Barilla builds a mill in Foggia and sells two: making in-house pays only above a threshold
27 September 2026, 10:13
On 18 September Barilla presented its new plan for southern Italy. In Foggia, where it already has a pasta factory, it will invest €100 million in a new-generation mill that will work for that factory: wheat in at one end, pasta out at the other, all on the same site. It is the model the group already uses in Parma, and construction is scheduled between 2028 and 2030. Giovanni Palopoli, head of operations for pasta and sauces, explains that the new site “meets the need to improve our logistical and production efficiency”.
Under the same plan Barilla is negotiating the sale of the Altamura and Castelplanio mills to Molino Casillo, and of the Marcianise plant to Pastificio Liguori. With a clause that says a great deal: those plants will keep producing for Barilla, under multi-year contracts, and the condition set by the group is that all the workers keep their jobs.
In Foggia the mill is being built. In Altamura the mill is being sold, and its semolina is still being bought. They look like opposite choices. They are the same choice, made in two different places.
Producing in-house turns a cost that follows sales into a cost that stays
A business that buys from a supplier what it needs for production pays by quantity: so many kilos, so many euros. If it sells more it buys more; if it sells less it buys less. It is a variable cost: it rises and falls with sales, and in lean months it lightens by itself.
A business that decides to make it in-house changes the nature of that cost. The raw material still has to be bought, and it stays variable. But along come the plant to be depreciated, the people who run it, maintenance, energy. These are fixed or semi-fixed costs: they are the same whether the departments are full or half empty.
In return, every item sold costs less and leaves more margin. It is a trade-off: less cost on every item, more cost every month, whether anything sells or not.
Below a certain level of sales, buying costs less
An example. A bakery products company has a turnover of €15 million and buys in a semi-finished product worth 40 per cent of revenue. If it made the product itself, raw material and processing combined would cost it 30 per cent: ten more points of margin. But it would need a plant and a team, for €1.2 million a year in fixed costs.
Is it worth it? It depends on how much the company sells, and the exact point can be calculated. If every €100 sold saves €10, covering €1.2 million of fixed costs takes sales of €12 million. That is the threshold: the additional fixed costs divided by the percentage saving, that is €1.2 million divided by 0.10.
Above €12 million, producing in-house pays. At €15 million the ten points are worth €1.5 million: once the plant is paid for, there is an extra €300,000 a year.
Below €12 million, producing in-house costs money. If one year sales fall to €10 million, the ten points are worth €1 million and the plant still costs €1.2 million: the company earns €200,000 less than when it was buying in, after spending money on the investment. The old supplier, meanwhile, cost less and asked for nothing in the idle months.
That is why the turnover to set against the threshold is not the figure from the best year. It is the figure that can be counted on in the worst years.
Barilla builds where volumes are certain, and buys where they are not
Measured this way, Barilla's plan adds up. A mill attached to a large pasta factory already knows how much wheat it will grind: what that factory's pasta needs, every day. Volumes are certain, fixed costs are spread over known quantities, and the threshold is cleared without effort. And the wheat travels less: that is the logistical efficiency Palopoli is talking about.
Where that certainty is missing, the same logic points the other way: the mill is run by someone who is a miller by trade, and who can also fill it with work from other customers. Barilla buys the semolina, and pays only for what it needs.
For a smaller company the reasoning is identical, with fewer zeros. Bringing a process in-house is not a sign of strength or independence. It is a bet on volumes: it pays off if sales stay above the threshold and is lost if they fall below it. And it is worth knowing where the threshold lies before signing the contract for the plant, not after the first bad year.
This article was prompted by a news item published in askanews.