The Cola That Won the Taste Tests and Lost the Public

On 23 April 1985 the Coca Cola Company retired the formula it had sold for almost a century and replaced it with a sweeter drink, later called New Coke. The decision followed a long contest with Pepsi, whose televised taste challenges had suggested that many drinkers preferred a sweeter cola in small sips. The company reportedly ran close to two hundred thousand taste tests, and most of them favoured the new formula over both the old one and Pepsi.

Within days the company faced a flood of calls and letters of protest. Some drinkers hoarded cans of the old drink, and critics treated the change as an insult to a national symbol. The tests had measured which sample won a single sip. They had not measured what drinkers felt about losing a drink they had grown up with. Seventy nine days after the launch, on 11 July 1985, the company brought back the original recipe as Coca Cola Classic.

Analysts have long disagreed about what the episode proves. One view holds that the company blundered by trusting narrow data over the attachment of its customers. Another view, which some executives later suggested, holds that the uproar renewed public affection for the brand. The evidence cannot easily separate the two, since the same events support both readings.

The tests had measured which sample won a single sip.

For managers the case carries a narrower lesson. A test is accurate about the question it asks and silent about the questions it does not. A sip test answers which drink pleases the tongue first, while a purchase depends also on habit, identity and trust. Research that is rigorous within its design can still mislead when the decision it supports is wider than the design.