What is the best way for companies to improve the overall performance of a complex transportation and logistics network?

In transportation operations, performance improvement often falls to focusing only on the bottom performers. Companies tend to default to this approach because they have literally dozens or even hundreds of facilities, drivers, dock workers and routes to manage.

Analyzing and understanding trends for so many facilities is extremely complex. The challenge is compounded by today’s manual process of seeking out patterns within confusing line charts (Figure 1), or alternatively individually reviewing dozens or hundreds of charts.

A common notion is that focusing on bottom performers will improve overall system performance. But this isn’t always the case. All things being equal, managing bottom performers would be the right approach, but in transportation operations, all things are rarely equal.

Given the amount of variables to manage, there is no one-size-fits-all model to improve facility performance that will fit every company. Below are three scenarios that illustrate the challenge of simply focusing on bottom performers as a way to improve the network as a whole. Each required deeper insights into data, improved decision analytics and revamped focus for successfully managing overall system performance.

The Rollercoaster

In this scenario, a carrier developed a financial metric to determine the overall operational effectiveness of its facilities. This approach considered the amount of shipments compared to controllable costs (driver labor, dock labor, cargo claims, work comp for recent injuries and the number of accidents). Calculations were applied on a daily, weekly and monthly basis, and used by senior management to rack and stack all of the facilities. In the case of several facilities, a distinct pattern occurred resembling a roller-coaster ride.

At one facility in particular, the rollercoaster effect seemed to occur on a regular cycle. This large facility was critical to the operational effectiveness of the entire system due to the transport volumes. Each time this facility hit the bottom of the list (showing up in the bottom 10 of more than 300 facilities), an improvement team of field engineers coupled with regional management and one or two facilities managers from other locations was deployed to the site.

Generally, the team would stay in place for three months which generally moved the facility from the bottom 10 to the top 10. While the improvement in performance was dramatic (and relatively quick), it required an investment of 15 people traveling full time for three months. After the improvement team completed its visit, performance stabilized and the facility would rank among the top 5% of the system. Then the trouble started.

Performance downshifted but still remained in the top 25%, and the results were erratic with increasing variation. Over the several weeks and months, the variation continued, and the overall performance trended downward. Eventually, the facility dropped to the bottom 10 and stayed until the entire cycle started over with deployment of the improvement team to the site. In fact, this process seemed to occur on a two- to three-year basis for their 15 largest facilities.

The intense focus on bottom performers prevented operations management from seeing the fall until the facility hit bottom. One key to managing hundreds of facilities is to track not only the bottom performers, but also to develop tools that identify downward trends within individual facilities (Figure 2). Early intervention with downward trending facilities is much less expensive to correct than waiting for the facility to reach the bottom.


For one carrier focused on the next-day regional market, on-time service was, at one time, a key operational metric: the goal being to deliver 98% of all shipments on the service date. The carrier enforced the on-time service standard strictly with no tenders or traps permitted as reasons for late delivery.

The system performance wasn’t hitting target by approximately 2%. The facilities were ranked by service, and the carrier’s VP of operations held daily meetings with the managers of the five poorest performing facilities. The meetings were tense and took a toll on employee morale. Ironically, the bottom five facilities were all improving on a week-over-week basis, yet the metrics were still showing poor performance relative to the other facilities. In actuality, the bottom performers were not dragging the on-time service metric down; the top and middle performers were not maintaining or improving their own on time service records, resulting in the overall 2% target shortfall.

A change of approach was needed. The focus moved to shifts in performance. The mean on-time service was charted weekly for the prior four months for each facility, and the statistical tests were used to identify shifts from the mean.

Fig 3 – On-Time Delivery Percentage Identifying shifts and trends in performance focuses attention on ILT (red line) instead of the bottom performer PTL (green line).

For instance, one facility had a mean on time service of 99.4% over the prior four months, but during the period when the system on-time performance was 96%, this facility had six weeks of 98.8% on time service (Figure 3). While the facility outperformed both the system average and target performance, it was actually a significant contributor to missing the overall system goal.

Yard Stick

A weights and inspection group (W&I) had more than 100 inspectors responsible for finding misclassified freight. These inspectors were ranked from highest revenue to lowest revenue on a weekly and monthly basis, given the company’s goal to increase revenue.

With each ranking, the inspectors in Chicago, Dallas, Atlanta and Los Angeles were ranked at the top. The bottom performers included inspectors in San Antonio, Oklahoma City, Miami, Queens, NY, High Point and Tampa. Focused entirely on the bottom performers, the W&I management team couldn’t drive improvements despite extensive training, coaching and performance management interventions.

Eventually, a new director of W&I was hired, and after several months of analyzing the process and the data, he changed the metrics. All inspectors were evaluated against total revenue, but high volume facilities such as Chicago, Dallas, Atlanta and Los Angeles had more opportunities for successful inspections. A linear regression model, an approach that analyzes the relationships between multiple variables, was developed which accounted for a number of factors including inbound shipments, outbound shipments, transfer shipments, etc.

This model, based on historical data, also predicted expected revenue. This new metric for expected revenue entirely changed inspector rankings. Additionally, control charts were developed to illustrate inspector trends.

The result of using a new yard stick showed that the inspectors in Chicago and Los Angeles were average. The inspectors in Atlanta, Oklahoma City and San Antonio were outstanding compared to expected revenue. Surprisingly, the inspectors in Dallas were performing significantly below expectation. Training, coaching and performance management were applied to these new bottom performers. As performance went up, interventions were eased. At the same time, the control charts ensured that the top performers stayed the top performers. As a result of this effort, inspection revenue increased 25% in just three years.

One side benefit of having the right metric was that modeling could take place for facilities that did not have inspectors. These models could project the revenue to determine where to deploy new inspectors rather than use a system average.


The amount of variables that affect transportation operations make it difficult for companies to hone in on the exact challenges affecting performance without a more analytical and comprehensive way to determine root causes.

One challenge with tracking individual facility or employee performance is the noise from so many charts. Looking at charts for 300+ facilities or over 100+ W&I inspectors isn’t practical. Instead, decision analytics tools that identify trends or shifts will narrow the focus to only those facilities experiencing changes in performance. Using these analytics tools allows for management by exception, eliminating the need to investigate facilities with tight process controls.

Effective management in the transportation environment requires more than a rank stacking of metrics from top to bottom. Ensuring the right metrics are in place and understanding them over a period of time to assess trends is essential to success.