The FTT defines ‘normally’ as ‘over 50% of the time’ – yet decides the case by analysing the product’s characteristics.
The latest instalment of the Mega Marshmallows litigation, Innovative Bites [2026] UKFTT 500 (TC), has been handed down by the FTT.
The case concerned large (5cm) cylindrical marshmallows marketed as ‘perfect for roasting, smores, or just snacking’. The FTT and UT had previously held them not to be confectionery under its ordinary meaning. The Court of Appeal, however, held that Note (5) [to VATA 1994 Sch 8 Group 1] was conclusive – effectively a deeming provision – remitting the matter to determine whether the products fell within Note (5) as ‘normally eaten with the fingers’.
Central to the remitted appeal was the meaning of ‘normally’. With minimal relevant case law proffered by the parties, both agreed on, and the tribunal adopted, a quantitative construction: holding that ‘normally’ means ‘more often than not’ (for example, ‘over 50% of the time’ at [29]).
This reminded me of the case concerning parents getting fined for taking children out of school: Isle of White Council v Platt [2017] UKSC 28, where the words ‘fails to attend regularly’ fell to be construed. The word ‘regularly’ could mean: evenly spaced intervals; sufficiently often or frequently; or in accordance with the rules. In that case, the Supreme Court held that it meant in accordance with the rules, rather than ‘with sufficient frequency’.
In the context of Note (5), ‘normally’ could be an evidential question of consumption habit frequency, but it could also be construed as something like ‘by design’ or as a standard – closer to ‘typically’ or even ‘suitably’, as a qualitative/normative hurdle, than ‘usually’ as a quantitative test. That could more closely resemble the typical consumer approach, rather than focusing on evidence of a percentage/calculation. It has also been suggested that ‘normally’ could mean not extraordinarily, i.e. it would not be absurd to do so.
In Innovative Bites, there were multiple possible consumption methods: (A) on a skewer; (B) eaten with fingers following roasting on a skewer; (C) part of a smore; and, (D) directly by hand. B and D were both with the fingers; A and C were not. HMRC sought to argue that smores were eaten with fingers, but the tribunal dismissed this argument as either (1) the marshmallows are between the biscuits which are held with the fingers or (2) marshmallow is an ingredient in the final smore (for example, a burger patty in a final burger).
Interestingly, despite proposing a ‘more often than not’ evidential test which read to me at least as a test of frequency, the tribunal undertook a detailed analysis of products’ characteristics. Roasted marshmallows were likely to be too hot and structurally unstable to be handled directly. Furthermore, the size of the marshmallows, the large packaging, and retail positioning, all strongly indicated that the product was not aimed at on-the-go consumption with the fingers. This is useful as obtaining numerical evidence would be difficult, and open to methodological criticisms.
The conclusion was expressed pithily as (A+C) > (B+D) such that the marshmallows were not ‘normally eaten with the fingers’ and therefore did not fall within Note (5). The appeal was allowed.
Home >Articles > Mega Marshmallows and the meaning of ‘normally’
Mega Marshmallows and the meaning of ‘normally’
The FTT defines ‘normally’ as ‘over 50% of the time’ – yet decides the case by analysing the product’s characteristics.
The latest instalment of the Mega Marshmallows litigation, Innovative Bites [2026] UKFTT 500 (TC), has been handed down by the FTT.
The case concerned large (5cm) cylindrical marshmallows marketed as ‘perfect for roasting, smores, or just snacking’. The FTT and UT had previously held them not to be confectionery under its ordinary meaning. The Court of Appeal, however, held that Note (5) [to VATA 1994 Sch 8 Group 1] was conclusive – effectively a deeming provision – remitting the matter to determine whether the products fell within Note (5) as ‘normally eaten with the fingers’.
Central to the remitted appeal was the meaning of ‘normally’. With minimal relevant case law proffered by the parties, both agreed on, and the tribunal adopted, a quantitative construction: holding that ‘normally’ means ‘more often than not’ (for example, ‘over 50% of the time’ at [29]).
This reminded me of the case concerning parents getting fined for taking children out of school: Isle of White Council v Platt [2017] UKSC 28, where the words ‘fails to attend regularly’ fell to be construed. The word ‘regularly’ could mean: evenly spaced intervals; sufficiently often or frequently; or in accordance with the rules. In that case, the Supreme Court held that it meant in accordance with the rules, rather than ‘with sufficient frequency’.
In the context of Note (5), ‘normally’ could be an evidential question of consumption habit frequency, but it could also be construed as something like ‘by design’ or as a standard – closer to ‘typically’ or even ‘suitably’, as a qualitative/normative hurdle, than ‘usually’ as a quantitative test. That could more closely resemble the typical consumer approach, rather than focusing on evidence of a percentage/calculation. It has also been suggested that ‘normally’ could mean not extraordinarily, i.e. it would not be absurd to do so.
In Innovative Bites, there were multiple possible consumption methods: (A) on a skewer; (B) eaten with fingers following roasting on a skewer; (C) part of a smore; and, (D) directly by hand. B and D were both with the fingers; A and C were not. HMRC sought to argue that smores were eaten with fingers, but the tribunal dismissed this argument as either (1) the marshmallows are between the biscuits which are held with the fingers or (2) marshmallow is an ingredient in the final smore (for example, a burger patty in a final burger).
Interestingly, despite proposing a ‘more often than not’ evidential test which read to me at least as a test of frequency, the tribunal undertook a detailed analysis of products’ characteristics. Roasted marshmallows were likely to be too hot and structurally unstable to be handled directly. Furthermore, the size of the marshmallows, the large packaging, and retail positioning, all strongly indicated that the product was not aimed at on-the-go consumption with the fingers. This is useful as obtaining numerical evidence would be difficult, and open to methodological criticisms.
The conclusion was expressed pithily as (A+C) > (B+D) such that the marshmallows were not ‘normally eaten with the fingers’ and therefore did not fall within Note (5). The appeal was allowed.
.tmb-editorspic.jpeg?sfvrsn=7981419_1)
