Discussion

OFD platforms are a new addition to the foodservice sector and provide consumers with access to many foodservice partners. There are three main OFD platforms in Australia, each with a large number of foodservice partners. OFD platform revenue represents a small percentage of the overall foodservice sector but recently observed revenue growth is expected to increase. We found that these platforms are most commonly used by young working adults with higher disposable incomes and that there is evidence of a shift away from traditionally popular takeaway options such as fast food to a greater variety of food types. Convenience is the main driver of OFD platform use. An increase in the desire for convenience food is reported to be a result of rising household incomes, urbanisation and a reduction in the time available for activities such as cooking. More recently, a move towards urban living to access better education and employment opportunities has resulted in greater constraints on time and available cooking space^{(Reference Sievert, Lawrence and Naika57)}. More generally, reasons given for consuming meals prepared away from home include the desire for fast and filling meals, the cost and effort of cooking at home and a lack of time^{(Reference Blow, Patel and Gregg58–Reference van der Horst, Brunner and Siegrist60)}. Whether these are drivers of OFD usage specifically remains unclear due to lack of available evidence.

OFD platforms are disruptors to the foodservice sector; they are altering the way that many of us choose and order takeaway meals and they present unique public health nutrition policy challenges. The availability of takeaway options has shifted from the physical to the virtual with hundreds of restaurants available at the tap of a screen⁽⁶¹⁾. In addition, OFD platforms have introduced an additional and new route to the market as they do not fit into the categories of foodservice outlet, food retail outlet or manufacturer. We believe the relative newness of OFD platforms at the time of this review was a factor in its absence from public health nutrition policies. As a new entrant with a novel business model, this is likely to add an additional challenge and complexity to national policymakers.

Examination of Global authority recommendations^(21–24) and translation into Australian policy initiatives showed that, although the initiatives address some of the best practice policy domains, only menu labelling is mandatory and this is based on state-specific legislation rather than a national scheme. We also found no evidence that public health nutrition policies explicitly target the advertising and promotion used by those in the food service sector. The HSR and Healthy Food Partnership address front of pack labelling, mass media campaigns, reformulation and availability/portion size but are voluntary and/or collaborative initiatives with the food industry rather than government-led regulation. Furthermore, a significant body of research suggests that initiatives predicated on voluntary action have limited success^{(Reference Moodie, Stuckler and Monteiro62)}. The unique OFD platform model and the relative absence of a robust national policy and policy process suggest that including OFD platforms in current initiatives will be challenging. Yet, the absence of policy to address OFD platforms enhances what many already consider to be an uneven playing field between suppliers of food^{(Reference Caraher and Hughes63)}. For example, the lack of information available for smaller independent restaurants limits consumers’ ability to make an informed decision when choosing between several foodservice outlets with and without this information provided⁽⁶⁴⁾.

We have identified three main ways worthy of further consideration. Consumers would be able to make more informed decisions if OFD platforms were required to provide the method and criteria used, if any, to label food options or foodservice partners as ‘healthy’. If not used already, this categorisation should be objective and based on nutritional content rather than a label selected by the foodservice partner. Adoption of nutritional information on the menus of foodservice partners would provide a transparent and objective categorisation method. A policy option that is likely to be highly feasible for OFD platforms is to incorporate nutrition information at least for foodservice partners who already provide this to comply with kilojoule menu labelling. The coupling of this nutritional information with the information already on the OFD platform is likely to be straightforward given that the foodservice partners will already have this information. It could also provide insights into how to best incorporate smaller foodservice partners as part of a phased approach to include all foodservice partners. Further, it would align with potential plans for a national approach to kilojoule labelling⁽⁴⁵⁾. Second, although the HSR is currently targeted at packaged food, there is evidence that it can be applied to fast food^{(Reference Dunford, Wu and Wellard-Cole49)} indicating that it may be feasible to include HSR for food and drink options through OFD platforms. Research demonstrated that determining and displaying portion size was difficult^{(Reference Dunford, Wu and Wellard-Cole49)} and so implementation would be particularly dependent on the cooperation and collaboration of multiple foodservice partners. A third potential policy option is participation in government-led schemes such as the Healthy Food Partnership; this could provide an opportunity for meaningful dialogue between stakeholders (e.g. OFD platforms, the government and the public health sector) to increase the collective understanding of how OFD platforms could make a positive impact on diet and health outcomes and implement actions that align with the goals of the partnership. However, as the foodservice working group has ceased operating, it is difficult to determine what actions will be implemented and therefore how OFD platforms can be included in this initiative.

Public health nutrition policymaking has been shown to be a hotly contested arena with multiple vested interests^{(Reference Cullerton, Donnet and Lee20)}. The recent emergence and growth of OFD platforms add to the current challenges faced by policymakers. OFD platforms operate in a policy and legal ‘grey zone’, and there are similarities between OFD platforms and other regulatory entrepreneurs. For example, Uber resisted being labelled as a taxi service, which enables them to operate outside of the regulation associated with this classification^{(Reference O’Sullivan65)}. Likewise, it is unclear if OFD platforms are categorised as manufacturer, retailer or foodservice outlet, and therefore they fall outside the policies we identified as targeting these types of organisations. OFD platforms also make use of technological advancements that may fall outside current and future regulation. In the US, ground robots have been used to make commercial deliveries on university campuses and trials of autonomous vehicles and drones to deliver orders are ongoing^{(Reference Lambert66–Reference Hwang and Choe68)}. Other technologies such as voice recognition could further increase the convenience of OFD services^{(Reference Lambert66)}. Like Uber, the OFD industry and main operators within it have also become so large that regulation restricting use in any way, particularly increased costs or reduced convenience, would likely be opposed by OFD platform organisations and the large number of consumers that use and support these services^{(Reference Pollman and Barry18)}. In order to design and implement effective policies, policymakers must be able to at least match the pace of development set by OFD platforms.

OFD platforms could be well placed to positively impact the food environment through the application of technology, which underpins their business model, in areas of delivery and marketing. In relation to delivery, OFD could improve accessibility to healthier food. It is well established that areas of greatest disadvantage have a higher proportion of quick service restaurants and unhealthy takeaway options^{(Reference Thornton, Lamb and Ball69)}; OFD platforms enable consumers to purchase from foodservice outlets outside of their immediate geographical location, potentially increasing the healthy options available to consumers in more disadvantaged areas. OFD platforms could also positively impact on population health through marketing and promotion. This is particularly important as young adults are key users of OFD platforms and growth in obesity is predicted to be higher in this group than in older ages^{(Reference Hayes, Lung and Bauman70)}. This could be implemented using choice architecture; a method of changing the environment to encourage or ‘nudge’ consumers towards healthier choices^{(Reference Saulais, Sloan, Legrand and Hindley71)}, during the order process. For example, when presented with available menu options, choice architecture techniques including setting healthy items as defaults, restructuring the menu to highlight healthier options using methods such as promotional tagging, or recommending a healthier alternative to a previously ordered meal can all be implemented to encourage healthy options. These changes could be implemented with relative ease and, due to the high number of foodservice partners, would almost certainly have a larger impact than changes to individual foodservice partners. Adopting these strategies to increase healthy eating would require action from the OFD platforms themselves. It is argued in political theories of corporate social responsibility that companies should use their power and influence as a business in a socially responsible way^{(Reference Garriga and Melé72)}; in the context of OFD platforms and population health, this could include investigating and implementing methods to promote healthier food choices. However, although they may be motivated by evidence of consumers who are increasingly wanting healthier food choices^{(Reference Gustafson73)} and one platform has pledged to add nutrition information to all menus, there is little evidence that those in the food industry have taken meaningful action to promote a healthier food environment^{(Reference Brownell74)}. It is likely that policy inertia, opposition to policies by powerful commercial interests and lack of demand for policy action by the public, are powerful disincentives to meaningful change^{(Reference Swinburn, Kraak and Allender75)}.

In this review, we found no information about the marketing strategies used by individual OFD platform providers. However, we did find widespread use of promotional tagging to identify ‘healthy’ options. The technological capability of these platforms enables the collection of consumer purchase behaviour which could be used to inform targeted advertising campaigns through email and app notifications. While not unique to OFD platforms, these digital promotion techniques go beyond standard print and media to use a mix of social media channels, influencers and food bloggers^{(Reference Dunford, Trevena and Goodsell76)}. Comprehensive research on, and identification of, digital marketing strategies already used would facilitate a comprehensive discussion on what policy options might tackle this aspect of the food environment.

There were some limitations of our review. Due to the paucity of published research at the time of this study, we included data from countries outside Australia (UK and USA) and thus some findings may not be as applicable to the Australian context. For example, while some OFD platforms aim to provide menu labelling for hundreds of restaurants in the UK (before extending this to all countries), there may be additional challenges when applying this to Australia due to the greater proportion of independent restaurants^{(Reference Lee, Hallak and Sardeshmukh77,Reference Nikolaou, Hankey and Lean78)} . In addition, we found very few peer-reviewed papers related to OFD platforms. The majority of the information used to inform this narrative review was sourced from the grey literature, news articles, websites and market research reports. Although through our search strategy, we took care to only include sources from reputable organisations, the lack of peer review evidence could have resulted in some bias. We have highlighted throughout the review when information has been sourced from the OFD platforms themselves, as this could reflect marketing strategies rather than evidence-based data and should be viewed with caution. However, several academic databases were searched, and thus this highlighted the newness and lack of academic research on OFD platforms and the urgent need for this review and continuing research in this area.