A request that reads "please send a stabiliser sample" is not a request anybody can answer well. It names a category, not a problem. What comes back is a reasonable guess, and reasonable guesses fail at the first trial more often than not.

The difference between a sample that settles a question and one that wastes a production slot is almost entirely decided before anything is despatched.

WHAT A USEFUL REQUEST CARRIES 1. The application What you are making 2. Process conditions pH, heat, shear, time 3. The outcome wanted Measured, not adjectives 4. The constraints Label, cost, plant 5. Scale and timing Volumes and dates A first trial worth running
A request missing any one of these usually produces a sample that cannot be judged, because nobody agreed in advance what success looked like.

1. The application, in specifics

Not "a beverage" but what kind, at what solids, sold in what pack, held for how long, at what temperature. The same functional ingredient behaves differently in a retorted drink and a cold-filled one. Naming the category narrows almost nothing.

2. The conditions it has to survive

This is the part most often left out and the part that most often decides the answer. At minimum:

  • pH of the finished product, and whether it shifts during processing
  • Heat treatment, with temperature and hold time, not just the word pasteurised
  • Shear, including homogenisation pressure if there is one
  • Water activity or moisture, where it matters
  • Anything already in the formulation that competes, such as salts, acids or other hydrocolloids

An ingredient that performs at pH 4.2 may do very little at pH 6.2. Knowing this in advance changes what gets sent, and often changes the recommendation entirely.

3. The outcome you want, measured

"Better texture" cannot be tested against. A target viscosity range, a firmness value, a sedimentation limit after a stated period, or a shelf life in days at a stated temperature can be. If you do not yet have a number, say what you would accept as a pass, and how you intend to judge it.

A useful test. Write down how you will decide the trial succeeded. If you cannot write it in one sentence with a number in it, the trial will end in opinions.

4. The constraints

Constraints are not obstacles to be discovered later. They are part of the brief.

  • Label requirements, including anything you are trying to move away from
  • Cost ceiling per unit or per tonne of finished product
  • Plant realities, such as what dosing equipment exists and where in the process it can be added
  • Regulatory positioning for the market you are selling into

5. Scale and timing

Indicative volumes and the date you need an answer by. This is not a commercial question at the sample stage. It determines whether a bench sample is enough or whether the trial needs quantities that behave the way production behaves.

What good looks like

A useful request is often six or seven lines. It might read: a 200 ml cold filled dairy drink at pH 6.6, 12% total solids, UHT at 137 degrees for 4 seconds, homogenised at 180 bar, currently separating within three weeks against a required twelve, cost ceiling in place, needs to stay clean label, decision needed within six weeks.

That can be worked on immediately. The category alone cannot.