Back in the day, the Mr Delivery booklet was a fridge fixture for most South African households. After a technology-led transformation, Mr Delivery became Mr D Food – a 100% online business with 95% of all orders being placed on its mobile app.
Today, the app has had over two million downloads, boasts more than 700,000 monthly active users and Mr D Food delivers to more doors across the country than any other delivery service. Mr D Food is one of South Africa’s resounding technology success stories. It’s a true South African business, built by South Africans.
It’s also a leader in e-commerce and a hub for ambitious engineers seeking to work for the best of the best, said Mr D Food’s Jonathan Muir, VP of engineering.
Muir and Monde Masiko, head of product & user experience, outline tech highlights that took the maverick start-up from analogue to digital.
“In 2014, takealot.com bought Mr Delivery for its ability to be last mile delivery hubs for e-commerce parcels… with a food business on the side. Globally, online food delivery was gaining traction, so the decision was made to develop the food side into a true online business. We were given three months to bring a minimal viable product to market.
“We used learnings from the old system to inform our design. There were lots of moving parts and it wasn’t just about building an app for customers to order on,” said Masiko.
“To get to a maximum order-to-door time of under 40 minutes, we needed to synchronise our three-legged marketplace – restaurants, drivers and customers – and develop apps and technology for all three sides to create a seamless experience.”
“We knew we needed to ensure we had one centralised technology platform. We ran the paper-based business in parallel to implement key learnings from the old system. We also launched in one branch in order to learn – and execute – on what our customers really wanted,” said Muir.
5 decisions that set the business on its path to success
- Mr D Food’s CEO, Devin Sinclair, focused on an app-only strategy from the start. In 2015, no one else was doing this. It was a pivotal factor that led to our success. “We leap-frogged past a traditional website to a totally mobile-centric strategy for customers, restaurants and drivers – betting that this would be the natural and best experience for our users in the future. We had to convince customers to migrate to and trust the app,” it said.
- The group adopted the strategy of trying to be as agile and fast as possible. “Time to market was key. Our competitors were close on our heels, so we needed to accelerate our roll out. We used Amazon Web Services (AWS); ‘going cloud’ enabled us to have a solid infrastructure in place from day one. It was another pioneering decision in the context of the time,” it said.
- It emulated how Takealot deployed its infrastructure using microservices to create a scalable business for the long term. “We also made the critical decision to use Python – despite the fact that everyone was using Java. The productivity of Python 3 exceeds anything else and fits our iterative philosophy,” it said.
- The business went with NoSQL rather than relational databases to further facilitate rapid acceleration without constant changes to the database structure.
- The group built its business on iteration and data. “Fail forwards fast – try it and see how customers respond. All our decisions are data-driven; everything is measurable. If you can’t measure it, you can’t manage it,” it said.
Along the way, seemingly small things posed big problems, said Masiko, highlighting the menu in particular as being his paradox.
“Think of a menu. There are so many different layers, with optional extras and price modifiers. We needed to fit all this functionality into a working screen.” The team spent lots of time on the ground with the restaurants and in the field doing deliveries.
“E-commerce usually has a three or four day delivery window. Mr D Food’s delivery time is 34 minutes. That’s a radical difference.”
Muir said that using stateless architecture in Mr D Food’s microservices is a game changer. “This enables us to containerise those services using DockerTM and then on top of that we have elastic container services, which allow us to scale elastically.
“We can scale the size of the platform depending on demand. For example, we can scale up the restaurant list view – our most popular page – drastically on a Friday night.”
Muir said every pay-day Friday is bigger than the last, with more than 10,000 requests per minute at peak times. The business has grown by more than 10 times in the last 2.5 years. The main strategy has always been ‘fewer taps to order’ to make the whole customer journey as quick as possible.
Then the customer side needs to speak to the restaurant and driver arms of the business. Getting the logistics right is an ongoing game of chess, Muir said.
“The radius of restaurants around the customer needs to be 5km away at most. Remember, that doesn’t mean 5km in a straight line. We’ve had to optimise real world road driving distances to make the network more efficient and empower drivers to make more deliveries.
“We recently introduced machine learning to better predict food preparation times based on granular historic data – for example, one serving of sushi may be faster to make than ten pizzas. This gives accurate delivery estimations to drivers and customers, enabling drivers to cut down on waiting times and customers to manage expectations.”
Machine learning and AI is a big focus for the team. “There’s so much we can do in the recommendation and personalisation space. We have deep insights into our customers’ behaviour – the next step is to use these to prompt people with smart suggestions according to food preferences during the day and in the evening,” said Muir.
This is a natural progression for the business, which has grown over 100% in the past year and expects to continue on its growth trajectory.
Masiko said he still pinches himself sometimes. “We launched in Somerset West and had about twenty orders a day and we wondered if our bold bets were going to pay off. But we kept launching in other areas, kept learning and making changes, and now we’ve had over 2 million app downloads and deliver hundreds of thousands of orders a month in all nine provinces in SA.”
Muir said he wants to double the Mr D Food engineering team in the next year.
“It’s a very agile environment. Headed up by a young, 35 year old CEO, we are a dynamic, entrepreneurial team that moves fast and loves change. We all believe in ownership – owning what we’re good at and putting our names on something we believe in. It’s about experimenting. There’s so much complexity that needs solving – we can guarantee our people will never be bored.”