The GM of Bel’s US cheese business on doubling Babybel output, the impact of GLP-1 drugs, and how AI is reshaping the way people shop.
The US is Groupe Bel’s biggest market, a third of the French dairy giant’s sales, and is expected to deliver more than half its growth over the next five years.
The maker of Babybel and The Laughing Cow announced a $200m investment in March to double Babybel output in South Dakota. Negar Sedghi, GM of Bel’s US cheese business, spoke to Just Food about capacity, GLP-1 and «marketing to algorithms».
Capacity and growth
The US is Bel’s biggest retail cheese market, expected to deliver over 50% of group growth in five years. Bel has two US cheese plants (Brookings, South Dakota, and Wisconsin) plus co-manufacturing and some imports, and is doubling Babybel capacity.
Volumes have grown 6–8% a year over three years, with net sales faster. Growth is driven by a portfolio of unique brands (Babybel’s playful «reveal»), innovation and communication, plus expansion into under-represented channels: out-of-home is ~50% of US consumption but only 3–5% of Bel’s sales, so the firm aims to more than double its presence there (e.g. a two-count Babybel flow-pack offering 8–10g protein).
GLP-1 and healthier snacking
Around one in eight Americans are on GLP-1, and new US dietary guidelines specifically call out dairy — especially full-fat — for greater consumption, with ~80% of Americans not eating enough dairy, fruit and vegetables.
GLP-1 curbs appetite, making nutrient density and protein essential to protect muscle, and shrinks portion sizes — a fit for Bel’s portioned products. It changes behaviour for whole households. Snacking overall is declining under GLP-1, guidelines and awareness, pressuring salty and sweet snacks and driving M&A.
Bel says it has «been on this trend for 100 years», with no sodas or chips in its portfolio.
«Marketing to algorithms»
Sedghi believes AI will transform shopping: «It’s all about marketing to algorithms as well as marketing to humans».
Shoppers will ask an AI agent (she is «a Claude user») to build lists from a need state rather than a category, so brands must appear in the «consideration set» even when «cheese» isn’t mentioned.
«Garbage in, garbage out» — brands need accurate information out in the world (agents draw heavily on sources like Reddit reviews), making marketing more about data management than «beautiful pictures».
Retailers are building their own agents (Walmart’s Sparky), so the product detail page (PDP) and the data fed to retailers’ databases become critical. Bel considers itself at the forefront, with strong AI partnerships.
Source: Just Food
