How AI Has Helped Food Delivery Cos Reach Greater Heights

Food delivery companies all over the world are using AI to reach greater heights. The same is true about apps such as Zomato, Dunzo, Borzo etc in India. If you are in India and remember the Swiggy app of yore, you may also remember how it was just another option when other apps didn’t work or delivery partners were not available. Today, Swiggy boasts of an order volume that has grown over 200%. Applying AI to its workings, the company generates terabytes of data week after week. The delights of using the app today are such that it has converted dedicated users of rival apps to being their own loyal customers. How 

did they do it? In their own words, this was achieved by real-time, micro-optimization of dynamic demand-supply, over and over many times during the day. With this they were able to provide the urban consumers with hitherto unexperienced levels of convenience.  

The How

Swiggy achieved its goals by creating a three-way hyper-local marketplace wherein the company matched the demands of the consumers with supply from vendors i.e. restaurants, cafes, and stores as well as deliver executives (mostly people looking for quick money and side jobs). They used AI across this marketplace to deliver a delightfully seamless customer experience to achieve unparalleled growth and also drive operational efficiency. Infact, they are now so dependent on AI for their growth that they consider it impossible to go back to the time wherein they relied just on human intelligence to achieve their goals.

Behaviour of the consumer

Collection of consumer data helps Swiggy get hold of the behavioral aspect of things. By knowing their customer’s behaviour through this data they can deliver personalized experiences. This is achieved using Catalog Intelligence with which ML models help enrich the Swiggy catalog with meta-data. For example Classifying foods on offer as vegetarian, egg, or non-vegetarian and even categorizing similar products under sections for example: salads, soups, main course, rice, breads, and dessert. Use of Customer Intelligence helps the company in customer segmentation on the basis of their affordability (derived using their past buying behaviour) and also log customer churn i.e. when a customer stopped using the app or service. Customized and relevant content is shown to the customers  using catalog intelligence and customer intelligence. You may notice the app showing you your previous order and prompting you to re-order the same. This happens as the app may have noticed you ordering the same food from a particular place again and again. This even includes showing you restaurants nearby you depending on your current location and not your usual home or office location. Their Live Order Tracking feature is one of the most popular features among customers. They even get to know of delays and when the delivery partner is at their doorstep.

Vendor X Food Delivery

To tap in the right vendors Swiggy uses AI for time-series based demand prediction models which help the restaurants plan ahead to meet the demand of the customers. The Company also uses ML to cut financial losses by identifying and preventing abuse.

With such advanced uses that are being improvised upon as you read this blog, the food delivery landscape is bound to change even more in the times to come.

Online Grocery Retailers Ride The AI Wave for Customer Acquisition

As Machine Learning has become more accessible, more retailers are leaning towards adopting it for customer acquisition. The same is also true about grocery retailers who are trying to strengthen their relationship with customers using AI and ML. AI uses personalization and other tools to provide better experiences to customers.

Hyper personalization: Long and never-ending product lists are fast getting replaced by personalized offers to entice customers into buying more. Online retailers are bidding to meet customer expectations in unique ways by making the customer experience better. Not only products but product recommendations need to be in tandem with the requirements of the customers, even before they know they want a particular product. AI steps in here by offering experiences customized to individual consumers rather than a discount or offer available to all. This is done on the basis of what products an individual is most likely to buy. It is a win-win as the customer gets gratification, the retailer strengthens the bond with customers, and also records an increase in sales.

ML for personalization: For online retail AI relies on ML algorithms trained with behavioral data to understand customer requirements in a better way. E-commerce players can use this data for personalized product recommendations to customers and present customers with a user-oriented shopping experience. This goes over and above just selling a product.

Diderot effect for AI in online retail: Have you noticed how online e-commerce stores gently nudge you towards a container to go with the new pasta packet you just bought or cooking oil spray to go with the potatoes you just purchased? This is what is referred to as the Diderot effect which is defined as follows – obtaining a new possession often creates a spiral of consumption which leads you to acquire more new things. It is basically an impulse buy on the part of the customer. Though this kind of selling is well-known to marketers in the physical retail space, in e-commerce, such behaviour can be brought about by personalization which is done by analyzing clicks and purchase history or searches of the users. These are then used to make not only relevant but near-apt products which are almost certainly bought by them.

Besides other tools for personalization towards the goal of customer acquisition include email marketing customized for each customer, welcome texts which are personalized, e-shop navigation according to customer visit history, chatbots etc.