How Can Fast, AI-Powered What-If Scenarios Impact Last Mile Logistics?

When logistics professionals think about what-if scenarios, they may be more likely to think of them in connection with processes that model the entire supply chain, not just the last mile. Even if you’re picturing them as a last-mile process, you may be imagining them as part of a slow, cumbersome planning process built on manual processes and calculations. 

If this were a few years ago, thinking about what-if scenarios in those terms would be spot on. But technology has progressed a lot since then—and what-if scenarios can now be applied to last mile logistics processes quickly and efficiently to help planners model different ways of servicing customers across the last mile. 

Of course, not all what-if plans are created equal. Performing these calculations quickly and accurately requires powerful computational abilities and strong predictive capabilities—usually requiring technologies like AI and machine learning that can produce more nuanced predictions than even the most experienced human planners. Luckily, modern cloud-based technology makes it possible to do exactly that. 

This is a moment in which technology is fundamentally reshaping the way that we get goods from warehouses and distribution centers to their final destinations. As what-if planning technology is increasingly leveraged to tackle the problems of the final mile, what impacts could that have on distributors, wholesalers, and other delivery organizations?

How Do What-If Scenarios Work?

Before we get into the impact that faster what-if scenarios can have for distribution operations, let’s talk a little bit about how they actually work. Traditionally, to run a what-if scenario, planners would have to model a particular delivery territory, including locations of distribution centers and customer delivery sites and customer and driver constraints around timing, capacity, day of week preferences, etc. From there, you’d have to tabulate likely travel times manually based on point-to-point calculations for drive times between stops. 

It should be easy to imagine how time consuming this could be. Even with legacy software solutions that do some of the work for you, it can be a slow, tedious, and deeply inefficient process. Most businesses simply can’t afford to spend such a huge chunk of their time and resources on running these scenarios more than once or twice a year—which means that, most of the time, businesses wind up running territory plans that don’t directly address the current market conditions. 

In the world of AI and cloud computing, however, these kinds of scenarios don’t have to take months. With the best last mile delivery software, you can test out dozens of different scenarios per day and choose the right one for your particular business case at the moment. This does more than just save time—it puts you in a position to adapt much more quickly to changes in your customer roster, commodity prices, transport network, driver roster, etc. In other words, you have the tools you need to stop running outdated plans. 

Use Cases for What-If Scenarios

Like we said above, increases in speed and power for this kind of technology means that it has new use cases that would have been difficult to imagine even a few years ago. There are a handful of things that might fall under that umbrella for delivery planners at large enterprises:

• Sales territory planning: Building sales territories that are at once efficient in the way they service customers and balanced in terms of the potential revenue for each seller can be a huge challenge. But when you’re able to run what-if scenarios as needed to model potential changes to your territories, you can iterate on them as needed—making them more efficient over time and ensuring that you stay responsive to changes. 
• Strategic planning: Weekly and monthly distribution plans for servicing customers have many of the same pitfalls that we saw above with sales territory plans—if you can’t model adjustments in how you’re delivering to your customers, it becomes incredibly difficult to keep up with changing conditions. Luckily, what-if scenarios let you rapidly model the effects of delivering to particular customers more or less frequently, of changing the location of a distribution center, of changing which driver handles which customers, etc. In this way, you can find the most efficient path forward before your plans go out of date. 
• Holiday planning: If you’re a beer distributor, for instance, the entire month leading up to the Super Bowl might require you to deliver totally different volumes than usual. To prepare for this kind of seasonal shift in how you manage your deliveries, you can leverage what-if scenarios based on projected peak season order and customer data. Using these projects, you can create strategic plans in advance and activate them at the right time. 

Again, these sorts of use cases are most powerful when plans can be run quickly. A long lead time between testing out possible scenarios has the potential to slow down your planning enough to virtually ensure inefficiencies. By contrast, testing out possible distribution scenarios in one sitting to determine the best fit for your needs right now can put you in a position to boost efficiency from end to end.

What to Look for in Your What-If Scenario Technology

When you have the right technology for running what-if scenarios and generating optimal plans based on AI-powered predictions and historical data, you can become much more adaptable. But how do you know if you’re leveraging technology that’s up to the task? Here are a few things to look for:

• Speed: We’ve beaten this dead horse over the course of this article, but we can’t overstate the importance of speed. The ability to run scenarios in minutes significantly shortens planning lead times and enables you to adapt much more quickly to changing market conditions. 
• Ease-of-use: Speed is mostly a matter of processing power on the back end, but the total time it takes to complete the process is also deeply dependent on how long it takes planners to configure these scenarios in the first place. When planners have to spend months getting trained on a system and then hours and hours grappling with the technology to produce the desired results, you’re almost back where you started. By the same token, if you have to bring in outside specialists to work on your scenarios because the planning solution is too specialized, you can wind up with silos and bottlenecks that hamper planning workflows. That’s why it’s so critical to find delivery software that enables you to operate what-if modules with minimal training and minimal fuss.  
• Direct integration with routing: One of the pitfalls that can come with running what-if scenarios on sales territories, for instance, is that territories that look efficient on the planning screen but completely fall apart when they’re translated into actual route plans. Simply put, the only way to make sure your plans actually work is to have your what-if planning fully integrated with your routing. If you’re actually generating hypothetical route plans for every scenario you run, you can be sure that the plans you ultimately adopt will work in practice, not just in theory. 

The ability to quickly and easily generate scenarios, run them, and adopt the best plans for the current market conditions isn’t just a pie in the sky prediction for future technology. It’s something that can be achieved right now, and it has the power to have a huge impact on distributors and other businesses that grapple with complex last mile delivery problems. The trick is finding a  logistics application that actually empowers you to work quickly, efficiently, and intelligently.