Aims: Online food delivery platforms play an increasing role in our food environments and shaping dietary choices, yet little is known about whether these platforms validly categorise and promote healthy food outlets. This study aimed to quantify the healthiness of food outlets classified as ‘healthy’ on the Uber Eats platform in Victoria, Australia.
Methods: We conducted a web scrape of the Uber Eats Australia platform between November 2022 and January 2023. Data (including outlet name, category, and location) were extracted using an automated Python script. Outlets classified by the platform as ‘healthy’ were reclassified using the DIGIASSESS index (1), a validated food environment scoring tool developed for online food delivery platforms. A combination of automated and manual classification was used. Descriptive statistics were used to determine the proportion of each outlet type and overall healthiness.
Results: From the web scrape, 12,938 unique food outlets were identified, of which 1,408 were categorised as ‘healthy’ by the platform. Classifying the healthiness of these outlets using the DIGIASSESS index, we found that most were classified as “less healthy” (n=1,123, 79.7%) or “unhealthy” (n=180, 12.8%). Only a small proportion of outlets were classified as “healthy” (n=106, 7.5%). The most common outlet types within the ‘healthy’ category were “Independent – Takeaway” (n=166, 11.8%), “Independent – Cereal-Based Café Meals” (n=136, 9.7%), and “Service Station Convenience Stores” (n=104, 7.4%), all classified as “less healthy.”
Conclusions: The Uber Eats’ ‘healthy’ category does not accurately represent nutritional evidence of the healthiness of food outlets, potentially misleading consumers and hindering informed decision-making. Clearer, scientifically-informed, standardised criteria for categorising food outlets on online delivery platforms are needed to improve transparency, prevent misinformation, and support public health goals.