The situation
A sauce manufacturer needed a substantially reduced oil mayonnaise for a retailer range. Previous attempts had matched the viscosity of the full fat product and been rejected on eating quality: the panel described them as thin in the mouth despite measuring correctly on the instrument.
Why it is hard
Oil in a mayonnaise is not only fat content. The emulsion is a dense packing of oil droplets, and that packing is what gives both the viscosity and the mouthfeel. Take oil out and the droplets have room to move, so the structure is lost even when a thickener restores the number.
Viscosity and mouthfeel are separate properties, and a hydrocolloid that matches the first does not restore the second.
A product can be correct on every measurement and wrong in the mouth, which is the hardest kind of failure to argue about in a meeting.
Mustard is useful here because it is already in the recipe for flavour, and mustard seed material carries mucilage that is genuinely surface active rather than merely thickening. Using more of what is already declared avoids adding a line to the label.
What was done
The emulsion was rebuilt rather than thickened. Droplet size was treated as a formulation target in its own right, since smaller droplets at lower oil give back part of the structure the packing had provided. The emulsifying contribution of the mustard fraction was used deliberately rather than incidentally. Acid stability was checked separately, because the system sits at low pH for the whole of its life.
What it shows
That oil reduction is a structural problem and not a thickening problem, and that the measurement which looks like success is often the one hiding the failure.
