Lotus Bakeries Bets €500m on Biscoff Plants While Nestlé Shuts a Chocolate Factory

Two plans landed weeks apart. On 21 July, Nestlé said it will close its chocolate plant in Diósgyőr, Hungary, by the end of the year. On 7 August, Lotus Bakeries said it will spend at least €500m on three Biscoff plants by 2030. One firm is shutting a plant. The other is building three.

The Lotus number is large for a firm of its size. €500m is worth more than a third of the group's entire 2025 sales. It is the biggest spend in the firm's history. All of it sits behind one brand.

Price rises do not fill a new plant

Lotus reported first-half 2026 sales of €749.1m, up 14%. Net profit rose 23.5% to €98.1m. Trading profit rose 19.4% to €131m. Biscoff sales and Biscoff volume both grew by more than a fifth.

That second number is what pays for the plants. Most food and drink majors grew this year by charging more while moving fewer packs. Lindt grew on price with volume falling. Mondelez took European pricing negative in the second quarter. Unilever's foods volume fell. You can raise a price in a week. Filling a new line takes years of real demand.

Biscoff is now among the five biggest biscuit brands in the world by shop sales. More than 13 billion were eaten across 70 markets last year. It is over half of everything Lotus sells. In the US it is the fastest-growing brand in both biscuits and spreads. More than one in ten American homes now buys it.

Lotus owns the biscuit and rents the rest

The €500m buys plants for two products only: the biscuit and the spread. Everything else that carries the Biscoff name is made by someone else. Mondelez makes the co-branded bars under Cadbury, Milka and Côte d'Or. This year it added Toblerone, Suchard, Freia and Marabou. Froneri makes the ice cream, now in 13 European markets, with Australia and New Zealand due late this year.

Mondelez also makes and sells Biscoff inside India. That cut out the import cost that had kept the brand a premium niche there. Chief Executive Jan Boone wants to copy that model in Brazil. He calls it "too important a country to remain niche for Biscoff". A family-run Belgian firm is using the world's largest biscuit maker as its plant network.

What the Hungarian plant could not do

The Diósgyőr site made hollow chocolate figures for Smarties, KitKat and Milkybar, and shipped them to more than 20 markets. It was still too small to keep. The site was worth under 3% of Nestlé's Hungarian sales and 0.4% of its output there. It sold hard for a few weeks around Easter and Christmas, then sat quiet.

Trade press has called the closure the first of many. The read is simple. Cocoa, sugar and energy all cost more now. A line that runs a few weeks a year cannot carry its own fixed cost. A line that runs every week can.

The bill and the wait

Lotus has told the market it will spend more than €250m in 2026 and 2027 alone. The new hall in Lembeke, Belgium is being built. Lotus calls that site the largest single-product biscuit plant in Europe. In Thailand, the first new line at Chonburi is due before the end of 2027. The US hall in Mebane, North Carolina only comes online in mid-2028, with full output in 2029.

So Lotus is betting demand holds for three more years while the concrete sets. Two risks sit against that. Own-label biscuit spreads are already on shelf in most European grocers. And Biscoff is sugar, refined flour and palm oil. That puts it in the path of every rule now being written on processed food.

What this sets up

For anyone weighing where to put plant money, the test on show here is worth copying. Lotus sorts its lines by how many weeks a year they run. It builds only the ones it can keep busy. Build what runs all year. Rent the rest.

For buyers and backers, the licensing model is the sharper lesson. Lotus keeps the brand and the two products it makes best. It hands chocolate, ice cream and India to partners whose plants are already standing. Reach arrived without the cost of owning every line that carries the name.

Watch three things from here. Whether Brazil gets the India treatment. Whether own-label spreads slow the volume line before 2029. And whether any other mid-sized brand owner copies the split: own the core, rent the rest.