As I write these lines, I’m at home waiting for a delivery from Ikea, which said the package would arrive sometime between 9 am and 5 pm today. It’s close to 2 pm, and I’m still without news from the carrier, FedEx. The probability that my Swedish design furniture will arrive today is decreasing as the hours go by, inversely correlated with my anger levels. I stayed home, instead of heading to the office, for nothing. (The package arrived the next day, without notice from the company or the carrier.)
The frustration caused by failed or delayed home deliveries is shared by many. Roughly 5 percent of online deliveries don’t reach their destination on the first attempt, according to a survey of 304 retailers and 2,020 consumers in the US, Germany, and the UK by research consultancy Loudhouse. Each late or failed delivery costs retailers an average of $17.78, the survey found, and additional delivery attempts increase traffic congestion—though by how much is difficult to measure.
As our dependence on home delivery rises, retailers and carriers are working to reduce inefficiencies in the last mile—the crucial step when packages are handed to the recipient. They’re focusing on three main obstacles to successful, on-time delivery: making sure parcels arrive where and when they’re supposed to, ensuring there’s someone to receive them when they arrive, and providing delivery locations where exact timing is less critical.
Stay in the know with our Transportation newsletter. Sign up here!
Many carriers try to reduce the risk of failed deliveries by giving recipients the flexibility to postpone deliveries, indicate when to leave packages with a neighbor, or ask that parcels go to a nearby access point. DHL says it has invested heavily in online interfaces including mobile apps, chatrooms to reach customer service representatives, and social media bots that, for example, let you track a package via What’s App.
But many of the challenges encountered at the last mile are the result of missteps that take place long before parcels end up in a FedEx of DHL truck. The key to better last-mile delivery, logistics experts say, is giving retailers and carriers more and better info. “You can never provide a carrier with too much data about how to reach you,” says Andrew Williams, CEO of DHL Express Canada.
One big cause of delivery failure is inaccurate or incomplete address information, according to research by the University of Washington’s Urban Freight Lab. Perhaps the sender forgot to provide an apartment number, or the carrier needs to punch in a code to enter the building.
The problem can also occur at checkout if online shoppers incorrectly enter their addresses, or the location isn’t recognized by the retailer’s website. Loqate, a location verification service, says that big data can help. It and similar services check a mailing address against a location database in real time to make sure the address is valid, verifying whether the street is spelled correctly, for instance, or whether the building number actually exists (Try checking your address here.)
Address validation companies can also enrich a shipping address with additional information like neighborhood density and usage—residential or commercial—and traffic patterns. This helps the carrier plan its delivery route more efficiently, increasing the likelihood that a package will be delivered on time.
Even more enhanced data about delivery locations is on its way, thanks to mapping services that give more details about streets, traffic, and even building interiors. What’s the best time of day to avoid congestion on that block? Is there parking available nearby? What about an elevator to reach the sixth floor? Where is the mailroom?
“That rich array of enhanced location data tells you so much more about the property and any restrictions about delivering to it, whether it's on the street side or the property itself,” says Matthew Furneaux, who heads retail operations at Loqate’s parent company, GBG.
That information could become particularly useful for deliveries performed by autonomous vehicles. With data maps of curbs and indoor layouts, and emerging indoor navigation technologies that use WiFi and Bluetooth as an alternative to GPS for wayfinding, a robot could deliver your lunch at your desk after successfully passing through your building’s security and making its way through your office. While the technologies are still being developed, there is enough interest in enhanced navigation tools by mapping giants Google, Here, and others to suggest indoor deliveries will eventually benefit.
Better location data also helps retailers deliver products to places around the world where formal addressing systems aren’t so reliable, or where people don’t have an address at all. E-commerce in emerging economies could be worth $4 trillion by 2022, according to Boston Consulting Group. Yet billions of these new customers may be unable to receive a package at home.
“We're having to go out and find, state by state, city by city, who are the local providers that we can partner with to ensure that our customers can deliver to their customers in different parts of India, for instance,” Furneaux says. One of those partners is What3words, a company that provides geocoded three-word addresses for every three-square-meter of the earth’s surface. Do you want, for instance, to send flowers to John Lennon's memorial in Central Park but don't think your carrier will recognize Strawberry Fields? Go to What3words and punch in ///battle.boxer.stray.
Even when all the address details are accurate and fully detailed, deliveries can face delays, often lengthy ones, inconveniencing and annoying customers. Logistics researchers are turning to machine learning for help ensuring closer-to-on-time deliveries. Algorithms can collect data from past trips and take into consideration factors like traffic or building accessibility to help carriers determine the best delivery route. Unfortunately for residential consumers, such an approach is most helpful for recipients who receive a large volume of packages and therefore generate a large amount of data points—most likely other businesses.
“You're trying to predict the availability of a customer during a certain time period on a particular day. And the number of observation points that you have on that [residential] customer is not that big at all, says Matthias Winkenbach, who heads the Megacity Logistics Lab at MIT.
Some solutions seek to dodge the problem of home deliveries altogether. One solution is package pickup sites, which can eliminate many of the common causes of missed deliveries, such as the requirement for a signature or overly large packages that don’t fit in the mailbox. France’s national postal service, La Poste, has hundreds of parcel lockers set up in train stations and other public places. The University of Washington’s Urban Freight Lab is testing common lockers in commercial buildings and near transit stops in Seattle. (Unlike Amazon’s branded lockers, these can be used for all packages.) An initial pilot at the Seattle Municipal Tower found that the system reduced the number of failed deliveries to zero and reduced the overall delivery time by 78 percent when compared to door-to-door delivery.
Other solutions take aim at places with greater barriers to last-mile delivery, like neighborhoods that haven’t been mapped or countries without a reliable postal addressing system. In India, MIT’s Megacity Logistics Lab is assessing the feasibility of tapping local knowledge, such as enlisting neighborhood stores to get packages to their destination. “They know their neighborhood, they know who's living where,” Winkenbach says. “That's sometimes turned out to be a much more efficient way of overcoming those last 50 feet.”
Those solutions tend to be site-specific. Something that works in one region or one town may not work in another, and carriers may want to work with a variety of local partners, he says. It’s a strategy that Amazon has embraced by handing last-mile delivery over to small businesses, including in remote parts of the world.
While technology and data have a role to play in improving deliveries, the human side of logistics is essential and often undervalued. Many retailers would be surprised to learn that their goods are already being sold overseas through informal delivery networks based on word of mouth and trust. And that’s where the future of home deliveries probably lies—somewhere between data-powered delivery routes and the street smarts of a seasoned driver.
“The technology doesn't know that this driver has been on the route for five years and knows that Mrs. Smith wants the delivery done at the side door every time,” says DHL’s Andrew Williams. “There's technology that augments the local knowledge the driver has, and that's when it really is a perfect match.”