Eliminating the Gap Between Planned vs. Actual Route Performance

By Jemel Derbali, Wise Systems Co-Founder

Route planning is an exercise in fortune telling. Whether companies use route optimization software or a map and paper to build their routes for the next delivery day, chances are that the biggest efficiency challenges they face arise from the disconnect between how routes look ahead of time and how they work in reality.

For companies that use vehicles to make multiple delivery stops each day, at the most fundamental level, route planning involves using distance to figure out the shortest way to get to each stop. It gets more complicated when accounting for time windows, as each stop has a different time at which the driver needs to arrive. For companies that have more than a couple of stops per day per vehicle, there are virtually infinite solutions to this optimization problem. But most route planning doesn’t go beyond  these two variables, and even companies that are highly sophisticated in calculating their route plans learn quickly that this is not sufficient to ensure operational efficiency.

The gap in the planning data combined with day-of, last-mile challenges combine to create a host of issues that can undermine even the most carefully designed plan. As soon as a delivery driver begins their day, the challenges begin. Traffic, whether predictable or not, can make a short distance into a long journey. Parking constraints and unprepared customers at certain times of day can make it impossible to stay on time. Delays cascade throughout the day. And suddenly a well-planned route looks increasingly suboptimal.

Drivers face a difficult choice in these situations. The planned route, if followed, should represent less mileage, but the effect of being off pace can be significant. If a driver misses a time window, a delivery may be refused and revenue will be lost or costs will double as the stop has to be added to the following day's route. Savvy drivers will change sequences to make sure they don't miss high priority customers. But this often increases distance (and maybe time) as drivers aren't able to recalculate the most efficient routes given the changes, making the routes more expensive when factoring in fuel, overtime, and other costs associated with longer routes.

The upshot is that the gap between companies' planned and actual performance is terribly expensive, driving up many of the logistics costs that eat into margins, and compromising customer service. In order to improve actual performance and realign to plans, there are three primary steps that high-performing organizations  must take.

First, they need to understand what happens on the ground at a highly granular level of detail. Too often companies don't actually have data that provides visibility into what really happens as drivers execute their routes every single day. They simply send out routes each day, and then deal with aggregated customer and driver complaints and high bills at the end of the month without understanding the connections between them. Companies need solutions that collect data on their daily delivery operations. Those solutions also need to provide the ability to analyze that data for insights into root causes, and provide recommendations for ongoing improvements and adjustments.

Second, companies need to adopt real-time dynamic routing solutions to help them intelligently adjust routes when they hit the inevitable daily obstacles described above. Drivers are already overburdened, and only an intelligent and dynamic routing solution can make the most efficient decision in the moment, taking into account all of the known parameters associated with each stop

Finally, companies need to leverage the data they are now gathering and put it into action. This ground-level data often reveals the predictability of many of these obstacles, e.g. parking is always bad in this particular customer location at noon, and this customer is always understaffed at 9AM. Sophisticated companies often spend significant time and resources trying to use the data they gather to adjust plans, but often this feedback loop is slow and can't keep operations ahead of the challenges they face.

Luckily, machine learning technologies make it much faster and more efficient to automatically predict and adjust around future delays based on operational data. With modern solutions in place, high-performing teams have the ability to eliminate gaps between planned and actual routes, and reap the bottom-line dividends. By reducing the impact of unplanned delays, the dividends accrue rapidly, minimizing fleet miles, overtime, and customer refusals, while improving customer service.