Read this guide in
SOY BEVERAGE STABILITY SCORECARD
Soy Milk Stability Guide: Protein, Oil Droplets & Cold Storage Scorecard for Buyers
Record soy beverage controls, undisturbed sediment, re-mixing, physical methods and actual storage checkpoints. Download a free paired Excel record.
Use this scorecard when a soy beverage looks uniform at first pour but the team still needs to understand standing, sediment, re-mixing or cold-storage behavior. It keeps sweetness, formulation variables and physical observations visible in separate rows.
The scorecard is designed for a defined product, process, package and target market. It is not a universal stability test, shelf-life protocol or approval limit.
Define the product and control
Write the product format, soy source, protein or solids direction, flavor, serving condition and target market. Keep a current or sugar-containing control beside the candidate sample. If the control and sample use different bases or processes, note that before reading the result.
Record the structure variables
Capture filtration or separation, particle or droplet clues when measured, pH or ions when relevant, viscosity or cloud method, mixing order, homogenization, heat-treatment sequence, cooling, fill and package. These fields help explain a visible result without assigning it automatically to the sweetener.
Make the comparison traceable before recording a result
Our rule is to keep the actual control and candidate visible together. A new soy base, mineral premix, dilution, heat sequence or package is a formulation or process change even when the purchasing question started with a sweetener. Record it rather than assigning the whole difference to monk fruit or stevia.
| Comparison field | Retain with both samples | What the record prevents |
|---|---|---|
| Identity and composition | Control/candidate code, lot, formula revision, soy base, actual sugar system and every other change | A different base or revision being treated as the same sweetener-only trial. |
| Process and package | Actual mixing, separation, homogenization, heat/cooling/filling sequence, vessel and fill condition | A processing or package change disappearing from the comparison. |
| Checkpoint and handling | Defined clock start, actual elapsed time, temperature/exposure history, pack identity and opening or sampling history | A calendar label or repeated reading being mistaken for an independent stored sample. |
| Method and criterion | Observation or analytical method, preparation, units, project criterion and any missing result | A visual impression becoming a laboratory result, a universal pass limit or a fabricated zero. |
Observe standing samples before re-mixing
Look at the undisturbed package first. Record bottom sediment, an upper creamy layer, other separation, cloud or color change, and the photo identity with the actual viewing setup. Then perform the agreed re-mixing action and record what remains. Keep the motion, duration or count and the vessel consistent enough for the intended comparison. A sample that looks uniform after shaking has answered a re-mixing question. It has not erased the standing observation or established why separation occurred. If recurrence after re-mixing matters, define that observation and its actual elapsed time separately.
Keep the five observation tracks separate
Use the relevant rows instead of collapsing everything into a single stability score. Leave an unmeasured result blank or explicitly unmeasured. A real zero needs a defined method and unit; not applicable is a different entry.
| Observation track | Useful record | Conclusion it cannot provide alone |
|---|---|---|
| Undisturbed appearance | Sediment, upper layer and other separation before movement; photo, geometry, temperature and elapsed time | Protein identity, mechanism, microbial safety or commercial shelf life. |
| Re-mixing | Actual re-mixing method, remaining material and later recurrence if checked | The original standing condition or the cause of a visible layer. |
| Protein measurement | Named method, sample preparation, measured fraction, units and result for the identified sample | A result interchangeable with a glass-bottom photograph or another method. |
| Other physical measurements | Actual particle-size, viscosity, pH or other selected method with preparation, temperature, units and result | Proof that all measured particles are oil droplets, or that one value explains every defect. |
| Taste and mouthfeel | Matched preparation and serving temperature; sweetness, aftertaste and gritty or thin sensations separately | A physical stability, process, safety or storage approval. |
Define the storage route and the sample unit
Keep unopened chilled, unopened ambient and after-opening records separate when those routes are relevant to the defined product. A product intended and validated for one route does not acquire another route from this scorecard. Record actual exposure and opening history rather than supplying a universal refrigerator temperature or storage calendar. Decide which observations use an untouched pack and which need a separately identified portion or retained pack. Opening, pouring, shaking or a destructive analytical step changes the sample history. Multiple readings or photographs of the same pack are repeated observations, not independent packs or batches. Keep a method-specific baseline and actual elapsed times with later readings. This record does not decide whether a stored sample is safe to taste. Physical appearance, re-mixing and sensory checks belong alongside the project's separate food-safety and shelf-life arrangements; they cannot replace them.
Read analytical research at the method level
The 2024 study of commercial plant-based drinks used several different measurements during storage. Its protein-solubility work compared nitrogen measurements before and after a defined centrifugation step; particle-size work used its own preparation and instrument. That is a useful reason to retain method and sample preparation with a number. Those measurements are different from watching sediment at the bottom of a bottle. We use the paper to explain that distinction, not to copy its storage duration, instrument settings or acceptance limits into a new soy beverage project. It supplies no Hengliyuan sweetener dose, chilled-storage guarantee or universal oil-droplet result.
Download the paired soy beverage checkpoint record
The free Excel has 32 paired fields and 64 blank control/candidate input cells on one sheet. It keeps sample identity, composition, process, package, actual checkpoint, handling, standing appearance, re-mixing, selected physical methods, taste and the next decision together. Use a separate copy for each matched comparison at a storage checkpoint and retain the same sample identities across the records. It is a blank comparison form with no automatic score, formulas, example results or universal pass values. Enter the actual method and units with any measured value. It downloads without an email address or account, and completing or sending it is optional for first contact.
Let the failed row choose the next sample
If sweetness fails, keep the physical system steady while narrowing the sweetener comparison. If sediment or layering changes, review the base, particle or droplet state, mixing, homogenization, heat history and storage. If aftertaste changes only in the soy base, keep the real base in the next sensory track. Change one main decision at a time. Keep the current control while testing the next decision. If several materials or process steps changed together, report the observed difference without crediting one sweetener for it. The existing soy formulation guide explains the broader trial logic; this page and its worksheet keep the actual comparison record usable.
Start with the product, market and observed problem
Send the soy beverage format, intended sales market and the main observed problem. Add the base, control/candidate difference, process, package, storage route and available records. Unknown details can be clarified; the worksheet, a completed test program and every purchasing field are not prerequisites to contacting us. Keep ingredient specification, proposed product name and label claims in the market review alongside the sample discussion. A physical observation or published study does not approve the formula or label. Packaging, estimated quantity, required documents and sample arrangements can follow as the proposed ingredient or blend scope becomes clearer.
Useful fields to add as the comparison develops
- Product format, target market and main observed problem
- Control and candidate identity, soy base and actual changes
- Process, package and storage route
- Actual checkpoint, temperature, opening and handling history
- Undisturbed view before re-mixing, then the re-mixing method and result
- Selected analytical methods, units and missing measurements
- Matched sensory preparation and the next decision
Common questions
What should a soy beverage stability scorecard include?
Keep sample and formula identity, actual composition and process, package, storage and handling history, undisturbed appearance, re-mixing, selected physical methods and sensory observations. Retain methods, units, missing results and the next decision beside the control and candidate.
Discuss this requestDoes visible sediment identify a protein problem?
No. Visible sediment is an observation. A protein-related conclusion needs a suitable analytical method, sample preparation and result; a bottle photograph cannot identify the material or its mechanism by itself.
Discuss this requestShould we shake a sample before photographing sediment?
Record the undisturbed view first. Then apply the agreed re-mixing method and record its separate result. Keep any later recurrence and its actual elapsed time identifiable.
Discuss this requestDoes a particle-size result measure only oil droplets?
Do not assume that. Retain the actual method, sample preparation and what the method characterizes in the defined soy beverage. Keep a generic particle measurement distinct from an identified oil-droplet result.
Discuss this requestCan the same opened bottle be counted as a new replicate at every checkpoint?
No. Re-reading it provides repeated observations with an opening and handling history. Identify separate packs or batches when those are the intended independent units, and retain separate portions for disruptive measurements.
Discuss this requestIs a uniform first pour enough to conclude cold-storage stability?
No. It describes that sample at that observation. Use the intended product route, package, actual checkpoints and exposure history; appearance alone does not establish safety or commercial shelf life.
Discuss this requestDoes the Excel calculate a stability score or supply pass limits?
No. It contains 32 paired recording fields with blank inputs and no formulas or automatic ranking. Use the project's agreed criteria and actual methods, and distinguish missing results from zero and not applicable.
Discuss this requestMust we complete the worksheet before requesting a sample?
No. Start with the soy beverage format, target market and main problem. Share the base, process and observations you have; the record and further packaging, quantity, document and sample details can follow.
Discuss this request