A Framed Field Experiment was implemented in France in order to compare the relative behavioural responses and then the induced nutritional effectiveness of seven front-of-pack logo formats: The Guideline Daily Amounts (GDA), and six ‘choice logos’ such as Green Keyhole or Traffic Lights. From a consumer point of view, while GDA requires a demanding global-diet heuristic, the choice logos require an easy product comparison heuristic. Our six ‘choice logos’ are different after 3 criteria: aggregate vs. analytical information, shelf vs. all products point of reference, multicolour logos vs. ‘only green’ logos. We measure the effect of each logo on the nutritional quality of actual consumers’ shopping baskets in a controlled experimental shop. We use a standard criterion aggregating the overall density of free sugar, saturated fatty acid and salt. We find that different logo formats generate different nutritional impacts. Some choice logos have better nutritional impacts than GDA. Aggregate and multicolour logos induce best responses. Shelf-referenced logos are not more efficient but they trigger very different behavioural trajectories than logos referenced on all-products.