Despite promising interventions to lower people’s daily sugar consumption, such as health- or taste-focused labels, the consumption of sugar-sweetened beverages (SSBs) continues to rise. To improve the effectiveness of existing labels, the way people process sugar amounts in grams (g) as displayed on beverages seems to merit elucidation. For example, do people perceive the difference in the amount of sugar, and thus in the subjective sweet taste, between two beverages according to Weber’s law? Additionally, is that perceived difference the cause of their beverage choice? In order to investigate these questions, participants in this online experiment first had to estimate the sugar difference between two beverages based on grams and then decide whether they would switch to a lower-sugar beverage. We found that participants’ different estimates followed Weber’s law. The choice of the lower-sugar beverage, however, depended on how large they personally perceived that difference. In other words, the choice was independent of the ratio. These results show that future labels, rather than indicating the total amount of sugar, should indicate whether the reduction, for example in the amount of sugar compared to another beverage, was perceived as significant by others.