What made this possible wasn’t consumer demand alone.Tom Szaky, Founder and CEO, TerraCycle and Loop, September 2025
While the CFO priced the reusable SKU at a premium and the CMO read the tracker promising shoppers would pay it, neither owned the shelf where the decision happens, so what Canada delivered was less a consumer verdict than the reading of a shelf that had barely existed.
P×R=V: the R of an offer is whether the responsible option was ever choosable, and an option absent from the shelf has no P to multiply.
While Loop ran the same experiment in seven countries with the same brands and the same proposition, six markets shut down and one reached five hundred stores, and the brands read the six closures as proof that shoppers say one thing and buy another, although what they had tested was a shelf of five products.
Loop launched between 2019 and 2021 with Unilever, Nestlé, Coca-Cola and Häagen-Dazs, across Canada, the United States, the United Kingdom, Japan, Australia and France. By 2026 only France remains: 500 stores, more than 400 products, a modest profit. (CBC, May 2026)
Szaky named the variable in May 2026: at five products per store both sales and returns stayed low, while at fifty sales jumped and returns climbed from 15% to 80%, in the same country, chain and shopper, so the moving part was the range rather than the customer.
Canada was not short of intent. 83% of Canadians wanted less packaging on groceries and 65% said they made an effort to choose reusable or recyclable brands. Loop Canada offered them 98 SKUs online, many out of stock, at a five-dollar deposit. (Kraft Heinz / Fuse, 2021)
CBC asked Procter & Gamble, PepsiCo, Nestlé, Loblaw and Tim Hortons to comment in May 2026. None would speak on the record. Szaky reports they all said the same thing: the demand did not translate into sales, so scaling could not be justified.
They ran the test, took the answer at face value and filed it under consumer preference, while France was running a different test.
Loop · TerraCycle
Seven markets, one survivor · Same brands, same proposition · The variable sat with the offer.
Return rate once Loop carried roughly fifty products per store in France. At five products it was 15%. Same country, same chain, same shopper, and sales moved at the same threshold. The variable sat on the retailer’s side of the table rather than the customer’s. (Tom Szaky to CBC, May 2026)
Canada · 2021
Loop opened in Ontario on 1 February 2021, exclusive to Loblaw, e-commerce only. 98 SKUs listed, many out of stock. Five-dollar deposit per container. Nestlé had spent about a million dollars developing a single Häagen-Dazs tub. Loop was losing roughly a million dollars a month globally, Szaky telling reporters he expected two more years of that before break-even, and the pilot closed. 83% of Canadians had told a Kraft Heinz survey they wanted less packaging on their groceries.
France · 2019-2026
Carrefour ran an e-commerce pilot in 2019, then became the first retailer anywhere to take Loop in-store. Monoprix and Coopérative U followed. September 2025: 370 products across 345 supermarkets. May 2026, the state to date: more than 400 products across 500 stores, at a modest profit. Buy anywhere, return anywhere, no cleaning required. Carrefour put more than 50 of its own private-label products on the platform, the gesture of a retailer done with piloting.
Six markets produced one finding and drew one conclusion from it, and only the finding holds, because a five-product shelf failing to move says little, while the leap to the claim that shoppers say one thing and buy another found no support in the data. France ran the experiment the others stopped short of, with fifty products on the shelf, priced below the disposable and returnable anywhere, and the return rate climbed from 15% to 80% with the shopper unchanged. Szaky’s own reading stays the bluntest available: what made this possible wasn’t consumer demand alone.
The Market Signal
France out-legislated the other six rather than out-consuming them. The Anti-Waste and Circular Economy Law requires medium and large retailers to give at least 20% of store space to bulk or reusable packaging by 2030, with subsidies attached, and obliges packaging EPR operators to put 5% of the levy they collect into funding reuse systems. Szaky calls it the carrot and the stick, and says no other market has fitted both. The United Kingdom came closest without it. Tesco had the best assortment, the best purchases and the best returns of any Loop partner, and the national rollout was cancelled a fortnight out when Brexit, Covid and the cost-of-living crisis took the oxygen. The offer was working while the context collapsed, and two weeks of political attention separated a case study from a footnote.
Return rate and unit economics
Sell-through, return rate and cost per turn of the container decide whether reuse is a business or a grant. At fifty products per store the return rate is 80% and the loop closes, while at five it is 15%, and each unreturned container is a unit of capital left in a cupboard. The performance case is real, measurable, and it stops short of being a referendum on the customer, since Loop France runs at a modest profit on the same brands, products and proposition that lost a million dollars a month elsewhere.
The offer architecture
P×R=V: the R of an offer is its choosability, meaning stocked deep enough to be found, priced below its conventional twin, and returnable where the customer happens to be, each of them a company decision taken before any shopper is consulted. Six markets skipped all three and then read the outcome as a preference, even though the intention was measured and present, with 83% of Canadians asking for less packaging and 98 SKUs listed for them, half out of stock. A responsibility impossible to choose fails at the shelf, upstream of any market verdict.
The full case
PDF · 7 slides