins. A catalogue formula cannot be made exclusive after the fact. We also decline exclusivity in markets we cannot police, because a promise that cannot be checked is worse than no promise. What the clause is worth depends on how narrowly it is drawn.

Medicinal products, concepts that depend on a disease claim, and formulations built on an ingredient whose identity or legal status cannot be documented. Those three sit outside what a food line can defend. A bowl format this line does not run is a different matter: it is scoped as a project with its own tooling and validation before anyone quotes, and that scoping is a normal conversation rather than a refusal.

The destination market, the bowl format and net content, the flavour direction with a reference product if you have one, and the target price band. Those four let us write a process schedule and a sampling plan. Without them, a sample is a guess that costs both sides a month, and the guess is usually wrong in the direction of more salt and less solids.