Welcome to the XPO Logistics third quarter 2019 earnings conference call and webcast. My name is Melissa, and I will be your operator for today's call. At this time, all participants are in a listen-only mode. Later, we will conduct a question-and-answer session. If you have a question, please dial star one into your telephone keypad. Please note that this conference is being recorded. Before the call begins, let me read a brief statement on behalf of the company regarding forward-looking statements and the use of non-GAAP financial measures. During this call, the company will be making certain forward-looking statements within the meaning of applicable securities laws, which, by their nature, involve a number of risks, uncertainties and other factors that could cause actual results to differ materially from those projected in the forward-looking statements.
A discussion of factors that could cause actual results to differ materially is contained in the company's SEC filings. The forward-looking statements in the company's earnings release or made on this call are made only as of today, and the company has no obligation to update any of these forward-looking statements, except as to the extent required by law. During this call, the company may also refer to certain non-GAAP financial measures as defined under the applicable SEC rules. Reconciliations of such non-GAAP financial measures to the most comparable GAAP measures are contained in the company's earnings release and related financial tables. You can find a copy of the company's earnings release, which contains additional important information regarding forward-looking statements and non-GAAP financial measures in the Investors section on the company's website. I'll now turn the call over to Brad Jacobs. Mr. Jacobs, please go ahead.
Thank you, operator, Good morning, everybody. Thanks for joining our third quarter earnings call. I'm here in Greenwich today with Matt Fassler, our Chief Strategy Officer, and Tavio Headley, our Senior Director of Investor Relations. We delivered beats this quarter on both adjusted EBITDA and adjusted EPS. We also generated strong free cash flow. We achieved these results despite a soft macro, which is reflected in our updated revenue guidance. Our main lines of business produced important gains in the quarter. In logistics, we increased adjusted EBITDA by 11%. We continued to buy truckload capacity at better-than-market rates. We grew revenue in managed transportation by 22% with a tailwind from XPO Connect. We outperformed the market in Europe, Our LTL adjusted operating ratio was our best third quarter OR ever.
We remain firmly on track for our LTL business to generate at least $1 billion of EBITDA in 2021. Our proprietary technology is continuing to drive benefits for our customers and our shareholders. Our entire management team is highly focused on executing on 10 major levers to significantly improve our profit over the next several years. The majority of these 10 initiatives relate directly to our technology. Company-wide, we're using our tech to optimize the $5 billion variable cost opportunity in our $6.5 billion annual labor spend. For example, in LTL, we rolled out game-changing tools that use machine learning to reduce our $1.3 billion annual line-haul spend, our $650 million spend for pickup and delivery, and also reduce the $365 million we spend annually in dock operations.
In the warehouses where we've introduced advanced automation, like collaborative robots and autonomous goods-to-person systems, our teams are working 2 to 4 times more productively. On the revenue side, we're applying data science to capture pricing opportunities across our transportation modes. In total, the 10 levers represent a pool of approximately $700 million to $1 billion of potential profit growth by 2022. These are self-driven, company-specific initiatives that are largely independent of the macro. We're excited about the size of the prize and the meaningful potential uplift to our profitability. Now I'll turn the call over to Matt to discuss the quarter in more detail. Matt?
Thanks, Brad. As I review the numbers, you'll see that we're moving full speed ahead with our tech-driven initiatives across our company, with many milestones marking progress in Q3. We're confident that these initiatives will deliver significant benefits for our shareholders, both today and over time. I'll start by walking you through the third quarter numbers and our strategic focus by business unit, beginning with our transportation segment. In North American LTL, we drove strong profit growth despite sluggish demand. Yield rose 2.9%, excluding fuel. Price increases on contract renewals grew by 4%. Both of these are a continuation of the trends we saw in the first half of the year, reflecting rational market pricing. Tonnage declined by 3.1% year-over-year. This was driven by lower weight per shipment from many of our industrial customers, consistent with the macro backdrop for the U.S. industrial sector.
By contrast, weight per shipment for our customers in the consumer sector remained steady. Our adjusted operating ratio improved by 460 basis points to 80.8%. Excluding gains on sales of real estate both this year and last, our adjusted OR improved by 210 basis points to 83.5%. As Brad said, it's our best third quarter adjusted OR ever. The improvement in this ratio reflects our continued improvement in yield as well as tight cost control. Tech-driven optimizations of pricing and labor are beginning to contribute to these results. Within freight brokerage, we drove volume while remaining disciplined on margin. Our freight brokerage top line declined by 14% year-over-year, which mirrors Q2. Excluding the reduction in our brokerage business with our largest customer, our underlying freight brokerage revenue declined by just 3%.
Our net revenue margin eased across freight brokerage by 60 basis points to 18% as contractual rates declined and the spot market stabilized. In truck brokerage, specifically, the decline in our load count moderated to 4%, which is an improvement from Q2. Importantly, excluding our brokerage business with our largest customer, our growth in load count was more than 20 percentage points better in the high teens. This is an acceleration from Q2 and reflects a significant gain in share. We're continuing to source truckload capacity at prices better than the market. In the third quarter, we procured capacity at about 3% better than the DAT benchmark. September was the strongest month in the quarter for truck brokerage, both in terms of the year-over-year revenue and net revenue trends.
Our digital freight marketplace is enhancing our ability to source competitively, deliver compelling value, and create an appealing user experience for our customers and carriers. Turning to our last mile operation. Revenue overall declined 19% year-over-year. This reflects the impact of winding down our postal injection business in Q1. We grew our core heavy goods business and excluding postal injection, last mile revenue growth accelerated to 4% as we onboarded new business in recent quarters. Net revenue margin in last mile increased sharply year-over-year, primarily due to mix as we discontinued our postal injection business earlier in the year and as we increased margin for our core heavy goods business. Given our recent new business wins, we expect last mile organic revenue growth to continue into the fourth quarter.
As we mentioned earlier, managed transportation posted strong sales growth in Q3 and remains one of the fastest-growing lines of business at our company. Our European transportation business had a revenue decline of 1.8% in the quarter. FX was responsible for about five percentage points of the decline. Transportation revenue trends in Europe measured in local currency were relatively consistent quarter-to-quarter, which is a solid accomplishment in a tough backdrop. We continue to generate the most growth in the U.K. and Spain. Looking across our transportation segment overall, revenue was down 5.8%, adjusted EBITDA was up 2.1% and adjusted EBITDA margin rose by 100 basis points to 12.4%, driven by improved profitability in North American LTL, managed transportation, and last mile. FX constrained this number by about $2 million, or a little less than one percentage point. Turning to the logistics segment.
In North America, our 4% revenue growth reflected a strong food and beverage peak, as well as ongoing strength in consumer packaged goods and aerospace. If we exclude our business with our largest customer, our underlying logistics revenue growth in North America moves up to about 7%. In Europe, logistics revenue declined 3.5%. FX had a negative impact of about five percentage points. Once again, e-commerce was our strongest vertical in European logistics. We're continuing to build on our market-leading position as the largest outsource provider of e-fulfillment logistics in Europe. Adjusted EBITDA for logistics as a whole rose 11% year-over-year in the quarter, and adjusted EBITDA margin rose by 100 basis points to 9.4%. The combined impact of FX and the downsizing of our largest customer cost us approximately $11 million of EBITDA.
We believe that the exclusion of these items more accurately frames the underlying growth in our logistics business. We remain excited about the growth we're seeing from our XPO Direct shared distribution platform. This offering brings together the best of our supply chain and transportation capabilities, and both our current and prospective customers share our optimism for its growth outlook. New business wins declined 2.7% year-over-year, improving from a 5.2% decline in the second quarter. Year-to-date, new business wins are up in the mid-single digits in North America and down by a similar rate in Europe. Company-wide, our sales pipeline remains above $4 billion for the third consecutive quarter. We have an important seat at the table for new business opportunities that we expect to pay off when the macro rebounds. Moving down the income statement.
Interest expense rose to $75 million from $51 million a year ago, reflecting our debt issuance earlier in the year to fund our share buyback program. Our effective tax rate improved to 20% from 26.2% a year ago, due primarily to foreign currency losses recognized. We now expect an effective tax rate for full year 2019 between 23% and 25%, which is below our prior expectation of 25% to 28%. Our weighted average diluted share count was 102 million on September 30th, compared with 137 million shares a year ago. The year-over-year decrease reflects our share buyback activity. There was no meaningful change quarter-to-quarter, as we didn't repurchase any shares in Q3. Since we launched the program in December, we've bought back 35.2 million shares at an average price of $53.42, for a total cost of $1.9 billion.
Our diluted earnings per share was $1.14, up from $0.74 a year ago. Our adjusted diluted EPS rose to $1.18 from $0.89 a year ago. Our buyback activity was $0.15 accretive to adjusted EPS for Q3, and should continue to prove nicely accretive for the year. Cash flow from operations in the quarter was $278 million compared with $288 million a year ago. Gross capital expenditures increased to $177 million from $145 million a year ago. Net capital expenditures were $70 million compared with $115 million a year ago. All in, free cash flow was $257 million, a sizable increase from $173 million a year ago. We had asset sales of $107 million, including $94 million from real estate, and the remainder from equipment. $71 million came from the sale of the office property in Portland we discussed on our last earnings call.
We sold off underutilized space and leased back a more appropriate footprint at an attractive rent. The sale of Portland resulted in a gain of $14 million. In total, we booked a gain on sale of assets of $33 million in the quarter, of which $29 million related to real estate. $26 million of this was in our LTL business, including most of the Portland gain. We had a sequential benefit to free cash flow from trade receivables programs of $56 million in the quarter. For the full year, we continue to expect these programs to deliver an incremental benefit to free cash flow of $125 million-$150 million. Now, I want to bring you up to speed on some of the 10 key profit initiatives we mentioned, all of which are underway.
They represent a potential profit growth opportunity of $700 million-$1 billion by 2022, largely independent of the macro. I'll start with some context about our technology organization. We've structured our tech organization to deliver a number of important advantages. First, as a global leader with about $17 billion in revenue, we have the resources to invest in innovation on a large scale. We make one of the largest tech commitments in our industry, investing approximately $550 million a year in technology. Second, while our technology strategy is centralized under our CIO, the team itself is embedded across our operations. This provides constant feedback loops that engage our operators and customers. Third, the way our business is structured, we can deploy innovations globally across multiple operations in a relatively short time.
Fourth, because our technology organization is run as a single entity, innovations developed for one business unit can have widespread benefits. We have numerous instances of solutions created for one purpose and then applied with great success in other operations. Our XPO Smart optimization tools are a case in point. Turning to the 10 key profit improvement initiatives, I want to focus on the progress we've made with the ones that are enabled by our proprietary technology. They are advanced automation and intelligent machines deployed for our contract logistics customers, our XPO Smart workforce productivity tools, which we're applying across our logistics and LTL operations. LTL process improvements for the optimization of pickup and delivery and our line haul network, pricing analytics, and our XPO Connect digital freight marketplace. I'll give you a little more color on each of these initiatives.
Across our supply chain offering, advanced automation delivers critical improvements in speed, control, safety, accuracy, and productivity. Productivity improvements are particularly valuable in tight labor markets, where wage inflation and labor shortages can erode margins. Many of the U.S.' largest markets for warehousing and distribution, such as Atlanta, Columbus, Dallas, Eastern Pennsylvania, and Northern California, have significant labor constraints. In the third quarter, we increased cobot deployments in our warehouses in preparation for this year's retail holiday peak. Once peak passes, we can repurpose these cobots for new customers and startups in 2020 and on from there. With XPO Smart, we've already increased workforce productivity in our logistics warehouses and on our LTL docks where it's deployed. This technology provides a bird's eye view of operations. Our managers can optimize labor levels through the day, enhancing customer service in the process.
Before we launched XPO Smart, we had been managing warehouse labor spend through a combination of tribal knowledge and reactive analytics. Now we have deep visibility into performance, and we're solving productivity gaps in real time. The results to date are significant. Labor efficiency improvements in our warehouses of more than 5% on average, increases in motor moves per hour on our LTL docks of more than 10%, and employees who are more energized and engaged in producing a winning performance. To date, we've rolled out XPO Smart in approximately 100 of our logistics warehouses in North America and expect to roll it to the remaining 200 sites over the next 12 to 15 months. Europe has a five-site pilot underway. In LTL, we've deployed the technology on docks in approximately 20 locations and expect to have it in all 290 of our North American LTL service centers by year-end.
We also have broader LTL tech initiatives focused on simplifying our processes with real-time analytics, driving better productivity, enhancing customer satisfaction, and positioning XPO as the clear innovation leader in the LTL industry. In line haul, bypass optimization is letting us deploy trucks deeper into the network so that loads go directly to their destination. The freight is handled fewer times, which saves time and cost. In the third quarter, we saw solid improvement in our load factor, due in part to these line haul initiatives. Our optimization of pickup and delivery adjusts our routing to traffic in real time, accommodates late-breaking customer requests, and generally manages surprises. We've created a more intuitive interface for our dispatchers, and we use machine learning to predict loading and unloading windows based on customer service histories. Our dispatchers have visibility into the profit impact of route adjustments, which enhances their visibility to manage exceptions.
We're piloting this technology this quarter with most of the benefit coming in 2020 and beyond. On the pricing front, we're using elasticity models in LTL to adjust for customers, lanes, and current conditions. Our algorithms incorporate data from past customer behavior and real-time market dynamics. The goal is to price as intelligently as possible and balance the network. For small to mid-sized LTL accounts, we've greatly improved the software that we use to price lanes. Our local account executives can deliver prices in real time, which aids our ability to capture share. Currently, we provide quotes on nearly 50% of the freight in the U.S. LTL market and convert less than 1/4 of those quotes into revenue.
In the first half of 2019, without the benefit of our optimization tools, we captured $1.9 billion of revenue on $9 billion of rate quotes out of an estimated $18.9 billion total market size. As the math suggests, we have a large EBITDA opportunity here. Moving on to XPO Connect. As many of you know, this is our proprietary digital freight marketplace with multimodal architecture. It's best described as an umbrella operating system for transportation with a shipper interface, a price engine, a carrier interface, and a mobile app called Drive XPO for truck drivers. We currently have over 37,000 carriers registered on XPO Connect. Downloads of the app doubled from quarter to quarter for the third period in a row. In other transportation initiatives, we've introduced electronic carrier scoring using KPIs, which creates service transparency and encourages on-time pickup and delivery and delivers better bounce rates.
We've launched a carrier loyalty program, which is helping to strengthen our carrier relationships and improve overall performance and retention. Finally, we've rolled out a new app specifically for last-mile drivers. As we close, I'm proud to share that XPO is the first global logistics company to join MIT's Industrial Liaison Program. Our collaboration with MIT adds another node of connectivity with emerging tech. It's an opportunity for us to realize new levels of productivity for our customers while providing input into the future of robotics and machine learning. We're also proud of our partnership with the Susan G. Komen Foundation. Soon, you'll see XPO trucks on the road sporting a familiar pink ribbon. We're supporting Komen in the fight against breast cancer because this is a cause that resonates with our employees and is central to our corporate culture.
With that, I'll turn it back to the operator, and we'll take your questions.
Thank you. At this time, we'll be conducting a question and answer session. If you'd like to ask a question, please press star one on your telephone keypad. A confirmation tone will indicate your line is in the question queue. You may press star two if you'd like to remove your question from the queue. For participants using speaker equipment, it may be necessary to pick up your handset before pressing the star key. Our first question comes from the line of Bruce Chan with Stifel. Please proceed with your question.
Yes, gentlemen. Thank you for your time. Just a quick question here on the LTL yields. When I look at some of the mixed characteristics, revenue per hundred weight was down, length of haul was up fairly meaningfully. I think both of those tend to increase the revenue per hundred weight, which would imply that core yields are maybe closer to zero. I just want to get your take. Is that indicative of what's going on in the market and the competitive dynamics? Is that more of a function of some of your AI and machine learning-enabled pricing tools?
Bruce, we think that pricing in the LTL market remains healthy and rational. Our yield trends through the year have been consistent. Our price increases on contract renewals increased 4%. We think an indicator of consistency in the pricing environment. We are deploying our pricing analytics in LTL. As we've mentioned, our work in price elasticity has been essential for assessing any individual customer's history, as well as current market dynamics for optimizing yields. We think that pricing in the LTL environment remains consistent and healthy as it's been all year.
Okay, then just switching gears a little bit to the M&A picture, you guys have done some nice work on the margin front and the cost management front. It looks like valuations are starting to moderate a little bit out there. Is there any more appetite to get back into the acquisition fray? If so, are there any changes to your previous strategy as far as where you want to fill out some of that geographic and/or service white space?
We are in very early stages of conversations with potential sellers. We are very disciplined in general, and specifically with respect to M&A. In M&A, we have a very methodical process. We've begun that process. There's no timeframe. M&A is something that we like to talk about retroactively rather than prospectively. In general, to answer your question about the types of deals we like to look at are ones that are, A. strategically compelling. There's a reason to do it. There's a strong rationale for doing it. B. from a financial perspective, highly accretive, and something that has synergy so that, one, we delight the customer, but two, also there's a clear path to improve both the EBITDA and the free cash flow that we're acquiring. There's nothing to talk about of substance at this stage.
Okay, just one final question here before I hand it over on the contract logistics side. Can you maybe give us some flavor on how that business would respond if we were to see maybe a modest recession? Obviously, business volumes would be a little bit lighter, but I would imagine that there would be more of an appetite to outsource from shippers that are seeking some cost savings.
I think to some extent, you're already seeing how these businesses act in a downturn. We've been in a downturn. We've been in a downturn in general for quite a while, about a year in industrial. Consumer is still strong, but the rest of the economy has not been so hot for quite a while. Remember, the tariffs started in March 2018. Nothing really happened. People weren't paying close attention to them for a good two, three quarters. People started seeing the effects. That's been building up, and now you've seen the business, in our case, be able to generate positive year-over-year comps, even beats in a slow economy.
In supply chain specifically, there is an interesting phenomenon of what happened in the great financial crisis in the largest of the four contract logistics companies that we bought, New Breed, where their EBITDA actually went up, and up substantially in 2008 and 2009. We'll see what type of recession we have and when we have it. So far, so good.
All right, great. Appreciate the time, and congrats on a nice result in a tough market.
Thank you, sir.
Thank you. Our next question comes from the line of Amit Mehrotra with Deutsche Bank. Please proceed with your question.
Hi. Thanks. Good morning. Thanks for taking the question. First question, I just wanted to ask about the cadence of earnings with respect to the fourth quarter. The full year guidance, even at just the low end, implies a sequential uptick in the EBITDA from what you did in the third quarter, which I don't think has ever happened. I would argue the business is probably less seasonally four Q-focused now, given the loss of postal injection. I understand there's a lot of cost opportunities. If you can just walk us through maybe the puts and takes on the sequential walk that gets you there and just how confident you are in achieving the full year EBITDA guidance. Thanks.
I think you're absolutely right in mentioning postal injection. That was a big bad guy in the fourth quarter of last year. We don't have that this year. Just walking into the quarter, we've got that wind to our back with not having both postal injection in France. France, the whole yellow vest started.
Yeah, I'm talking about the sequential, Brad. The sequential walk is what I'm talking about.
Sequentially, last year we had the hit for postal injection and the hit from France. This year, we're not going to have that. The full quarter benefit of the cost out that we did so well in the third quarter didn't hit the whole third quarter. That layered in throughout the quarter. We will get the
Full benefit for all three months in the fourth quarter. The biggest contributors will be LTL, where we've got positive yield, got good cost control, and we've got Smart labor tools already showing good results, and they'll be rolled out to all the LTL facilities by Christmas. Secondly, European supply chain, where we've had growth in new business, and we've been very disciplined on labor costs. Last mile, we're cycling the postal injection. We've had new wins, and the consumer is still healthy here in the fourth quarter. The final thing that I would mention is the 10 levers. The 10 levers, six of which are cost control. The main one that's most exciting, and particularly for the fourth quarter, is tackling that $6.5 billion labor spend that we're very zeroed in on, making sure that we're spending that $6.5 billion very wisely.
Okay. Just to bow tie that, you basically are comfortable you have a line of sight to the earnings power of the enterprise being higher in the fourth quarter than it was in the third quarter. Is that correct?
That's correct.
Okay. Thank you for that.
That's reflected in the guidance that we gave out last night, where we brought down revenue, and we kept EBITDA, and we kept free cash flow. That's what we're seeing.
Got it. Okay. I appreciate it. Then just maybe one longer-term question, looking beyond this year into 2020. I'm not asking for specific guidance, no one has really given 2020 guidance in the trucking industry, I'm not asking you to do that. Just conceptually, given the success you're making in the new business wins and the pipeline new business is up, I don't know, 20% year-over-year or something like that. You're lapping the large customer loss at the beginning of this year. Is it fair to assume, Brad or Matt, that EBITDA growth accelerates next year? How should we think about it? There are some decent gains on sales this year impacting the outlook for 2020.
I just want to think as we look prospectively into 2020, given all the progress that you're making on the cost and the revenue side, is it conceptually right to assume EBITDA growth year-over-year accelerates in 2020 versus 2019?
We like the setup of how we're going to enter into 2020. We've got a lot of momentum from so many different projects that the team is laser-focused on. We're not going to give guidance now. We're going to give guidance once we finish the budget process and we're on the next quarterly conference call. We like the way we're set up going into next year.
Okay. I'll leave it there, guys. Thanks for taking my questions. Appreciate it.
Thank you.
Thank you. Our next question comes from the line of Chris Wetherbee with Citi. Please proceed with your question.
Hey, thanks, and good morning, guys. On the LTL side, could you give us an indication of what you guys are seeing so far here in the fourth quarter? I know it's early, but I just want to get a sense of what October trends kind of are looking like on a tonnage and pricing standpoint.
Similar as to what it was in the third quarter. Yield is a little bit better. Tonnage is a little bit worse. Of those two things, obviously, we prefer yield to be the better of the two, since there's no cost associated with it and it goes right to pre-tax. In the third quarter, we had the best OR ever. Best third quarter OR in the history of not us, but also Con-way, with or without real estate. With real estate, it was 460 basis points improvement. Without real estate, it was 210 basis points improvements. We expect to continue to see similar types of OR improvement for the balance of the year. Yield was up 2.9% in the third quarter.
We expect it to be more towards the higher end of the zone that we've had all year long at 3%-4%, at least that's how October is starting out. Price increases on renewals have been about 4%. That's the zone it's been all year, 4%-5%. I would say the star of the third quarter that we would expect to continue for the fourth quarter, given what we've seen in October, is really load factor from an operational perspective. Load factor was up 6.1% in the third quarter, and we've been getting better productivity from our new line haul modeling tools. I think for the fourth quarter, you should see OR improvement of about more than 300 basis points with real estate, more than 100 basis points without real estate.
Then as Matt was mentioning before on the workforce planning, we've now piloted XPO Smart in 20 LTL service centers. The results have been amazing. It had been very, very similar to what we saw in the warehouses, where we've had over 5% labor productivity fairly consistently, which is why we've accelerated the rollout. We'll be in all 290 LTL service centers in the next 60 days. We've got very exciting tech projects in the line haul, $1.3 billion of cost there. In P&D, $650 million of costs we're reducing there. In the dock, $365 million of cost there. Tech-enabled projects on all of those items. We've seen benefits from that in the first month of the quarter. We expect it to increase over the whole quarter.
Okay. That's very helpful. I appreciate it. You made a comment about buying it better than the DAT benchmark on the brokerage side, I think you said loads were up there. There's clearly some share gains going on. Can you talk a little bit more about the strategy on the brokerage side, what you think you might be able to accomplish from a market share perspective as we continue through this weak market? Does that strategy shift at all if we do see these early signs of firming in the truckload market maybe continue to a better, more firm market in 2020?
In brokerage, we've both lost market share and gained market share in the sense that if you take into consideration the largest customer that downsized their business with us in truck brokerage, where we were their largest truck broker, and they were our largest truck brokerage customer, now they're a small truck brokerage customer of ours, then we've lost market share. If you take that out, if you just pro forma out the largest customer downsizing, business was up. Business was up very nicely. Productivity was up. While including the largest customer downsizing, loads were down 4% year-over-year. Our headcount in brokerage was down 13%. We're seeing nice productivity improvement there. In terms of procuring capacity, that's correct. We have been procuring capacity in the third quarter at 3% better than DAT, and that's because of our XPO Connect tools.
XPO Connect is, I hate to use a cliché, but it's on fire. We had 37,000 downloads in the third quarter, which was double what it was over the second quarter. We're on track to be at 100,000 downloads by the end of the year. There's good things going on in freight brokerage. At the same time, we're not immune to what's going on in general with the truckload market.
Okay. No, that's helpful. One quick detail, one before I let you go, would be just gains on sales. How do you think about the fourth quarter? Any projections you can give us for the models?
Sure. The thought process is likely somewhere in the $20 million-$25 million range in total, and that depends on how deals end up tracking for us in Q4.
Okay. Thank you very much. Appreciate it.
Thank you.
Thank you. Our next question comes from the line of Scott Schneeberger with Oppenheimer and Company. Please proceed with your question.
Thanks very much. Good morning. E-commerce has been a big driver last year, expected to be again this year, but a bit of a different customer profile. Curious, Brad and Matt, could you address the approach this year and some of the trends you're seeing thus far?
Yes, Scott. E-com is still our largest and our fastest-growing vertical globally. It's a sticky business because we do it really well, and these are customers that value service and speed and accuracy a lot. We have very high renewal rates in that business. We are the largest e-fulfillment provider in Europe, and we capitalize on that position. Our reverse logistics business, the returns management business, is the fastest-growing subset of that fastest-growing e-commerce vertical. But we're also seeing nice gains in omnichannel. And all the tech investments that we've done to support e-commerce give us the ability to manage peak better for our e-commerce customers in periods like Black Friday, for example. And the smart labor tools have been very much appreciated by our customers because we're taking cost out. We're not just only the scale leader in e-commerce, we're also the cost leader in e-commerce.
We're going to continue pressing our advantage in e-commerce. It's working very well for us, and we'll keep doing it.
Thanks very much for that. The logistics margin expanded nicely sequentially. Could you discuss some of the drivers there? You've covered a nice job on this call discussing efficiency initiatives. I'm just curious what kind of trend we may see going forward in that segment. Thanks.
Well, you're right. EBITDA was up 10.9% in the third quarter in logistics. The margin was up 100 basis points. The revenue was down, though. It was down a half percent year-over-year. A big chunk of that was the downsizing of our largest customer, which is not all over with. We're going to see some more pressure of that going forward. I think we can see more downward pressure on revenue there, but still a strong ability to grow the bottom line strongly, healthily, despite adverse revenue trends. Capacity's been growing. We have 195 million sq ft today. That's up 5% year-over-year. It's down sequentially because of the downsizing of our largest customer.
The smart labor management tools actually originated in our logistics business before we saw it working. They said, "Hey, let's try that in LTL as well." We've got about an eighth of our 800 or so contract logistics facilities using smart labor tools right now. We are seeing 5-plus % labor productivity in those 100 or so sites. We are going to roll out smart labor management tools to all of our North American sites by the end of next year. Remember that of our $6.5 billion labor spend that we're working to improve, $3.5 billion of that is in supply chain, is in logistics. To answer your question about what's growing the most there, it's reverse logistics. That's the fastest-growing part of our company as the subset of e-com, which is the fastest one beneath that. In XPO Direct, XPO Direct continues to be on track.
We see that as a billion-dollar revenue run rate by the end of 2022. Customers love it. They love it because it lowers their cost, and it brings their inventory closer to their customers. Customers want faster shipping in a cost-effective way, and that's what XPO Direct does. A lot of positive wins there on the bottom line. Top line, I expect to be challenging.
Great. Thanks for that. I'll turn it over.
Thank you.
Thank you. Our next question comes from the line of Kevin Sterling with Seaport Global Securities. Please proceed with your question.
Thank you. Good morning, gentlemen.
Morning.
Brad, can I dig deeper into a little bit of your longer-term plans? You talk about $1 billion in EBITDA and LTL by 2021. It sounds like to me you can get there through technology improvements, through reduction in outside power, lane density. Obviously, pricings and yield seem to be pretty good, but as you know, tonnage is falling. I guess barring an all-out recession or kind of a big stinker, you're pretty comfortable you can still get to that $1 billion in EBITDA outside of just some industrial kind of slowing, if you will. Am I thinking about that right?
You're thinking about it exactly correct. Every sentence you just said, I agree with. The proof is in the pudding. We have been in an industrial recession for the last year, and I say that just based on the objective fact that not only us, but the entire LTL industry, has seen negative tonnage growth in each of the last four quarters. In a contracting market, we've still been able to post excellent numbers in LTL. I attribute that partly to the things you mentioned that are intrinsic to us, that we're doing in order to improve the business every day. We've got a great management team in LTL that's very focused on efficiencies and productivity and cost out, customer service. Also to the nature of the LTL business, which is very different than the truckload business.
In truckload, you have so many tens of thousands of carriers spreading out and sharing the capacity. In LTL, there's a handful of carriers, less than 10, that control 89% of the capacity. There's more discipline, there's more pricing power. Secondly, in the truckload business, when rates start going up, it's so easy for capacity to come in, for supply to come in. You just need a CDL and a few trucks, and bingo, you're in business. In LTL, you need a network. You need pickup and delivery vehicles. You need service centers. You need the line haul trucks. You need the whole back office. You need all the technology to plan the whole network. A lot of moving parts. That moat prevents new capacity from easily coming in. The LTL business, from our point of view, is a pretty good business.
It's a business that can perform well even in tough times, provided that the industry and managements of the leading companies are disciplined. I think 10 years ago, maybe they weren't so disciplined, and maybe they weren't so knowledgeable about how all the different levers affected the P&L. I think today, every single management of the top 10 LTL companies is very sophisticated and understands levers very well, particularly price. I expect them to be disciplined in any future downturn in just the way they've been disciplined in the last 12 months.
Got you. Thank you very much. It sounds like we're driving, or you're driving toward a sub-80 OR in LTL. Is that your ultimate goal?
It'd be nice. We're moving in that direction. It's not our main report card. Our main report card is growth in operating income year-over-year, growth in EBITDA, return on capital, and profit margin is one way to do that. We do have a major effort on utilizing our backhaul better and to reduce our empty miles. For there, you have a little pressure on yields. That does pressure the OR. I'm not going to apologize for our OR. We had the best operating ratio for a third quarter ever in the third quarter, even without including even giving 0% credit for the real estate profits we had. OR improved 210 basis points year-over-year. You're right, it's almost sub-80, but not quite, 80.8.
Yep. Okay. Lastly, any update on CFO search? Where does that stand? Anything you can share might be new there or incremental?
CFO search is like M&A. It's binary. We've either done it, or we haven't done it, and we prefer to speak about it retrospectively rather than looking in the future. The search is active. We're seeing candidates. Like in general in life, we're disciplined about it, and we're looking for that superstar CFO, and when we find that person, we'll hire them, and we'll announce it promptly.
Got you. Okay. Well, Brad, Matt, and Tavio, thanks so much for your time this morning. I appreciate it. Take care.
Thank you.
Thank you. Our next question comes from the line of Ariel Rosa with Bank of America. Please proceed with your question.
Hey, good morning, guys. Congrats on some of these results here. Brad, to start, maybe this is going back to Amit's question, looking at 2020, you're lapping the loss of your largest customer. For this year, the low end of your guidance suggests kind of 7% EBITDA growth. Given that you have some of these tailwinds, and as you said, we've been in an industrial recession for the past 12 months, and you no longer have that loss of the largest customer, is there any reason to think that 7% is kind of a baseline in terms of what we can see in terms of EBITDA growth?
Is there any reason to think that kind of where the macro is that that level of growth that we've seen this year, given all the headwinds that you faced, could be kind of a baseline as we look forward to next year or to 2021?
Nice try, Amit, we're not going to be giving next year guidance on this call, just like all of our competitors aren't giving guidance either. I like the setup going into next year. I like the very large number of internal initiatives that we've got that are idiosyncratic to us, that are company specific, that are self-held, that are largely independent of the macro, and the deep focus we've got on the management team, senior management team, and the middle management team on executing with a high level of accountability on each one of those levers. I like the way the organization is functioning, we're not going to give specific guidance at this point in time for next year.
All right, fair enough. Well, I tell you what, let me ask then different question. Moving on to the $1 billion EBITDA target for 2021, coupled with the, I think you said $700 million to $1 billion of savings from these 10 tech-driven initiatives. How do we think about those two things interlocking with each other? Is the $1 billion from EBITDA, is that additive to the $700 million to $1 billion from tech initiatives or do they kind of overlap?
They do overlap. Look at it like this. We have in the normal play of things, we have headwinds, and we have tailwinds. The main headwinds are inflation, cost inflation in particular. We have pricing pressures with our $6.5 billion of labor costs. The positive winds are things like price, volume, and everything that goes to margin expansion. Good management teams know how to function in good markets and bad markets so that you achieve all those, and the good guys outweigh the bad guys, and you grow your profits like we've done every year for many years now. In terms of additional pockets besides the normal to and fro of what I just described, we as a team have spent a considerable amount of time to say, what can we do above and beyond the normal way we run a business to improve it?
Besides normal operational excellence, what kind of strategic levers can we push that will move the needle quite a bit? We devised 10 major levers. That's a pool of $700 million to $1 billion of potential profit improvement by 2022 that we feel confident we're going to get a good chunk of that. We're not going to get $1 billion. We're not going to achieve 100% of every single lever, but we will achieve hundreds of millions of dollars from that. You can take that to the bank. These are things that have been well thought through that are very well grounded. A lot of those levers are tech-enabled. More than half of them are. What are we trying to achieve with those tech-enabled initiatives? We're trying to do things that make the customer very, very happy with us and that improve our bottom line.
We're trying to shorten distribution cycles and speed up fulfillment time and increase the accuracy of orders and inventory. We're trying to reduce the stocking cost for customers. We're trying to increase labor productivity. All these things accomplishing customer delight and bottom-line improvement. If you look at those endeavors, you can slice and dice them in different ways. I would put them into 4 categories: automation, XPO Smart, XPO Connect, and a bunch of categories that all fall into LTL. Let me briefly explain each one of those. If you look at automation, intelligent machines, advanced automation, the main needle movers are goods-to-person systems, where we've improved productivity by four to 5X, and we're in a hurry to roll them out more. The cobots, where we've increased productivity by 2X, and the robotic arms.
We've compressed order fulfillment times, in many cases, from five days down to one. These are big hits with customers, and definitely is a big hit with them around peak. XPO Smart. XPO Smart is using dynamic data science and machine learning so that we can use the information that we get from customers, use our internal big data information, not only about that customer, because sometimes we keep better records than the customer keeps about themselves, but also use records data from similarly situated customers and all the customers we have in our managed transportation business, so that we can predict what the likely volumes are going to be the next day, the next week. We can get the right staffing levels. We get the right length of the shifts. We can manage overtime correctly. We can get the right ratio of temporary to permanent workers.
We get the right ratio in LTL of drivers to dock workers. In logistics, where we've got, as I was talking about before, $3.5 billion global labor spend, we want to roll out to the other seven-eighths of the sites all over the next couple of years. In LTL in general, we're looking at three big pools of spend. We're looking at the $1.3 billion we spend in line haul. Here we're using new line haul modeling technology to build more intelligent routes for the 2.6 million miles we drive every day. We're already seeing benefits of that immediately. You saw that in the load factor this quarter, up 6%. We're attacking the pickup and delivery, $650 million spend.
There, what we're doing is we're designing intuitive tools for the dispatcher so that they can use dynamic route optimization to do three things: improve route density, reduce miles per stop, and reduce the cost per stop. Pricing, which is the most impactful of all the levers, not just in LTL, but in general, because there's no cost with it. It goes right to EBITDA. In pricing, we're using AI-based pricing algorithms that incorporates past customer behavior and real-time, in-the-moment market conditions. We can find just the right price for our RFPs, so that we're not pricing ourselves too high and losing business. We're not pricing ourselves too low and leaving money on the table. The fourth bucket is Connect, the digital freight marketplace, where we're using machine learning to match the right carrier to the right shipment at the right price.
Our competitive edge in the digital freight marketplace is the fact that we use a multimodal solution. We're not just offering brokerage or just offering one thing, but we're showing the customer truckload, intermodal, LTL, managed trans, last mile, global forwarding, expedite, most of which are on the platform, and the rest of which will be on the XPO Connect platform over the next year or so. XPO Connect's been growing at an explosive rate, and you've already seen the fruits of that. You saw us buying transportation and brokerage 3% below DAT. You saw us growing managed transportation in the quarter by 22%.
In short, I would say that nobody has as big of a sandbox to play in as we do to drive digital transformation, and that's going to benefit our sales, and it's going to benefit our customers, and it's going to benefit our shareholders.
Okay, terrific. That's really great color. Thanks for the time, Brad.
You're welcome.
Thank you. Our next question comes from the line of Brandon Oglenski with Barclays. Please proceed with your question.
Good morning, everyone, thanks for taking my question. I guess, Brad, coming off that answer, the way you guys calculate organic growth, it was negative in the quarter, but I think you said that the pipeline has increased, right? Can you just remind us the way you look at the pipeline and if that's netted against business that you think is not going to renew or that you've lost? Look, I'm not looking for guidance in 2020, but I guess when do we see resumption in growth in the business or is this just more a macro issue?
The pipeline definition is the precise pipeline definition that we use in our CRM, which is Salesforce.com. The pipeline stood at $4.3 billion at the end of Q3. It was up 19% year-over-year. It was up both here, it was up in Europe. Only the third time it's been over $4 billion since three quarters in a row. I wouldn't bring out the champagne yet because new business, the closed business was down. It was down about 2.7%. It was just shy of $900 million. Two things I would take away from that is, on the one hand, customers are taking more time to transfer, to convert. Secondly, on the other side, on the good piece of news, even though it was down, it was down 2.7%, it was down less than the 5.2% it was down in the second quarter.
I would look at it as the proverbial mongoose going through the snake. The snake is digesting the mongoose a little slower in this slower macro environment. With respect to your question about when is organic growth going to resume and be positive, that's a function of two things. That is a function of the macro. We're not completely immune to the macro, but it's also how effective we are at selling our services to our customers. I think on that score, we've got a very superior service offering across our modes, and we have leading positions in these fastest-growing parts of transportation logistics, and I'm very confident and optimistic about that.
Okay. Appreciate that answer. I guess what a lot of people are struggling with is reconciling the growth in EBITDA, and obviously you guys are pointing to improved margins as well in the Fourth Quarter. The lack of growth in free cash flow. I guess I want to look at it through the lens of gross CapEx, because it looks like to hit your net CapEx relative to your sale gains so far, you're actually going to be spending maybe north of $600 million this year. Longer-term question, what is the right growth level of CapEx, and what do you think the level of sales to offset that can be going forward consistently?
You're right. All through the year, we've expected gross CapEx to track in excess of $600 million. That is likely to be the case in 2019 as well. That's certainly the number that's embedded in our forecast and that we expect to hit for the year. As you know, our maintenance CapEx is dramatically lower than the gross CapEx that we're expending this year. We go into each year in our budget process considering growth initiatives, considering the environment, considering our various uses of capital and planning CapEx from a bottom up in that way. It's a bit similar to the other questions that we've gotten on 2020. We'll consider the environment, consider how compelling projects are versus other uses of capital, and come up with the number at that point in time.
As we've gone through this year and tightened our belts on costs and increased our discipline on capital allocation, particularly with regards to CapEx, one thing we're very happy about is that we've continued to invest in the tech initiatives that we've discussed through the call. If anything, our focus on those initiatives, our commitment and our investment in them has increased, and we found opportunities away from those strategic efforts to the extent that we need to rein in spending and increase discipline on both capital and SG&A.
Thanks, Matt.
Thank you.
Thank you. Ladies and gentlemen, this concludes our question and answer session. I'll turn the floor back to Mr. Jacobs for any final comments.
Thank you, operator. Thank you, everyone, for spending the hour with us. I would sum up the quarter in two ways. Tough external environment, which shows up in the revenue, but great bottom-line performance by our team, and I'd like to thank and congratulate our operators around the world for delivering a good bottom-line performance. Thank you very much.
Thank you. This concludes today's teleconference. You may disconnect your lines at this time. Thank you for your participation.