Plant-based vs animal-based orders

A simple measure to understand the impact of each nudge is to compare the percentage of orders month on month that were plant-based vs animal-based. The table below compares the percentage split of main meal orders from October – February that were either plant-based or animal-based, categorised by sales channel and intervention. We expand on these results below. 

Dine-In Interventions

Chi square analyses indicated that, compared to control, plant-based selection was significantly higher with the churro promotion for chicken (p=0.01) and paneer (p<0.01), but significantly lower with the environmental message for chicken (p=0.04) and paneer (p=0.01). The differences are represented in Figures 1 and 2.

Takeaway Intervention

Linguistic changes to the descriptions of items on the takeaway menu saw a sizable effect. Compared to takeaway sales in the control month, the proportion of plant-based meals ordered increased by 9 percentage points, from 55% to 64%, which was a significant increase (p=0.01). This change is graphed in Figure 3. 

Burrito of the Day Intervention

Burrito of the Day sales recorded the most significant change in purchase behaviour.  Compared to the control month where just over half of Burrito of the Day orders were plant-based, 100% of Burrito of the Day orders were plant-based during the nudge month. This was, of course, expected, since animal-based burritos were removed from the Burrito of the Day menu during the nudge month. However, there was no significant decrease in the overall number of orders during the plant-based only month (p=0.23) – in fact, there was a modest increase, from 72 orders to 83 orders , as shown in Figure 4.

It is worth noting that the Burrito of the Day menu was subjected to a particularly strong nudge, where animal-based options were simply not presented to consumers. We still consider this to fall under the definition of a ‘nudge’, since it is not changing the availability of meat-based options, nor increasing their price. Rather, it is simply prioritising the plant-based options for promotion. The meat option was still available at its regular menu price.

Taco Tuesday Intervention

In contrast, a negative effect was observed for plant-based options during the Taco Tuesday intervention, which involved highlighting some plant-based options on the menu.

Compared to Taco Tuesday sales in the control month, the proportion of plant-based main meals ordered decreased by 8 percentage points, from 55% to 48%. We suggest potential reasons for this in the discussion.

These findings affirm and expand upon existing literature on choice architecture nudges in several key ways. Firstly, descriptive nudges that serve to appealingly convey the sensory qualities of plant-based dishes stood out here as particularly effective. Simply adding descriptors to plant-based ingredients and dishes such as ‘fresh’, ‘crispy’, or ‘aromatic’ can improve their appeal and increase sales of those dishes. 

Burrito of the Day sales also demonstrated an effective nudge. Whereas in the control month, the menu (which provided a discount on one specific burrito per day) included plant-based and animal-based fillings, in the nudge month, the plant-based fillings were prioritised for promotion. The observed effects can be interpreted as the result of a substantial price discount on plant-based alternatives. When the price difference is significant, consumers are more likely to choose the plant-based option. 

However, no quantitative conclusion can be drawn from this experiment. It is to be expected that 100% of consumers chose the plant-based option in the nudge month (compared to 64% in the control month), since plant-based options were the only one listed on the menu. In this sense, more significant than the proportion of plant-based vs. animal-based orders is the fact that sales in the nudge month did not diminish – in fact, they increased by 15%. This demonstrates that centering promotions around plant-based options does not require sales to be sacrificed: in this instance, sales increased after the plant-based promotion was introduced. 

There is an interesting difference in performance of the promotional and the informational nudges. When plant-based dishes were promoted using in-restaurant promotional cards with the offer of a free churro, the percentage of plant-based orders marginally increased. However, when these in-restaurant cards were replaced with informational messages highlighting the environmental harm of animal-product consumption, the percentage of plant-based orders decreased. This implies that consumers respond more positively to interventions that incentivise plant-based consumption, compared to informational interventions that advocate for plant-based eating. 

That said, it is worth noting that our interventions overlapped with some Indian holiday months, which is likely to have extraneously impacted food choice during this time. Diwali, a festival that sees large numbers of Hindus gravitating toward vegetarianism, ends in November, so by December, a lot of Hindus begin eating more meat again. Additionally, Western habits such as higher animal product consumption are more likely to be adopted around Christmas and New Years’, particularly in major cities such as Mumbai. This could explain the higher consumption of chicken during the December nudge period, which coincided with the environmental message.  

The visual nudges implemented in the Taco Tuesday menu also resulted in a negative effect (i.e. an increase in the percentage of animal-based products). This might be explained by the fact that these particular visual nudges – such as a red box surrounding plant-based options – were not strong enough to push consumers towards plant-based choices. 

Other limiting factors should also be acknowledged. Some of these nudges, such as the free churro nudge nudge, relied to some extent on staff enforcement, which is vulnerable to human error. Local cultural and linguistic barriers may also have affected the efficacy of these nudges. Some may not have internalised the messaging of the environmental cards due to language and/or context barriers (Gracias Granny uses English in its restaurant rather than Marathi or Hindi). Moreover, the value of the ‘free churro’ promotion may have been diminished due to a relative lack of knowledge surrounding Mexican cuisine in Mumbai (if a customer does not know what a churro is or hasn’t tried one before, they may be less inclined to value its inclusion in the promotion). 

These findings are valuable because they provide real-world sales data on the effectiveness of menu nudges in promoting plant-based eating, addressing a gap in the literature. The results are widely applicable to the global hospitality and restaurant industries, offering practical insights into cost-effective strategies for reducing meat consumption without sacrificing sales. Moreover, this study was conducted in India—a country with a significant and growing impact on global food consumption patterns and an important context for sustainable food policy development.

This research aimed to investigate the effectiveness of a variety of plant-forward nudges in a real-word restaurant setting. Growth of meat and dairy consumption in rapidly developing LMIC such as India is likely to be disproportionately high, and so research into consumer food choices and effective plant-based choice architecture modifications is particularly relevant in these settings.

These findings validate the effectiveness of choice architecture modifications in nudging consumers towards more plant-based choices. It is worth noting that these findings are exploratory and influenced by unique factors related to the location, plant-forward nature, and menu of the restaurant in question as well as the timeline of the research itself. 

Future research on this subject could address these limiting factors. For example, they could extend the time period for which each nudge is active to better control for external month-on-month variation in sales. Moreover, it may be that frequent consumers at Gracias Granny are already plant-forward eaters. Implementing these nudges in different restaurant environments with different consumer demographics (for example, those with a more animal-product dominated menu) would therefore also be interesting. Overall, these findings should serve to encourage similar research on plant-forward nudge trials in real-world settings to explore how these interventions manifest in other environments.