Technology behind the wheel

Kimberley Taylor, Loop

---

In the 2013 bestseller, The Everything Store: Jeff Bezos and the Age of Amazon, author Brad Stone explains how the “flywheel effect” worked in the early stages of the retailer, which was started in 1994. Lower prices meant more customer visits, Stone says, and more customers meant increased sales volumes, which, in turn, attracted more commission paying third-party sellers to the site. This allowed Amazon to get more out of their fixed costs, like the fulfilment centres and the servers needed to run the website. And by upping efficiency, they could lower their prices even further. “Feed any part of this flywheel, they reasoned, and it should accelerate the loop,” Stone writes.

In 30 years, the company now has a valuation of $1.5 trillion, and had a 37.6% share of all ecommerce sales in the US in 2023. From just selling books, it has branched out into cloud services, digital advertising and streaming. Amazon Web Services, which has about 32% of market share, accounts for 16% of Amazon’s total revenue, but is responsible for 74% of the whole company’s operating income. Other than the flywheel, there are any number of reasons for its success, among them its high-performance culture, as well its “working backwards” model. This means imagining what the consumer wants, and then working backwards until the product team knows exactly what to build. Teams will produce press releases, or FAQs, and if the product described isn’t better (or easier, or quicker to use, or cheaper), then it won’t be built.

“Customers now expect faster deliveries, more transparency, and real-time tracking – and they want all that without paying a premium.”

Kimberley Taylor, Loop

These ways of doing business have also been described as the “Amazon Effect”. In its retail business, consumers enjoy same day delivery, a catalogue of around 12 million products and competitive pricing. It has also re-invented the shipping process, and delivers an average of 1.6 million packages a day in the US alone.

There’s no chance any other competitor will catch up (Walmart is at 6%), but as Cornel Rautenbach, CTO at local ecommerce company Bob Group, says, smaller brands need to offer personalised and value-added services as a differentiator.

Kimberley Taylor, CEO of local delivery management platform Loop, says Checkers’ Sixty60 has had a similar impact on the local market. “Here in South Africa, it feels more like the Checkers Sixty60 effect. Customers now expect faster deliveries, more transparency, and real-time tracking – and they want all that without paying a premium. This puts huge pressure on logistics companies to step up their game.”

Some successful brands wisely don’t try to compete with Amazon on Amazon’s terms, says Marc Kuo, the founder and CEO of Routific. “Giving customers the opportunity to talk to a real person, for example, is huge,” he says, or stocking locally-sourced, high-quality products that a shopper wouldn’t find on Amazon. He says Yuppiechef is a great example of an ecommerce business that has won over customers with personal touches, such as including a handwritten card with every online order. But once the item has been ordered, it then has to be delivered, and it’s here that many companies stumble.

“Giving customers the opportunity to talk to a real person, for example, is huge.”

Marc Kuo, Routific

Successful logistics companies are focusing on marketing, technology and operations, Taylor adds. “Tech is just one part of the equation, which is why they’re paying just as much attention to how they run their operations, how they manage their teams and how they communicate with their customers,” she says. What technology can do, at least in theory, is make it easier to track the movement of your goods in real-time and optimise the flow of goods from point A to point B.

The Traveling Salesman Problem is one of the most studied challenges in computer science and it was while working on this problem at university that Taylor developed the algorithm that would serve as the foundation for Loop.

The problem statement is as follows: given a list of cities, what is the shortest possible route between them that visits each city exactly once before returning to the starting point? What makes this problem so tricky is that we’re not just talking about large cargo trucks transporting goods from one location to another, we’re talking about a delivery driver having to make multiple stops along the way and with each stop comes more complexity. This might explain why this problem remains largely unsolved. Modern technologies like route optimisation software have enabled us to create solutions that are good enough. Streamlining manual route planning, these solutions make it easier to find the shortest, most cost-effective routes when travelling between multiple destinations, while also considering real-world realities and modern business constraints. Routific’s Kuo says that until the 2010s, most companies didn’t have the computing power to solve routing problems efficiently. Now, high-quality solutions are available and affordable for just about everyone, he says.

Loop focuses on optimising last-mile delivery, and its platform helps logistics companies handle deliveries more efficiently and at a lower cost using route planning and optimisation algorithms. Its analytics engine tracks delivery success, order volumes, time spent with customers and overall logistics performance, so customers can identify areas where there is room for improvement. Taylor says Loop customers like Pingo, OneCart, Parcel Ninja and Kauai have seen a 30% average increase in delivery volume, a 20% reduction in distance driven and a 50% reduction in the amount of time spent with customers. “We collect tons of data across the platform, which we then use to give clients valuable insights,” she says. This helps them improve their operations, better understand driver behaviour and assess how successful their deliveries are.

Routific, meanwhile, sells delivery route planning software. Users upload or import an order list with addresses and the software creates the most efficient routes. It takes into account factors like how many drivers are available, vehicle type, load sizes and customer delivery time preferences, among other things.

But it’s not enough just to create a mathematically efficient route – it has to make sense for the drivers. For example, if a driver is leaving a neighbourhood after making a delivery and sees one of their colleagues driving into the same suburb, they may think that it would have been smarter to put all the deliveries for that neighbourhood into the same truck. Kuo calls this problem, where routes are criss-crossing each other around a city, “spaghetti routes”, and says drivers may think that the routing software is inefficient. Delivery drivers following their own routes, going rogue, is said to be a challenge in the industry.

Intelligent route optimisation takes the human into account when planning routes – keeping your drivers happy – by considering certain practicalities in conjunction with a mathematically optimised route. For example, most route optimisation algorithms don’t take time-of-day traffic into account.

While this tech holds great potential for the industry, Taylor believes that one of the biggest issues, especially in a place like South Africa, is resistance to change, which is why change management is so important. Taylor says many logistics companies are stuck in their old ways and are hesitant to go digital because they’re worried about complexity and costs; they’d rather avoid the hassle of training employees on new systems. “Another big challenge is getting old systems to work with new ones,” she says. “A lot of companies are running legacy systems that don’t easily communicate with modern tech, causing operational disruptions.”

With the right technology, adds Taylor, “we’re going to see companies like Checkers shift from being a retailer that just happens to deliver to being a fullblown delivery company that sells just about anything.”