Determining the amount of cargo the freight forwarder Flexport moves, as well as the capacity it has secured directly with liner carriers, has been an elusive exercise, but an analysis provided to JOC.com suggests the company is moving roughly 80 percent of the US import volume it estimated in April 2018.
In a forwarding industry where market relevance is typically measured by one variable — volume moved on forwarders’ own contracts with ocean carriers — the topic of how much volume Flexport moves, as well as what percentage of that volume it moves under direct contracts, is of ongoing significance. This all matters because customer volume and large direct contracts with carriers traditionally translates into pricing power.
San Francisco-based Flexport’s volume has been a topic of widespread industry speculation and debate, primarily because the company is the most heavily venture capital (VC)-backed logistics services provider ever, underlined by its recent $1 billion funding round, led by the SoftBank Vision Fund. Flexport has not — and likely will not — disclose its container volumes, and as a privately held company, is under no obligation to do so.
But in April 2018, the firm intimated publicly that its 2018 volume would be at least 108,000 TEU, extrapolated from the 9,000 TEU per month it said it was moving at the time. The projection prompted skepticism among forwarding competitors, ocean carriers, and shippers, and an analysis of advanced manifest system (AMS) data from PIERS, a sister product of JOC.com within IHS Markit, also called into question whether Flexport was indeed moving the container volumes it was claiming.
'A real presence'
A more recent analysis conducted by Steve Ferreira, founder and CEO of freight audit services provider Ocean Audit Inc., estimates that Flexport actually moved 80,336 TEU of US imports in 2018. That figure includes volume moved under Flexport house bills of lading as well as volume moved via the company’s network of agent non-vessel-operating common carriers (NVOs). It does not include “cargo under management,” volume where Flexport is providing some service related to the lifecycle of a shipment, but where those TEU would be moving under its clients’ direct contracts with carriers.
Ferreira’s estimate compares with just 12,462 TEU that show up in PIERS AMS data for 2018. The source of that discrepancy is up for debate. Ferreira believes it’s due to Flexport masking data on its AMS filings, data his team was able to unspool. A Flexport spokesperson told JOC.com that the nature of its agency model means the volume under its control isn’t well reflected in AMS data.
Ocean Audit identifies and recovers ocean freight spend for 15 Fortune 100 companies and six of the top 20 global forwarders. Ferreira’s process isolates historic vessel voyages where 90 percent of the invoices contained errors that resulted in overcharges. According to Ferreira, in the majority of cases, the overcharge was expressed in such a way that a shipper’s audit process would not have detected the anomaly. Ocean Audit has a list of all vessel voyages impacted, so beneficial cargo owners (BCOs) can hone in on process improvement in this critical area of control.
Ferreira told JOC.com he considers Flexport’s volume impressive for a company its age (the company was founded in 2013), especially in light of skepticism about its size. Comparing his estimate of the company’s actual volume with PIERS data on other forwarders, Flexport in 2018 would have ranked 21st among US import NVOs, ahead of major names such as CEVA Logistics, Kintetsu World Express, Nippon Express, and Damco.
“So much talk in the LinkedIn chat boards tended to downplay Flexport’s actual TEUs, but this shows they have a real presence in the market and the possibilities their funding affords them could make them a formidable player,” he said.
The forwarder has said that extrapolated volume estimate from April 2018 is low, as its business ramped up in the second half of 2018. Ben Braverman, Flexport chief revenue officer, said in an early April podcast with the technology forum SaaStr that the forwarder brought in $472 million in revenue in 2018 and expects that figure to grow to between $950 million and $1 billion in 2019. According to a Flexport spokesperson, Ferreira’s estimates are inaccurate, as are estimates of annual volume in the 100,000- to 130,000-TEU range JOC.com presented to the company.
“As a policy Flexport doesn’t disclose volumes, but I can tell you our 2018 volume vastly exceeded these estimates,” the spokesperson told JOC.com. “We're confident of our volume; it's based on multiple metrics and data sources that together paint a complete picture.”
Flexport said Ferreira’s estimate is inaccurate because the forwarder “works with booking offices around the world that work with our local control tower teams on the Flexport platform. This approach gives us access to freight booking data that allows us to manage and control service outcomes for our clients. Many of our booking partners prefer to use their own credentials to submit U.S. manifest filings and we permit them to do so. As a result, freight booked on our contracts and manifested by our partners does not appear in PIERS or ImportGenius as Flexport volume.” ImportGenius is another AMS data provider originally co-founded by Flexport CEO Ryan Petersen (he currently is a minority shareholder in the company).
Defining NVO volume
Also coloring the discussion is a lack of clarity around what percentage of the company’s volume lies outside of US imports. AMS data capture import cargo that passes through a US port, even if that cargo has a final destination outside the United States. However, the data do not cover container volume that moves directly from a foreign port to Canada and Mexico or between non-US ports.
If a considerable percentage of Flexport’s volume doesn’t touch a US port, it’s difficult to independently gauge the company’s total volume. Flexport declined to specify how much of its volume falls into this foreign-to-foreign category.
Then there’s US export volume, which also doesn’t appear in AMS data. Braverman said in the SaaStr podcast that Flexport will “book 65,000 containers” of US export cargo for one customer alone, pulp and paper manufacturer Georgia-Pacific. The company didn’t specify how much volume that represents in TEU, nor what portion of that volume is through Flexport’s NVO business versus what is cargo under management.
Cargo under management also explains some of the discrepancy. Flexport confirmed to JOC.com that its internal volume figures include both NVO volume and cargo under management, but standard practice for freight forwarders is to only consider volume moved under its own bills of lading as “transport volume,” multiple forwarders confirmed to JOC.com. This may be a case where Flexport is deviating from such a practice, whereas Ferreira’s estimate of 80,336 TEU moved by Flexport in 2018 only includes freight moved on Flexport’s NVO contracts and on its agents’ contracts.
While the volume number Ferreira unearthed might seem high to Flexport’s skeptics and low to its backers, Ferreira said rigorous analysis of the company’s volume data shows that it may also have been incorrectly counting consolidation volume by considering each bill of lading as a separate container.
For consolidated cargo — also known as groupage — Flexport might have as many as 10 different invoices for the same container, and Ferreira suspects this is one of the underlying reasons for the difference between his estimates and those of Flexport. “A ‘groupage’ or consolidated container may contain two or more importers [as it relates to the NVO’s billing],” he said. “In Flexport’s case, there were significant numbers of consolidated or groupage containers which had 10 Flexport BCOs, each one with a unique Flexport bill of lading or invoice number.”
His theory is that Flexport, in those groupage scenarios, may have counted the same container 10 different times. “On the 10 BCO invoices, each invoice is a unique number and on every one of those unique invoice numbers is imprinted the same container,” Ferreira said. “So if I’m an external auditor and don’t know that much about container shipping, I may be counting invoices and saying, ‘Okay, it’s one invoice, it’s one container. It’s two invoices, it’s two containers, etc.’
“In reality, with NVOs, counting invoice numbers and the container number or numbers on those invoices, you are always going to have the possibility that the same container number is going to show up in another Flexport region with another BCO because of the groupage or consolidated ‘mirage’ effect.”
When Ferreira adds those duplicate containers back into his figures, he arrives at a number much closer to what Flexport said publicly in April 2018: approximately 100,000 TEU.
“Contrary to much of the industry, the vast majority of our LCL [less-than-containerload] shipments are owned by Flexport end-to-end so that we can better control customer service, reliability, and speed,” the Flexport spokesperson said. “All shipments in a container filled with LCL freight are on Flexport bills of lading, so you could see up to eight bills of lading per container. Therefore any analysis that sees containers on Flexport paper as representative of our full volume would be massively overweight on LCL and is based on a misunderstanding of our consolidation network.”
Then there’s the issue of holding back on advanced manifest data. JOC.com found in 2018 that Flexport had been incorrectly masking its name in the AMS data, and according to Ferreira, the company is still doing so (or at least attempting to). But those efforts have gotten “sloppy” due to how quickly the business grew, he said. “Quality control is the killer as to why Flexport has revealed so much,” Ferreira said.
The Flexport spokesperson said the company “does not and cannot hide its AMS data. The request to mask data was filed by Flexport in error and promptly rescinded.” But Ferreira pointed to multiple master bills of lading in 2019 in which the forwarder’s name was masked as the NVO.
Courting cargo owners
Working backward from identifying characteristics in shipment documentation, Ferreira also determined that Flexport’s US import minimum quantity commitments (MQCs) with liner carriers in 2018 amounted to a collective 55,980 TEU among eight carriers, with the largest proportion (25,000 TEU) committed to Japan’s Ocean Network Express (ONE).
Ferreira and his team were able to piece together Flexport’s volume and MQC information through identifying data fragments that didn’t get screened out, such as marks and numbers on containers, Flexport’s Standard Carrier Alpha Code (SCAC), the codes of its agents, and mentions of Flexport (or Flexport derivatives) in “arrival notice” and “notify party” fields.
Those MQCs, in aggregate, are a good proxy for the amount of volume Flexport believed it could move directly with carriers, and not with neutral NVOs, either through named accounts for specific customers or via its own consolidation services. Flexport works with agents in China, primarily D.T. Logistics Hong Kong and Flexible Transportation, volume that wouldn’t normally be moved via direct carrier contracts.
The Flexport spokesperson also said the MQC estimate is inaccurate, reiterating that its policy is not to disclose volume information. Again, the issue of what is discoverable in US import volume might be at play here. For instance, Ferreira discovered a 700-TEU MQC with APL in the current contract year that covers volume from five countries in Asia to Canada.
Ferreira also assembled a list of Flexport’s top customers in its capacity as an NVO and found the company has primarily eschewed the large-volume shippers typically targeted by both aspiring and established NVOs in the trans-Pacific trade. Instead, Flexport has relied primarily on small suppliers, rising names in e-commerce, and sellers in Asia using the Fulfillment by Amazon (FBA) program.
Many of these companies — especially newer e-commerce sellers — may have logistics managers with little experience in dealing with a range of NVOs, or might be too small to attract the direct attention of liner carriers. They might also be attracted to Flexport’s slick user interface.
Why is this volume information important? The $1 billion investment led by SoftBank confers a new level of legitimacy on Flexport and gives the company an unprecedented amount of capital to pursue market share. Other forwarders have had larger operating and net margins, but none have had so much cash on hand with so few strings attached. Unlike forwarders with private equity owners, public shareholders, or family-ownership structures, Flexport’s mandate is to spend cash, scale quickly, and not immediately worry about generating profits.
The latest investment round, Flexport’s fourth, brings its total funding to $1.3 billion, but sources in the VC world tell JOC.com this won’t be the company’s last. Those sources say Flexport likely needs to burn through that $1 billion over the next 18-24 months in order to get to its next — and presumably even larger — funding round.
How much cargo Flexport actually moves, and whether the company portrayed itself as larger than it was, still matters in terms of how large BCOs might perceive working with them. But if Flexport uses its capital to subsidize aggressive pricing, build a broader personnel footprint, and provide market-leading customer service, BCOs likely won’t care how the company got there.
The volume that Flexport moves also has repercussions on the overall market, since it ties into pricing power with the liner carriers from which it buys capacity, as well as reputational power with BCOs.
Ferreira suggested it might be good for BCOs to add Flexport in their request for proposal (RFP) cycle this year because there’s a “better than average possibility that, given the volume it is actually moving, Flexport’s buy rates with liner carriers will improve to give Flexport a better sell rate.”
Petersen said in a February interview with JOC.com that the company would seek to use its $1 billion investment to grow market share.
“We don’t need to make money on every single shipment,” he said. “We’ve always operated on the philosophy that we’ll be cash flow positive on every shipment. And now you can start to look at it strategically as a network and say, 'Where do I need baseload volumes?' and be willing to go below cost for that customer in order to get another customer on board. We haven’t had the balance sheet to play that game.”
Ferreira predicted that Flexport will use its latest funding to attract midsize shippers with rate and margin leverage. “Is it possible or probable that Flexport might take a loss on the underlying ocean freight in order to build volume and adjust its margins with ancillary services?” he said. “Even if it does not improve [their buy rates with the carriers], they are still a formidable force as they can respond to RFPs using artificial numbers, as if they had a huge TEU commitment to the [carriers].”
To scale quickly and develop a dominant market share in key trade lanes, Flexport will eventually need large-volume global importers to become customers. Its largest customer in 2018, Gerber Plumbing Fixtures, moved 4,178 TEU. Although top-five global forwarders would welcome such a customer, it wouldn’t make or break the forwarder’s balance sheet.
To put Flexport’s book of business in perspective, Ferreira compared it to a similarly-sized NVO, Orient Star Transport, according to his estimation of Flexport’s actual volume. Orient Star’s 50 largest customers accounted for about 28,000 TEU (23 percent) of its annualized volume, while Flexport’s top 50 customers made up about 30,000 TEU (37 percent) of its annualized volume, which means the loss of an important customer would have a larger effect on overall volume.
“Flexport still has some work to do to expand its client base to mitigate any possible defections of top-50 clients,” Ferreira said.