By the end of 2026, revenue in the worldwide online food delivery industry is expected to reach over $1.5tn. According to data published by Statista, this will grow to $2.05tn by 2031. What was once a niche market is now a massive global phenomenon, driven in countries like the UK by Uber Eats, Zapp, Just Eat and Deliveroo. As brands seek to expand their market share in a sector eyeing further expansion over the next five years, companies are leveraging every piece of data they can to gain that competitive edge.
In other words, food delivery firms are vying for our attention. One of the ways they’re doing this is by analyzing customer data to create more personalized and intuitive experiences. Notably, Deliveroo utilizes Amazon Web Services’ machine learning (ML) to better understand real-time ordering patterns. This has led to better engagement through, for example, targeted marketing campaigns.
This has been seen in other web-based sectors. For instance, when you check out current bingo promotions at iGaming platforms and see special offers tailored to new customers or, conversely, long-term users who play at certain times such as after 6 pm. For Deliveroo, the company leveraged Amazon SageMaker to build and deploy ML models that processed historical order frequency to display discounts from a personalized restaurant feed. Data such as price point and cuisine were taken into account so that each user gets a unique, tailored feed.
Rather than sending blanket marketing emails to every single person on a subscriber list, platforms are using business intelligence (BI) to better connect with customers. This could be anticipating when they might order – based on previous deliveries – to issue a special offer app notification, or create milestone rewards to build brand affinity.
DoorDash, for example, developed a unified consumer memory BI engine to map long-term customer behavior. Elements such as dietary habits and average basket sizes were monitored. One of the most interesting aspects DoorDash was able to identify was habit loops such as a user’s inclination to try a new cuisine on a particular day of the week.
This long-term memory was used to power the company’s AI assistant, named Ask DoorDash. By retrieving a user's stored preferences in real time during a chat, the AI agent can build personalized grocery carts or suggest meals. After combining this long-term memory with real-time session intent, the company reported a 24% increase in grocery checkout conversions and a 17% rise in basket sizes.
Similarly, Uber Eats has leveraged graph learning – an advanced form of ML – to map the complex relationship between users, dishes and restaurants. The company analyzed millions of order patterns and has uncovered specific taste preferences, allowing it to suggest cuisines or local venues that a user might like before they search for them. In its app, it has combined these smart recommendations with real-time updates on aspects such as kitchen prep time and delivery routes. The knock-on effect is long-term customer retention and higher order frequency.
In a trillion-dollar market that’s predicted to achieve huge further growth over the next few years, raw data alone is no longer enough. Using insights derived from business intelligence and machine learning, food delivery companies are shifting from generic menus to intuitive, personalized experiences in order to stand out.