DESIGN INTELLIGENCEJune 30, 2026·Mary · DEPIX Design Intelligence

Reese's cheapened the candy, kept the wrapper, and got caught.

The most expensive design decision Hershey made last year wasn't on a screen or a plinth. It was inside an orange wrapper that didn't change at all.

Sometime after a new chief executive arrived in August 2025, Reese's quietly began swapping its substance. Milk chocolate gave way to "compound coating." Peanut butter became "peanut butter creme." The packaging stayed identical, the price stayed familiar, and the brand assumed the object was the label. On 14 February 2026 the founder's grandson, Brad Reese, published an open letter saying he had to spit out a Reese's Unwrapped Mini Heart and threw the bag in the bin. His post drew 97,000 likes and over seven million views, reached Colbert, Good Morning America and the AP, and by 2 April Hershey announced it was reverting most products to the classic recipe. The reversal arrived in under eight weeks. The reputational bill is still open.

For a design audience the story is not a candy story. It is the cleanest demonstration this year of where a product's identity actually lives, and where it does not.

Hershey treated the wrapper as the product. That is the seductive error, because the wrapper is what every internal review sees. It photographs the same. It scans the same at the register. On a cost model, "compound coating" and milk chocolate sit in the same cell of the spreadsheet, separated only by a margin number that points one way. Every signal a company looks at on the way to "ship it" said nothing had changed. The only signal that mattered, the one that fires when a person bites down, was never in the room.

That is the trap of substance-led products. When identity is carried by material and not by silhouette, a cost-down substitution and a faithful product look identical right up until use. They share a shape, a colour, a logo, a shelf. They diverge violently in the one place the brand cannot stage: the customer's mouth, hand, or year-three living room. A clay model never tasted of wax. A render never went grainy. The packshot flatters the substitution exactly as much as it flatters the real thing, which is to say the asset a company stares at hardest is the asset least able to tell it the truth.

And the customer is a faster, harsher panel than any focus group. Hershey did not get a quarter of soft sales to diagnose. It got Brad Reese on day one, because a 97-year-old recipe is a sensory memory millions of people carry pre-loaded. The benchmark wasn't a competitor. It was the buyer's own remembered version, summoned in full the instant the wrapper opened. You cannot win a blind taste test against a memory the eater already trusts more than your new formula.

This is the argument for pressure-testing substance, not just packaging, before it is locked. The work a concept phase exists to do is to put the real material in front of real people in the real moment of use, while the change is still reversible and the cost of "this is worse" is a sample rather than a recall. A parallel design team renders and prototypes the substitution into its lived states, against the remembered original, and asks the only question that decides the launch: not "does it look the same," but "does it survive the bite." Hershey ran that test live, on national television, with the brand's own DNA as the prosecution witness.

The reformulation reportedly touched a sliver of the range, around three percent. It still cost a CEO his first impression and handed every rival a month of free contrast. The saving was real and the icon was real, and the design failure was believing the second could be quietly traded for the first behind unchanged graphics. Substance is not a line item you can edit in the dark. It is the product. The wrapper just told everyone where to look.