Melt and stretch are two different jobs

Every developer working on an analogue pizza cheese has had the same week. The cheese does not flow enough, so the formulation is opened up. It flows. And it has stopped stretching. So the formulation is tightened again, and the stretch returns along with the original complaint.

This is not a failure of tuning. The two properties need separate targets and can trade off, but they are not inherently opposed.

Melt is the structure letting go

When cheese is heated, two things happen at once.

The fat softens and then liquefies. Below its melting range it is a solid filler sitting inside a protein network, contributing rigidity. Above it, it is a liquid, and the network is now holding a fluid rather than resting on a solid.

At the same time the protein matrix itself loosens. The bonds holding it together are largely hydrophobic associations and calcium mediated bridges between casein molecules, and both weaken with temperature. The network stops behaving like a solid and starts behaving like a very thick liquid.

Melt, in other words, is the structure agreeing to let go. The more readily it lets go, the better the cheese flows and covers.

Stretch is the structure holding together

Stretch is the opposite request.

To draw a thread, the protein matrix has to stay continuous while it is being pulled, and it has to be able to align. In mozzarella that ability is built mechanically rather than chemically. The curd is heated and worked in hot water, which pulls the para casein into aligned fibres with channels of fat and serum lying between them. When that structure is heated it can be extended, because it is already organised in the direction it will be pulled.

So stretch needs a matrix that is continuous, hydrated, and still strong enough to carry a load at eating temperature. That is the same matrix that melt was asking to dissolve.

Why one moves the other

Once both are described properly the conflict is obvious. Changes to the matrix can affect flow and stretch differently; formulation and processing determine whether they trade off or improve together.

In a real mozzarella this is resolved by one protein doing both jobs at slightly different points on the temperature curve, and by the fibre structure that the process built in.

An analogue is built from selected ingredients, often using heat and mixing, rather than following the same curd-stretching process as mozzarella.

The usual assembly is a protein for structure, a fat for melt and eating quality, a starch for body and cost, and salts to make the protein hold the fat. Each can affect melt and stretch, with the direction depending on the formulation and process. Starch is the clearest case. It is excellent at controlling flow and it is the cheapest way to build body, but it forms a gel, not a fibre. A cheese carrying a lot of it can be made to flow exactly as required and will still not stretch, because there is very little in it capable of being drawn.

And the oven asks for two more things at the same time

Melt and stretch are the two that get named. A pizza oven is asking for four.

Browning. The colour on a baked pizza is a reaction between reducing sugars and amino groups, and it needs both present at the surface. A cheese built from materials that carry little reducing sugar will bake pale no matter how well it melts, and a customer reads pale as raw. The correction is easy to overdo, because browning happens at the surface and quickly while melting happens through the mass and slowly. Push the browning and the cheese can colour before it has flowed.

Oiling off. Free oil pools on the surface when the fat is no longer held. Holding it is the job of the protein, helped by emulsifying and buffering salts, which sequester calcium, open the casein up and improve how much fat it can carry. Too little and the pizza swims. Too much and the cheese is over emulsified: it holds the fat perfectly and has lost most of its flow.

So four properties, one formulation, and every lever moves at least two of them.

The oven is not one oven

There is a further complication that formulators meet late.

A conveyor oven at a chain, a deck oven in a pizzeria and a domestic oven finishing a frozen pizza are three different time and temperature histories. A cheese balanced for ninety seconds of high heat will behave differently over twelve minutes of moderate heat, because melt, browning and oiling off do not scale with time in the same way. Browning accumulates, flow reaches an end point, and free oil keeps being released as long as heat continues.

This is why a cheese that performed in one customer's plant can fail in the next one with nothing changed in the recipe, and why the oven is part of the specification rather than a detail about it.

What happens before the oven also decides this

Most analogue cheese is not sold as a block. It is shredded, treated so the shreds do not fuse, sometimes frozen, and stored for months.

Each of those steps acts on the same structure. A cheese firm enough to shred cleanly is a cheese with a stronger matrix. Anti caking treatment sits on the surface, which is exactly where melt and browning begin. Freezing moves water within the matrix, and what freezes and thaws does not always return to where it was.

A formulation judged only on a hot plate can therefore be correct and still fail, because the product that reaches the oven is not the product that was assessed.

What this makes the job

If all of that is true, then a specification written as a list of ingredients cannot work.

Say which of the four matters most, and in what order. A cheese for a frozen retail pizza and a cheese for a fast service chain are not the same product, and they are rarely ranked the same way. Ranking them is the first decision, not the last.

Specify on the finished pizza. Flow measured on a plate is a useful control test and a poor target. What matters is coverage, colour, thread and surface on the product a consumer receives, made on the equipment that will make it.

Give the oven and the format at the start. The bake time, the temperature and whether the cheese is applied fresh, refrigerated or frozen change the answer more than most ingredient choices do.

Expect the balance to be built, not chosen. No single material delivers melt, stretch, browning and oiling off together. They are held against each other, which is why this is a system rather than an ingredient, and why the last ten percent of the work is usually worth more than the first ninety.

If your cheese melts but will not stretch, or stretches but will not brown, tell us what you are making and what oven it goes into.