Soft-Serve Overrun, Draw Hold and Melt: Sweetener Testing Guide

Sweetness is only the first gate

Our view is simple: match sweetness, then judge the operating result separately. Monk fruit or stevia can restore perceived sweetness without restoring the original solids, freezing behavior or air structure. A cone that tastes right at the nozzle can still lose its shape during service. A 2025 Wageningen model study examined ice content and the viscosity of the unfrozen continuous phase in relation to hardness and melting. Those are separate properties, and the study supports keeping separate measurements. It does not select a commercial soft-serve recipe, prove a winning sweetener or provide results for your machine.

• Original model study: ice content, serum viscosity, hardness and melting

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.

Set the comparison before changing the sweetener

Keep a current or full-sugar control that represents the product you actually sell or are developing. A sweetness-only replacement helps reveal what sweetness adjustment leaves unresolved. A system trial can change bulk, body or viscosity as well, but record those changes together; it cannot isolate the effect of one ingredient. Use the same intended machine and comparison method. If a machine, preparation or setting has to change, label the result as a different operating condition rather than silently ranking it against the original run.

Use the actual mix as the overrun reference

For a suitable uniform product, weigh the same fixed volume of liquid mix and drawn soft serve, subtracting each container tare. Calculate overrun (%) = (net mix mass / net soft-serve mass - 1) × 100. Use each formulation's own mix reference and record the fill and leveling method; do not borrow the control's density for every candidate. The Guelph textbook describes equal-volume weighing and warns that inclusions such as nuts change what package weight tells you. Establish an appropriate sampling method before comparing particulate products. The free workbook below checks positive equal cup volumes and positive net masses; it cannot detect poor filling, collapsed air or an unrepresentative sample.

• University of Guelph: overrun calculations and particulate limits

Give each draw, hold and melt result its own endpoint

Record the immediate draw before describing what happens later. Start the observation clock at a defined event, such as completion of dispensing, and retain the actual elapsed time. Choose checkpoints around the intended service question; 5, 10 and 20 minutes are possible planning examples, not a required schedule or universal acceptance limits.

For a rate over an interval, keep both collection times and the mass change with the elapsed duration. If you express collected melt as a percentage, identify the starting sample mass used as the denominator. A sample that has not dripped by the last observation is recorded as not observed by that time, not as a zero-minute result or an unlimited hold claim.

Keep repeats and the service route recognizable

Record separate portions and runs so a repeat can be distinguished from another photograph of the same cone. Repeated observations of one portion show its history; they are not independent samples. If a measurement or tasting changes the portion, use a separately identified portion for the next comparison rather than treating the disturbed sample as untouched. Record waiting, packaging, loading, route, opening and arrival observations for an actual delivery comparison. A counter shape-hold test does not reproduce that route. Keep any chilled holding, heat-treatment and cleaning procedure within your validated operating and food-safety arrangements; this article supplies no universal service temperature, safe time or commercial shelf life.

Download the existing comparison and overrun Excel

Use the frozen-dessert worksheet to keep the control and candidate records together and calculate overrun from your own equal-volume measurements. It has 42 blank inputs, six formulas and 15 paired record fields. It checks missing inputs, unequal or non-positive cup volumes and non-positive net masses; a real zero-overrun result is retained. The file downloads directly without an email address or account. It is a comparison and overrun record, not an automatic machine qualification or a complete delivery-validation form. Attach the timed shape, collection and route observations with their sample codes when they need more space.

• Download the free frozen-dessert comparison and overrun calculator (.xlsx)
• Read the frozen-dessert scorecard and workbook methods

Choose the next trial from the defect you observed

A coarse draw, fast shape loss, early liquid and a slow-clearing sweet finish are different problems. Keep texture and temporal sweetness notes separate so the next sample addresses the actual defect. When several formulation or process variables changed together, describe the observed difference without assigning it to one sweetener. Use the existing controlled ice-cream trial guide for the ingredient comparison logic, then evaluate the soft-serve result on the intended equipment. A softer sample or higher overrun is not automatically a better product; judge it against the intended portion, eating quality and service route.

• Plan the controlled allulose and erythritol ice-cream comparison

Send the product, market and main defect first

Start with the finished format, target sales market and the problem the next sample should solve. Add the base, known sugar system, preparation, machine, measured conditions and observations you have. A completed worksheet, every purchasing field and a finished test program are optional at first contact. For an allulose-supported or erythritol-free direction, include the proposed composition and label plan for market-specific review; include the main sales states for a U.S. allulose project. The sample discussion does not settle final food-use, label, process, shelf-life or delivery approval. The separate supplier brief covers quotation scope, packaging, quantity and included costs.

Common questions