Importing is complex, products are made to order, and it takes on average six months to complete a transaction. That’s why the biggest opportunity for improvement in the import supply chain is the first mile. The first mile is defined as order placed to product received and includes purchasing and logistics.
I see potential there in three areas: process automation, digitization, and connectivity. Excel and email plateaued decades ago, but they are still the default in this industry. Where I see the greatest innovations is companies using applications (apps) which allow entirely new ways of doing work and connecting their supply chain.
A good example is Uber. They didn't try to create a better taxi, but they instead embedded their app with new technologies such as GPS, messaging, and navigation to create a completely different experience and outcome for passengers. Importers are still — figuratively speaking — standing on Olive Street hoping that a taxi comes by.
When we think about automation, we consider two opportunities. The first is where we can automate a physical activity, such as unloading or driving a truck. A lot of work is being done in robotics and full self-driving (FSD) to try to tackle that, and the pandemic was likely a major accelerant. The second is data entry. The supply chain runs almost exclusively on Excel, email, and documents. Much of this can be eliminated through process automation. The best solution is to make the work go away — for example, using an app versus manual creation. The next best solution is to enhance manual efforts with technology such as chatbots, robotic process automation (RPA), and optimal character recognition (OCR).
In terms of visibility, there needs to be progress in the following three areas:
1. End-to-end: There is great work being done in pockets (such as project44 for transportation); however, that's only one portion of the supply chain.
2. Exceptions: In most cases, it's more important to know where something isn't versus where it is.
3. Quality: Much of the data available for visibility is very low-fidelity. GPS, ML, tagging, analytics, and process automation must become embedded in all visibility solutions.