Soft-Serve Sweetener Testing: Overrun, Draw Hold and Melt
Direct answer: A soft-serve mix can taste sweet enough and still fail the next operating check. Overrun, draw state, holding time and melting are related, but they are not the same result.
Published by Guilin Hengliyuan Biotech
This guide is for product developers and ingredient buyers comparing monk fruit, stevia, allulose-supported, erythritol-free or broader natural sweetener systems in dairy, plant-based and private-label soft-serve projects.
Sweetness is only the first gate
High-intensity sweeteners can restore perceived sweetness without restoring the solids, freezing behavior or air structure that a soft-serve mix needs. A 2025 Wageningen model separated ice content, unfrozen-phase viscosity, hardness and melting parameters, while a 2025 Fiordilatte study compared overrun, melting, emulsification stability and sensory results together. These sources support a better test list; they do not predict a finished commercial mix. Sources: https://edepot.wur.nl/696739 and https://doi.org/10.1016/j.idairyj.2025.106339
Record overrun as an operating variable
Overrun describes the air incorporated during freezing. As a simple arithmetic example, 50% overrun means one litre of mix produces 1.5 litres of finished product; it is not a universal quality target. Pump or gravity feed, mix viscosity, temperature, beater condition and machine settings can all change the observed value. Compare the same mix on the same machine before using the number to explain texture or serving cost.
Separate draw hold from timed melt
The product coming from the nozzle is one observation. The shape after a project-defined service interval, the first visible liquid, the later melt stage and the serving temperature are separate observations. Keep day 0 or immediate draw notes apart from timed checks such as the buyer's agreed 5-, 10- or 20-minute window; those example time points are test design choices, not delivery limits or universal benchmarks.
Keep the real machine and base in the test
Write down the final format, dairy or plant-based base, fat and protein, total solids, stabilizer/emulsifier system, mix temperature, freezing or hardening route, machine model or feed type, draw temperature, serving format, package and service or delivery window. The 2026 Northern Ireland Food Standards Agency and CAFRE guidance is useful for reformulation planning, but its scope is Northern Ireland and its examples are not a global formula or legal answer. Source: https://www.gov.uk/government/publications/ice-cream-reformulation-and-technical-guidance
Use three comparison routes
A workable first round usually keeps the current or full-sugar control, adds a route that mainly restores sweetness, and adds a system route that changes sweetness together with solids, body or viscosity. For a plant-based project, a 2025 response-surface study of coconut milk, date syrup and FOS also measured overrun, melting, physical properties and sensory response in one model. That supports broader observation, not the study's ingredient combination as a soft-serve recipe. Source: https://doi.org/10.1007/s44187-025-00282-0
Build a scorecard a supplier can read
Use separate rows for sweetness onset, peak and aftertaste; mix viscosity; overrun; draw appearance; extrusion or hold shape; first liquid; melt stage or rate; temperature; iciness or creaminess; package condition; and storage change. Add the main defect in plain language, such as weak hold, fast first drip, long aftertaste or poor plant-base body. A supplier can choose a better sample direction when the defect is tied to a base and process rather than only to the word 'sweet'.
References and boundary
The test logic uses the Wageningen model study, the 2025 Fiordilatte study, the 2025 plant-based ice cream study, the Northern Ireland FSA/CAFRE guidance and public Arla Pro soft-serve operating fields. Arla's page is a supplier example, not independent proof. None of these sources provides a Hengliyuan dosage, a universal overrun or melt threshold, a shelf-life result, a delivery guarantee or automatic market approval. China projects still need a product-category and ingredient review against GB 2760-2024. Sources: https://www.arlapro.com/en-gb/arla-pro-soft-serve-ice-cream/ and https://www.nhc.gov.cn/sps/c100088/202403/bda120e678df4a49a8beb90852559d7c.shtml
Related Hengliyuan products
Send a soft-serve sample brief
Frequently asked questions
- Does a sweet soft-serve sample automatically have good overrun?
- No. Sweetness, air incorporation, mix viscosity, machine settings and draw conditions are different checks. Compare overrun and draw state in the real mix and machine.
- Is higher overrun always better for soft serve?
- No. Overrun changes volume, mouthfeel and melting behavior, but a higher value is not a universal quality target. The buyer's format, machine, serving style and sensory target still matter.
- Can a short melt test prove delivery stability?
- No. A timed melt observation can describe one test condition. It cannot prove shelf life, transport performance or a universal delivery window without the actual package, temperature profile and validation plan.
- What should buyers send before requesting a soft-serve sample?
- Start with your target market and product idea. Share any known packaging specification, estimated quantity, required certificates/documents and whether samples are needed; we can clarify the remaining details together. Add base, machine, draw condition, service window and main defect if available.
What should buyers verify before a commercial discussion?
This is a buyer education Resource. It explains a natural sweetener, application, sourcing or validation question before a commercial discussion.
Product fit, label treatment, regulatory status and finished-product performance still need target-market and real-base checks.