The expensive thing about a substitute trial is not materials, it is the months. Teams usually start with a candidate suggested by a supplier and iterate on dose, and when nothing works they conclude the application is hard. Most of the time the failure was set up in the first hour: they picked a material that cannot do the dominant job, and no dose was going to rescue it.
The six questions below are in order for a reason. Each one eliminates candidates, and answering them out of sequence loses most of the benefit.
Question one: which job is dominant?
Write down, in one sentence, the single function the product cannot lose. Not a list of requirements — one. "Suspends cocoa." "Sets firm at ambient." "Keeps acidified protein in solution." If you cannot name one, you are not ready to choose a material yet, and you should go back to the product brief.
| Dominant job | Materials that do it well | Materials that do not |
|---|---|---|
| Suspend particles | Low acyl gellan, some blends | Anything that only thickens |
| Protect acidified protein | HM pectin, soluble soy polysaccharide | Gellan |
| Elastic chew | Gelatin, konjac blends | Most others |
| Aerated structure | Whipping proteins and gelatin | Gums generally |
| Heat-stable set | Low acyl gellan, agar | Gelatin |
This table alone resolves a surprising proportion of failed projects. Gellan does not protect acidified protein; no gum whips; nothing reproduces gelatin's elasticity.
Question two: what will the replacement give up?
Every substitution loses something. If you cannot name what you are losing, you have not understood the current system well enough to replace it. Agar loses elasticity. Starch loses flavour clarity. Gellan loses spreadability in a jam. Write the loss down and get it agreed before spending money, because someone will notice it eventually and it is better that it happens now.
Question three: what happens at your pH?
pH eliminates more candidates than any other single measurement, and it is the one most often assumed rather than checked. Below about pH 3.5 several common hydrocolloids quietly stop performing. Acidified systems additionally have the protein problem described above.
Question four: what cations are present?
Network-forming materials need divalent cations. Sometimes the formula supplies them — fruit usually does — and sometimes it does not, particularly in systems built on demineralised water or purified concentrates. Analyse rather than assume, and re-analyse when suppliers or seasons change. This single variable explains a very large share of batch-to-batch variation complaints.
Question five: what does your process do to the material?
Ask specifically about shear, heat and dwell time — in that order, because each one can undo a network after it has formed.
| Process step | What it can do | Who gets caught out |
|---|---|---|
| High shear | Breaks the network before it rebuilds | Beverage lines |
| Retorting | Breaks acid-sensitive systems | Sauces |
| Long hot hold | Degrades some thickeners | Fillings held before filling |
| Freeze–thaw | Contracts networks and expels water | Frozen and chilled desserts |
A material validated before the pump and validated after it are two different materials.
Question six: is it really one material?
Often it is not. Combined systems are the norm rather than the exception in commercial products: pectin plus gellan for acidified milk drinks, native starch plus gum for clean-label sauces, protein plus gum for aerated products. Accepting this early saves months of iterating on single materials that cannot cover two jobs.
| Product type | Realistic answer | Single-material answer works? |
|---|---|---|
| Acidified milk drink | Protective colloid plus suspending agent | No |
| Plant-based meat | Methyl cellulose plus binder support | No |
| Reduced-sugar jam | Sometimes single | Sometimes |
| Clean-label sauce | Native starch plus gum | Rarely |
Running the elimination on paper
The point of the six questions is that they are answered at a desk. Work through them and most candidate lists halve before anything is weighed. What survives is usually two or three materials, and those are worth trialling properly.
Tools
- Dosage calculator — dose, hydration temperature and ion requirement for 32 applications.
- Grade selector — which grade matches the job you identified in question one.
- Formula pre-check — flags the pH and ion combinations that fail before you run the trial.
Frequently asked
Why start with the dominant job?
Because it eliminates fastest. If the dominant job is "keep acidified protein in solution", half the candidate list is gone immediately and no amount of dose work brings it back.
How many materials should be trialled?
Two or three, and no more. A list longer than that usually means questions one to five were skipped.
Should I use one material or a combination?
Assume combination until proven otherwise. Most commercial systems are combinations, because most products need suspension and something else at the same time.
Where do I start if nothing is working?
Go back to question three and actually measure pH and calcium rather than reading them off a specification. In practice that resolves more stalled projects than any change of material does.