Restaurant food straight to your sofa

Johannesburg, 17 Mar 2020

Mr D Food CEO Devin Sinclair.

With Massmart and Edcon announcing plans to close stores, and general business confidence and consumer spending indicating a flat-lining economy, South Africa’s retail segment is undoubtedly under pressure. However, one segment going against the grain, and experiencing growth in the retail segment, is food deliveries.

Naspers, which spends a lot of time and effort looking at long-term trends, has, in the past three years, invested $2 billion in its technology-enabled food delivery businesses across the world, and is present in at least 35 markets.

Naspers and Prosus CEO Bob van Dijk said at a press event in October 2019: “If you look at what’s happening in India or China, people are really changing the way they eat, the way they source ingredients. I think from here, there will be a growth path factor of 20 to 50 in food delivery.”

And, of course, the investment company has local experience to draw from thanks to its Takealot subsidiary Mr D Food.

Food delivery isn’t new. In fact, the Mr Delivery service can trace its origins back to 1992. But in 2015, a year after it was acquired by Takealot to assist with the etailer’s last mile delivery, the food delivery part of the business pivoted to an app-led strategy, rebranding as Mr D Food. Today, 95% of its orders are placed through the mobile app.

Since 2015, the local food delivery market has blossomed significantly, and rapidly. Statista estimates the ‘restaurant to consumer delivery segment’ in South Africa will generate $403 million in revenue this year; with a compound annual growth rate of 9.1%, it’s expected to reach $573 million by 2024.

“Going cloud enabled us to have a solid infrastructure in place from day one.”
Devin Sinclair, Mr D Food

Mr D Food’s CEO Devin Sinclair agrees that this is a growth segment. He says the company has experienced year-on-year growth of over 100%, growing by more than 10 times in the past two-and- a-half years. “In the last year, over R1.5 billion was ordered on the Mr D Food app,” he says.

The head office team has grown by more than 200 people in the last two years, and the company is looking to hire an additional 50 employees by June, particularly in data science, data analysis and software engineering.

But with rapid market growth comes increased competition, with the likes of OrderIn and Uber Eats also entering the market. Such competition is forcing innovation, which benefits the customer. Among some recent differentiators introduced – which are often quickly copied by rivals – are: time-sensitive promotions (helping to increase usage during quiet periods); real-time driver tracking; various payment options; flat-rate monthly delivery fees, the curation of premium restaurants, exclusive partnerships with key brands, and even alcohol delivery.

The internet’s last frontier

Food delivery platforms are disrupting the restaurant industry. These platforms enable not just traditional restaurants to sell to customers, but have also given rise to a new concept – dark kitchens. Also called cloud kitchens, these businesses are delivery-only kitchens, often operating multiple food brands from one location, sharing resources and ingredients.

SmartKitchenCo, based in Cape Town, is one such business. It currently operates four kitchens and is about to open a fifth. Each kitchen runs the back-end for five brands, including Jazzy’s Pizza, Good Burger, King Chicken and Jiro Poké Bowls.

Founder and CEO Jasper Meyre says that while there are relatively low barriers to entry – you need a certificate of acceptability and a health licence, which aren’t difficult to get – the margins aren't good at a small scale. “To make these kitchens profitable, you have to have scale and growth. You can't run with one brand either. It’s about economies of scale. All our ingredients need to be used wisely. We'll sell desserts across all brands, but the same dessert. If you're a single kitchen without much sales ability, the delivery platforms will take 30%-plus in commission. The typical margins of a well-run restaurant are 30%.”

The number of dark kitchens in SA is definitely on the rise. Look at the delivery platforms and you'll see there are lots of restaurants you've never heard of; they're all dark kitchens, he says. “Some are larger groups running multiple brands and kitchens, some are single operators, but they won't last.

“It's a bit of hype at the moment and everyone is jumping on it, but a year in and they'll realise. To be worthwhile, you have to invest a lot and make it large-scale.”

Meyre admits to not being a trained chef. He studied business and evolved his career from part-time pizza delivery guy, to owning a mobile pizza company trading at markets and events to where he is today. “I love starting businesses. For me, this was the last frontier of the internet – something had been done with taxis and accommodation. I thought the last big change was food and how the internet empowers the on-demand economy.”

Run effectively, operating multiple brands, the dark kitchen model, like any modern agile business, shares resources and is able to try things, grow them if the data is good, or shutter them if not. “I currently have five brands in the kitchen, but I've probably been through 15, testing the market, getting the feedback, all without costing too much money. If the feedback isn't great, we can pull the brand; if feedback is good, push it further, which helps to reduce a bit of risk. We don't need to redecorate the kitchen every time, or change the front of house.”

Where the dark kitchen has an advantage over the traditional restaurant is that operations are specifically set up for delivery platforms. “Restaurants aren’t designed for delivery, they’re either sit- down, or for you to come and collect fast food.

“We're doing damage to them, big time. The idea of starting a brick-and-mortar restaurant is a scary thought with all the money that goes into it. There will always be a place for high-end sit-down restaurants, because it’s an experience.

“We're delivery-only so we're providing a different service. We're trying to create a good experience through delivery. That comes down to packaging, sealing the bags so the drivers don't take anything. Many restaurants haven't thought about that; once the food leaves the kitchen, they're done. We know the food has to travel well, and the packaging has to be designed to carry the food well.”

SmartKitchenCo predominantly partners with UberEats, which passes on feedback and analytics. It recently scored investment from Tim Draper, a Silicon Valley-based entrepreneur, who was an early stage investor in Skype and Baidu. The plan, says Meyre, is to nail down the processes and technologies, prove the concept in Cape Town and then look at expansion into the US.

Even though the economy is stuttering, this non-essential service is growing. How? Sinclair provides some insight. “Consumers are increasingly time-poor and looking for on-demand service in all aspects of their lives. Many South Africans are too busy to shop, cook or go out for their favourite meals, and are drawn to the ease, variety and convenience of getting food delivered through an app. With an average delivery time of 34 minutes, once they weigh up the time and money it takes to drive to the shop or restaurant, pay for parking, stand in line, drive home, prepare the meal, the attraction and affordability of food delivery becomes clear.”

The drive for jobs

Given the high unemployment rates in the country, the growth of this sector is a boon to the gig economy, giving delivery drivers employment. Sinclair says the number of drivers delivering food for his company has tripled over the past two years to approximately 5 000, servicing 2 600 suburbs, in all provinces across the country. Drivers are either independent contractors or employees of franchisees.

“This provides income-earning opportunities for thousands of people, many of whom were previously unemployed. South Africa stands to benefit from this growth – as the culture of food delivery grows, so too will employment opportunities and income for drivers.”

The synergies created by Mr D Food being a part of the Takelot delivery group means that work isn’t just limited to delivering food at peak times on a weekend; there are also opportunities to perform deliveries for etailers Takealot.com and Superbalist.com, as well as card deliveries for FNB, during the day. “By providing drivers with multiple types of deliveries, we can leverage synergies that allow us to maximise efficiencies and scale of the driver network, as well as maximise driver earnings.”

He adds that while food delivery might have once been the preserve of a Friday evening treat, it’s now being used for everything from weekday breakfasts and office lunches, through to catering for sports matches at home or weekend binge watching sessions. This broadening of consumption times also provides expanded opportunities for restaurants and kitchens, and increased hours for the food staff that work within them.

Where’s the tech?

Mr D Food claims to have more than 700 000 monthly active users on its app. Underpinning such a delivery operation is a technology infrastructure that has to be robust enough to be available to customers, drivers and restaurants, accepting orders and payments for pretty much the entire day, every day. So, what does the Mr D Food architecture look like?

Sinclair says that in 2015, when the business was being restructured, the decision was taken to adopt a mobile-centric, app-only business. To achieve this, the infrastructure had to be cloud-based to enable agility.

“We used AWS. Going cloud enabled us to have a solid infrastructure in place from day one,” he says.

The use of microservices allowed for scalability, and this was considered vital as the business was expected to grow significantly. A stateless architecture was also adopted, meaning that containers could be deployed. On top of that sit elastic container services, he says. “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 on a Friday night.”

‘Pay-day’ Fridays in particular experience high demand, with more than 10 000 requests per minute at peak times.

In addition, NoSQL was used to help speed up growth, without needing to adjust the database structure. “We’ve always had a culture of test, learn and fail fast. We built our business on iteration and data. All our decisions are data-driven; everything is measurable. If you can’t measure it, you can’t manage it.”

Machine learning has also been added to the mix, to assist with predicting food preparation times, based on historic data. Drivers spend less time at the restaurant waiting, and customers have a more realistic expectation of when their delivery will arrive.

Machine learning and AI are a big focus for the team. AI is being used to help guide drivers on the optimal route from the restaurant to the customer, so that food arrives as quickly as possible. Personalised recommendations for customers are another use of AI. “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,” he says.