Why a "Coverage Approach" to RFPs Doesn't Pay: The Hidden Costs of Ghost Lanes

Conducting or participating in an annual freight procurement request for proposal (RFP) can feel inefficient and time consuming. Shippers need to establish rates for anticipated demand over the next year and carriers plan capacity to serve demand and build out their networks. But it’s often a time-consuming process—how can we find ways to reduce the burden?

I recently had the opportunity to sit down with Dr. Angi Acocella, research affiliate at MIT’s Center for Transportation & Logistics, to discuss her work as part of the MIT FreightLab. In searching for an answer for our earlier question on improving RFPs, Angi’s research focused on the impact “ghost lanes” can have on supply chains*. Here are some of the key insights that came up about her research during our conversation.

Steve: So Angi, what exactly are ghost lanes?

These are lanes procured via contract during an RFP, but on which no volume actually ends up materializing. The contracts established during the bid are never used. This inflates the RFP and takes focus and already overburdened resources away from more important lanes.

If there’s no volume, why are ghost lanes included in an RFP at all?

Often shippers use ghost lanes as a coverage strategy to ensure any anticipated demand that may come up has a contract rate on file with a provider. This strategy can help reduce shippers’ risk of exposure in the more volatile spot market.

And in some cases, volume does materialize. In theory, the shipper has a carrier lined up at a pre-set price if this occurs. Shippers expect the upfront costs of procuring all potential ghost lanes to be less than the downstream costs of searching for a carrier, negotiating a price, loading the routing guide, and setting up payment channels for the small number of potential ghost lanes that actually materialize.

However, this approach makes carrier operations more challenging. The contracted carrier has planned for this anticipated demand. When it does not materialize and ghost lanes result, it can throw carriers’ networks out of balance. This is closely related to previous work on carrier reciprocity in dynamic freight markets—the two most common reasons carriers reject freight are inconsistent demand and contract price competitiveness.

Steve: Are there hidden costs to shippers using a coverage strategy that results in ghosting their carriers?

The short answer is yes. Not only is the process inefficient during annual freight procurement, but it may have unanticipated costs down the road. These costs include the additional and unnecessary resources spent upfront during the freight procurement process and higher contract rates later on.

Steve: Let’s start with the size of the problem. How frequently do ghost lanes occur?

Ghost lanes happen at a frightening rate. We analyzed five years of truckload freight transaction data provided by TMC, a division of C.H. Robinson, to explore the ghost lane phenomenon and find possible solutions for this inefficient practice.

The information covers the complete tendering sequence for more than 3 million loads, tendered to almost 2,000 primary carriers. In addition, the dataset covers the outcome of the RFP procurement events over this time. That is, the contract price and carrier for all lanes in each shipper's network.

First, we identified ghost lanes and determined the percentage of total lanes procured that end up as ghost lanes each bid cycle. As I mentioned earlier, ghost lanes are defined as those that are procured during the strategic freight procurement event (the “first stage” of truckload transportation) and on which no loads are tendered during the execution of the contract (the “second stage” of truckload transportation).

This is depicted in the upper right quadrant of the Ghost vs. Materialized Matrix in Figure 1 below. Conversely, a planned, materialized lane is one that is also procured during the first stage, but on which at least one load is tendered during the second stage contract period (upper left quadrant in Figure 1).

Finally, lanes which are unplanned are those that are not procured during the first stage (no carrier is contracted on them) but volume appears at some point during the second stage before the next procurement event. These were outside the scope of this study, but were explored in other FreightLab projects.

Ghost vs. Materialized Lane Matrix

Figure 1: Ghost vs. Materialized Lane Matrix

Ghost lanes appear much more frequently than transportation practitioners may have realized. In fact, 65–80% of lanes contracted during the procurement event become ghost lanes. This is consistent year to year (see Figure 2).

Ghost vs. materialized lane outcomes as percentage of all lanes in the annual RFP

Figure 2: Ghost vs. materialized lane outcomes as percentage of all lanes in the annual RFP*

*Partial ghost lanes are those that have been contracted with more than one carrier and at least one of those carriers receives no volume; from that carrier’s perspective, it is a ghost lane.

Steve: So, what do ghost lanes cost?

Specifically, if a contracted carrier ends up with a high percentage of ghost lanes, does that impact performance or contract bid costs the following year?

We modeled carriers’ behaviors based on the rate at which their contracted lanes in a given year become ghost lanes and found they indeed respond negatively during the next bid cycle. A higher rate of ghosting to primary carriers in one year contributes to higher, less competitive contract prices from that carrier on average across all lanes won in the following year.

In fact, every 10 percentage point increase in ghost lanes to a carrier leads to 1% higher contract prices across all lanes from that carrier the following year, compared to what it would have bid at the lower ghost rate.

Therefore, not only are shippers paying the price of including unnecessary lanes in their RFP in terms of resources allocated, but their contract rates the following year for these carriers are higher.

Steve: Can ghost lanes lead to low carrier acceptance?

We did not find strong evidence that ghost lanes factor into carriers’ freight acceptance rates. However, this research along with previous studies on carrier reciprocity demonstrate the most important contributing factor to acceptance rate is the competitiveness of the contract price relative to the current market conditions. Accordingly, we controlled for this factor in our models.

While carriers are not deprioritizing freight after being ghosted at a high rate, they are requiring higher contract prices going forward.

Steve: How do we identify potential ghost lanes and remove them from the RFP before they occur?

We built a predictive model to characterize which lanes end up as ghost lanes based on their attributes, the market condition the year they are procured, and the carriers they’re assigned to.

Most (in fact, 85%) ghost lanes tend to be new lanes—that is, those that have not been bid out by the shipper the previous year (see Figure 3). For example, a new facility may be set to open the upcoming year and transportation needs may be expected. Delays may push the facility opening farther back and the demand may never materialize, resulting in a ghost lane.

Moreover, of the ghost lanes that are not new that year (i.e., they were procured the previous year) most had no volume materialize on them that previous year—in other words, they were ghost lanes the previous year too.

Ghost lanes' previous year volume

Figure 3: Ghost lanes' previous year volume

In addition, only 5% of new lanes actually have volume materialize on them. So, while 85% of ghost lanes are new that year, 95% of new lanes end up as ghost lanes (see Figure 4).

Overlap of ghost and new lanes

Figure 4: Overlap of ghost and new lanes

A major concern is that the established contracts do not perform well on the 5% of new lanes with freight that materialized. In fact, across all years, carrier freight acceptance rates for volume on new lanes sits at 73%.

Perhaps more importantly, the contract prices established on these new lanes are too high. On average, contract prices for loads that do materialize on new lanes are 13–40% higher than the spot market price for that lane at the time the load is hauled. This finding holds across new lanes that materialize at low volumes (<52 loads per year, roughly 80% of our sample) and high volumes (>52 loads per year). Carriers have likely factored into their pricing formulation that these lanes are uncertain and have established prices that reflect the risk they assume.

Ultimately, our results suggest the coverage strategy of establishing contracts on new lanes not only hurts the carrier’s network balance but also ends up being less cost effective than putting those loads in the spot market .

Steve: How can shippers reduce the effects of ghost lanes?

Ghosting carriers has future cost implications for shippers. The research demonstrates ghost lanes are an inefficient and resource-draining part of the freight procurement process, and the coverage strategy is not as effective as shippers may have hoped. So, we suggest some strategies to eliminate them and improve procurement efficiency.

The main recommendation I can make to shippers about to enter an RFP process is to remove new lanes from the main procurement events. Instead, wait to procure new lanes once the volume does materialize with specialized “mini bids.” Alternatively, they can turn to digital brokerage options that are better suited to cover unexpected demand.

Second, review the previous year’s ghost lane rates to incumbent carriers to identify those that may need strengthened relationships and more competitive contract prices to overcome any potentially harmful ghosting behaviors from the past.

Steve: This has been enlightening, is there more research to come on this topic?

While this research highlights the startling frequency and effects of ghost lanes, it does have limitations. For example, while we modeled carrier freight acceptance and pricing decisions, we lacked visibility into the carriers’ existing network of business, which is likely a significant contributor to the behaviors we modeled.

We encourage interested parties to reach out to discuss this research and related projects. For more information on other FreightLab research, contact the authors.

For a more detailed account of the research described in this post see the working paper: The Hidden Cost of Not-So-Friendly Ghost Lanes by Angela Acocella and Chris Caplice, July 2022.

*Data source is from TMC, a Division of C.H. Robinson.

Steve Raetz
Director, Research & Market Intelligence
Stay up to date with our latest blog posts

Our carrier experts will answer your questions, provide advice and offer insights.