Creating one-to-one relationships at scale.
Working with one of the category leaders in the fast food delivery market, we combined the use of machine learning driven insight and journey driven communications, to deepen 1:1 relationships and drive market share.
Reading time: 2 minutes
The problem
The food delivery industry is reaching maturity and our client, one of the category leaders was facing several big challenges from their competitors - who were spending heavily on above-the-line media in an attempt to gain market share.
The data showed that order frequency, whilst strong, had headroom to grow. Our brief was to support the delivery of incremental order volumes, revenue and loyalty.
The hypothesis
Deep learning had flagged that the most active customers had a common trait: whilst ordering from a large number of restaurants, they tended to stick to one or two types of cuisine.
Basically, it showed that we’re creatures of habit when it comes to the food we love, more so than the restaurant that makes it.
The same deep learning techniques flagged that the people who only ordered from a handful of restaurants were significantly less likely to place an order than people who were less risk averse and so weren’t afraid to mix up their restaurants and their cuisines.
Our hypothesis was that by prompting more risk averse people to order their favourite Pizza, Thai, Chinese or Indian from a new restaurant, we could increase their adventurousness and drive deeper loyalty.
Our strategy
Whilst we could group people by cuisine, each person was unique, with a favourite food - so it was critical to reflect this in personalised communications.
We created a series of interchangeable recommendation widgets that were sent to customers at the right time in their journey encouraging them to order their favourite food from a new restaurant.
Individual variants were created based on the recipient’s favourite type of food. Every part of the communication was personalised from their name to the local restaurants serving their favourite food. We then delivered an action-inspiring message that challenged them to try something different.
The results
The initial experiment delivered a significant uplift in engagement with 75% higher click-to-open-rate against the benchmark which proved that this type of personalised content was more engaging.
To see if there was a positive impact on orders we scaled-up the initial experiment.
We ran a series of time-bound tests over a number of European markets and saw an average click-to-open-rate uplift of 31% and an ordering uplift of 5%.
Our hypothesis was proven.
By encouraging customers to order the food they love from more restaurants they interacted more frequently.
The customer experience now satisfied a need, rather than unknowingly fighting against it.
Previously campaigns were designed to introduce people to types of food they hadn’t ordered before - which we now know wasn’t resonating with people.
The campaign was so successful that this approach has now become an automated part of the CRM programme across all markets - always-on and continuously encouraging people to be that little bit more daring at dinner time!
Like our content? We’d love if you shared it!