The most instructive thing about the Tropicana disaster of 2009 is that nobody changed the name. The word Tropicana appeared on the new carton exactly as it had on the old one. The typeface changed. The image changed. The orientation of the logo shifted from diagonal to horizontal. But the name — Tropicana, five syllables, rolling and tropical and memorably strange for a company that processes oranges in New Jersey — survived the redesign entirely intact.
And sales still dropped 20% in the first two months.
In January 2009, PepsiCo launched a redesign of its Tropicana Pure Premium line, commissioned from the brand agency Arnell Group for a reported $35 million. The iconic orange-with-a-straw — the image that had anchored the carton since the 1980s — was replaced with a glass of orange juice, cool and clean and photographically competent. The carton caps were redesigned to resemble orange slices. The logo was reoriented. The effect was, by any conventional measure of packaging design, professional and contemporary. By February 2009, Tropicana had pulled the new design entirely and reverted to the original. The estimated revenue lost to the sales decline: approximately $20 million. Total cost including the redesign and reversal: north of $50 million.
The explanation that circulated in marketing departments afterwards was that consumers hated the new design aesthetically. This is incorrect, or at least insufficient. Consumer research on the new design showed mixed aesthetic preferences — some people liked it, some didn't. The problem was not that the glass-of-juice image was ugly. The problem was that the orange-with-a-straw was a brand element, not a design element. The distinction matters.
David Placek's framework for this kind of failure involves what he calls the "visual morpheme" — a non-verbal element that functions exactly as a phoneme functions in spoken language. A morpheme is the smallest unit of meaning in a word. The orange pierced by a straw was a morpheme in Tropicana's visual language: a compact, instantly recognisable unit of meaning that carried the brand's entire positioning in a single image. Not "juice," not "orange-flavoured beverage," but specifically "Tropicana" — the thing in the tall carton with the straw in the orange that you pick up without reading the label. Remove the morpheme and you have not redesigned the packaging. You have erased the brand's most basic unit of recognition.
The phonetics of the word Tropicana are, as Placek would note, excellent. T is a voiceless alveolar plosive — crisp, energetic, attention-grabbing. The TROPIC- cluster is highly processed in the English-speaking brain as tropical, warm, exotic. The -ana suffix adds a rhythmic, musical quality: Tropicana, Americana, Pollyanna. There is nothing wrong with this name phonetically, which is why changing it was never on the table. PepsiCo knew the word was valuable. What they didn't realise, until the sales data arrived, was that the orange-with-a-straw was equally valuable — and equally part of the name, in any functional sense of the word.
This is Rory Sutherland's central insight about the Tropicana case: consumers do not experience the name as a word in isolation. They experience the name as a cluster of sensory signals — the word, the image, the colour, the weight of the carton, the cap that opens like a fruit segment. When consumers in a supermarket aisle reach for a product they buy habitually, they are not reading. They are pattern-matching. The brain reaches a purchase decision in approximately 0.3 seconds for habitual purchases, according to research published in the Journal of Consumer Psychology. In 0.3 seconds, the conscious mind has not had time to process text. It is processing gestalt: the whole visual impression of the product location in the category.
The old Tropicana carton had built a 0.3-second recognition signal over decades. The new carton required the consumer to actually look at the shelf — to read, to evaluate, to confirm. In a supermarket aisle with forty-seven orange juice options, the request to deliberate is the request to reconsider. And reconsidering, when you are holding a $5 carton of juice, frequently means putting it back.
Scott Galloway would note the mathematics here without mercy. Arnell Group charged PepsiCo $35 million to destroy a recognition signal that had been built over thirty years. The recognition signal was not sitting in the word "Tropicana." It was sitting in the orange. The agency was paid to solve a problem the product did not have — a packaging design that was dated but functional — and created a problem the product had never faced: consumer confusion about which carton was the one they wanted.
The broader lesson from Tropicana's eight-week catastrophe is that a brand name includes everything a consumer uses to find the product. For some brands, that is primarily the word. For packaged consumer goods with decades of shelf presence and habitual purchase behaviour, the "name" is a multimedia experience: word, image, colour system, packaging architecture, and the muscle memory of reaching for a particular carton in a particular supermarket bay. When Tropicana removed the orange-with-a-straw, they removed the brand's fastest identifier — the element the shopper's visual system processed before the text was readable.
The reversal was immediate and necessary. PepsiCo's speed in pulling the redesign was, in retrospect, the most rational thing about the whole episode. They had data — dramatically negative data — and they acted on it without sunk cost paralysis. This is rarer than it should be. Most companies that commission £3.5 million rebrands take eighteen months to admit they made a mistake. PepsiCo took eight weeks.
The question for anyone naming or renaming a product is not just "what should the word be?" but "what are all the elements the customer uses to recognise us?" For some products, the answer is mostly the word. For others — particularly those with long shelf presences and habitual purchase patterns — the visual language is as phonetically loaded as the name itself. Understanding which elements are morphemes and which are decoration requires the kind of honest consumer observation that most brand audits are not designed to deliver. The tools to analyse a domain name's phonetics are at domainsleft.com — but the discipline is the same: identify what actually does recognition work, and protect it.