The better is the enemy of the good: How Delivery Hero is helping restaurants to constantly improve their service

Delivery Hero has the largest network of restaurants in the online food ordering and delivery space, partnering with more than 150,000 restaurants around the world. Collaborating with so many partners, how does Delivery Hero actually maintain this huge network on a daily basis to keep both restaurants and customers happy?

How to distinguish good from exceptional restaurants

“Every restaurant is a very important partner for us. Hence, we provide them with knowledge and data – and we are striving to enhance our partnerships every day,” says Agron Rexhepi, Global Director of Sales at Delivery Hero. Agron and his team use data-driven insights into restaurant performance to help our partners grow and improve.

Tracking data on restaurant performance and customer satisfaction provides Agron’s team with meaningful information on what works well, and what doesn’t: “We take a very close look at operational key performance indicators, since they determine how customers will experience online food delivery and how they will judge the restaurant.” These indicators include, amongst others: fail rate (e.g. orders that were declined or not delivered by the restaurant), issue rate (e.g. customers receiving the wrong food), and customer feedback submitted via rating (zero to five stars) for food quality and service.

Delivery Hero’s account management program ensures that all qualified restaurants have a dedicated person of contact who will consult them on their performance. Based on data and experience, account managers also advise restaurants on how to gain loyal customers, for example by reworking the menu structure, lowering the minimum order value or improving the quality of a certain dish that has led to negative feedback in the past.

Standardized procedures allow restaurant partners to optimize their operations as well as their presentation on the platform, eventually leading to higher customer satisfaction and loyalty.

Educating, rewarding, removing: How to ensure customers will only choose from the best

“We will do everything to support the restaurants in improving their operations. However, we won’t sell any additional promotion features to restaurants that don’t fulfil our operational criteria, since we don’t want to highlight a restaurant that is not exceptional,” Agron notes.

Restaurants with great operational performance will get a higher ranking, whereas consistently poor performance can result in a restaurant being removed from the platform.

Our local teams decide on these sanctions and implement features that create transparency for the user. For example, users of our Turkish platform Yemeksepeti can filter the restaurants by several criteria such as speed of service and quality of food. Yemeksepeti also implemented a banner that cautions customers when they are about to order from a poorly performing restaurant. As a result, Yemeksepeti is experiencing an outstanding reorder rate and order frequency for restaurants with good performance.

Automation and machine learning will shape the future of restaurant partnerships

Agron knows: “In the end, happy customers are all that matters – both for restaurants and for us. Hence, we will continue to improve our processes, for example by driving automation and implementing self-learning algorithms.” To reach that goal, experts from Delivery Hero such as the Data Science team have implemented machine learning into the process. We aim to ensure real-time data exchange with our restaurant partners. This will enable them to improve and adjust quickly in cases where there are changes or issues with the order.

Is machine learning a way to ultimately make restaurants better and customers happier? Why not – Louis Armstrong already knew:
♪ ♫ They’ll learn much more than I’ll never know. And I think to myself: What a wonderful world ♪ ♫