Who's the Vice President of Product and Growth, Brock Abednejad, who's running the Escrow division, and Habib Ullah, who is running the Loadshift division. You may address any question in the Q&A to any of us in the room at the end of the financial results presentation. In the first half, Freelancer Limited delivered a gross payment volume of AUD 576.9 million, which is down 1.6%. Freelancer GMV was up 0.8% to AUD 65.2 million. Escrow GPV was AUD 489.4 million, down 13.2%. Group net revenue was 27.1%, down 7.3% on PCP, with Freelancer revenue of AUD 22.3 million, down 4.8%.
Escrow revenue is AUD 4.8 million, down 17.2% on PCP. The Escrow division was profitable for the first half, and the core marketplace and Loadshift now effectively at a break-even position. They're in operating profitability and breaking break-even overall. We had a significant turnaround in profitability with NPAT at negative AUD 300K versus AUD 3.1 million in the PCP, and FX tailwind of 6.5%. The group had positive net operating cash flow of AUD 1.3 million for the first half, ended with cash and cash equivalents of AUD 23.1 million, up 1.8% on the first quarter. Moving by segment. The Freelancer revenue was AUD 22.3 million, down 4.8%. Freelancer GMV was AUD 65.2 million, up 0.8% on PCP.
The segment, excluding Loadshift, was operating positive and break even for EBITDA in the quarter. The difference being the unrealized FX translation. This is with a decrease of AUD 1.7 million in marketing costs year-on-year, as the marketing has got quite profitable in the last 12 months. In the second quarter, we added 1.8 million new users and 280,000 new projects in the marketplace. The average project size lifted to AUD 275, up 4.1% on PCP in the quarter. Note, in the calculation of average project size, we're including everything in the Freelancer segment and the Logistics segment, given that the financial metrics all report together. The average project size includes all enterprise customers and Logistics division.
In the enterprise customer segment, there's actually quite a significant lift in the volume of low-value projects from the large computer and printer company that we have that we've talked about for a few quarters. That is counterbalanced to some extent by the Loadshift division, where the average project size is closer to AUD 4,000. We've put it all in together because the financial metrics are all being reported together in terms of revenue and GMV in this segment at the moment. On the contest side, the average entries per contest was very strong, and liquidity in the marketplace is extremely strong. I would challenge anyone to find a marketplace that was stronger for labor in terms of liquidity anywhere.
There's about 300 entries per contest, and contests go anywhere from $10 up to $10 million. For $10, you can get a logo done or some simple graphic design done, or photoshopping or what have you, as an example. We could do it in 2,700 other categories we have. At the high end, the largest contest we have active right now is for $6 million U.S. dollars, roughly AUD 10 million Australian dollars, which is for NASA and the National Institutes of Health in gene editing in the central nervous system of humans.
It's a pretty exciting contest. It shows that the high end and the sophistication of the work we can do, compared to any competitive peers, is pretty unrivaled, as well as the quality of the freelancers entering, which in this case will be very high-end research institutes and scientists and so forth. The focus for 2023 for Freelancer, we've talked about this several times before in the quarter leads and so forth, is four major points. The first is really personalization to drive the core marketplace conversion. What we're doing here with personalization is really tailoring the experience, taking advantage of some of the advances in things like AI, which can really look at user-generated content and really provide a highly tailored experience. It's pretty amazing.
Our AB testing, in various parts of the funnel, all the way from traffic to posted projects, to awarded, to ultimately paid and completed, are showing, you know, quite dramatic lifts actually, when we apply very high levels of personalization. That's very encouraging. As a result of this, actually, we've bucked seasonality quite well over the last couple of months. We're currently heading into the northern hemisphere summer. Usually, that's a bit of a downtrend in the metrics, and we've bucked that with the improvements we've made with personalization. That's doing quite well. That's really driving sort of conversion of the funnel. We've also made improvements in the way we do recommendations to clients.
We can do that in a very highly personalized way. We've got a lot more coming here. We can actually go quite a long way with how we do personalizations recommendations, given the fact that a lot of the interactivity of the website is text-based. Following that, sort of chat and video, and these are all areas in which you can apply things like AI very well and very strongly. We also have a very large showing in some of the open corpuses for training of the AIs, given the fact that we have a lot of data on our website and we have more users than any of the competitors. We also optimized how we provide notifications and so forth.
This has led to a statistically significant increase in GPV and GMV, and we intend on driving that further. We're also, you know, really on a roll at the moment in delivering product features, you know, for many quarters. I apologize. I'm sorry to interrupt. Yep. You're not sharing your screen. Thank you, Amit. Sorry about that. I'll mention. Where are we? I'm just trying to find the screen sharing here. Alex, do you know where it is? You use Hangouts. It's at the bottom. Can you stop video security in this chat? It'll be a green button. There we go. Apologies for this. Sure. There we go. Okay, is this coming through now? Yep. Okay. We're really deploying a lot of product quite quickly.
We've, you know, for many quarters, we talked about, you know, improving the front-end infrastructure or the back-end infrastructure and so forth, and we've got through that now. We're really powering through the, through the release of features. The project clarification board is a way in which you can post a project, then Francis can ask clarifying questions, which improves the conversion of projects. We also have got a feature we pushed a little while ago with quotations that we continue to iterate on, which allows freelancers to issue quotations naturally like they would in the real world, where someone can ask them for some work, and they can send them a quotation. This now supports fixed price and hourly projects. Shortly, it will also support many other ways in which you can pay for work.
We've had a lot of success with the collaborative features. We talked about this last quarter. We've got a number of things which allow sort of interactivity between the clients and the freelancers. One is with groups, which is simply very similar to sort of what you expect with sort of a Facebook group style interface. This is growing very, very strongly. It grew at about an average of 6% month-on-month through the first half of 2023, and will continue to grow. Our audio and video calling is growing very strongly. It grew about 300% in terms of call count in the first half. All this interactivity is designed to increase the average project size. At AUD 275, that is still a relatively small amount of work.
If you are in the Western world, and you hire one, full-time staff member in the U.S., U.K., Canada, Australia, and so forth, you're probably paying $40,000-$45,000 minimum wage, no matter where you are in the world, for an annual wage. $275 is a relatively small amount of work compared to what businesses will pay, to employ even one person. Even if you're a freelancer in India, and you're a software developer, you probably earn $1,500-$2,000 a month. You know, really, this is the lever I think we can pull that to really drive the growth of the business.
You know, if you've gotten down funnel conversion, and we've had some pretty tremendous lifts, actually, with the personalization results where you've seen some double-digit increases in up funnel metrics for conversion. You know, slowly filtering down to the down funnel metrics. That's great, you can get a big lift, I think, pulling the average project size and adding a zero to the end of it. You can see there's the lift in the sort of call connected count, and that will, I believe, continue as we continue to really hone and push out the, you know, collaborative features like this. We also have improved our acquisition funnel.
I don't know if you remember, the second half of last year, we pushed out a new predictive long-term value model for paid acquisition of customers through the paid search engine marketing channels and so forth, and that requires a bit of time to train and calibrate and so forth. That's fully live and active now, we've actually delivered now five-year highs in terms of search engine marketing profitability, which is up 81% on PCP. This is all in the face of, you know, cutting back about AUD 1.7 million of spend on the marketing side as well. The paid acquisition channels are doing pretty well.
We also have got some personalization that we've deployed in all the funnels for paid acquisition, that's showing some strong uplifts. In particular, a series of AB tests we had showed a 20% increase in new paying clients, and also, one test showed, you know, post-project conversion, about 14% on mobile apps. Another showed new paying clients and five eyes up about 16.5%. We continue to deploy that across the website and hone that and improve that. That's good.
I think we have got a very clear pathway, at least for the rest of the year and probably into next year, in terms of what we can do with personalization and what we can do in terms of funnel optimization, and in terms of what we can do with driving the average project size. The second major thing that we're focusing on product is really getting the design from consistent to delight. Over the course of the last year, you know, we had to deploy a new front-end interface that really basically involved rebuilding the entire website while a public company, which is obviously a very challenging thing to do. That interface is now fairly consistent. We've got our own design system called Bits, and now we're taking that from consistency to delight.
Where I want to get to at the end of the year is to be able to hold up the app and hold up the website and be on par with sort of the, you know, the world-leading examples of great UX and great design. You know, examples like, you know, GitHub or Discord or the like, and that's really what the goal is of design and UX to get to by the end of the year. We're powering through that, and I think anyone that uses the website will see the changes. We've got a lot of positive feedback from the freelancers on this.
We actually had a whole bunch of quotes from the Freelancers in this report. The ASX, for some reason, doesn't like us to pull it out. Anyone who uses the website will see that the interface is improving in sort of leaps and bounds. The enterprise division in the first half, the revenue was up 37% on PCP. The second quarter, the growth was a bit slower than past quarters. That almost entirely was attributed to timing with sort of NASA payments. If you actually had the NASA payments we had in the first quarter, and you added them to the second quarter, you would be on par with Q2 with Q1.
We did have a couple other things that kind of contributed to the second quarter being a little bit soft. One was that we had a very large enterprise client that we really dedicated a lot of work and energy into. We've talked about in previous quarters, this is one of the it's a trillion-dollar tech company that we were working for, you know, very hard for a very long time to activate for a very large engagement. Now, what actually happened was they had a couple other vendors in the program, and they shut the whole program down the day after we passed vendor onboarding. They have told us that there is a actual engagement for us, and post summer here in the Northern Hemisphere.
We've got a kickoff meeting to kind of figure out what that engagement might exactly be in August. I'm pretty hopeful and pretty positive we'll get something. They told us that uniquely, we have the broadest offering, and we actually have the best pricing by far. We'll see what it is, but it, you know, big companies sometimes make some decisions to cut whole divisions, and that was before that we actually generated AUD 1 of GMV, and three other vendors have been totally cut from that program.
We do put a bit of effort into that, but I think that will turn around because I think it's quite promising that the number of meetings we've had since then, and it also pays a little bit of money for something else. I think, I think we'll kinda get there on that. The other two things are, one is that the, in India, we've got quite a number of BPOs. In fact, the who's who of BPOs and back office sort of style organizations, in the, you know, we're signing MSAs that we're supplying labor to.
Somewhat a victim of our own success, that volume has been ramping quite well, we had to change the operating model because when we first went into India, what we did was we were running the payments directly out of India to our Australian entity. As that became bigger and bigger and along with computer and printer company, that's also ramping the volume quite significantly. We, you know, basically we're accruing quite a large withholding tax issue because of that model. What we've done now is we've fully committed to India. We have an incorporated entity in India, and we've got a new model where we get paid directly in rupees with a pure agent model.
we had to have a little bit of disruption going through client by client, or enterprise customer by customer, moving across to that new model. We haven't fully completed that yet, but there was a bit of disruption, obviously, getting them across to a new model. Ultimately, I think that our Indian clients will be much more happy to be paying in rupees than, you know, paying FX and paying an Australian entity. There's a number of other advantages, because there's quite a complex tax situation in India, with tax collected at source, and equalization levy on top of that. The good news is we've got, I think it's significantly better for India. It's been paid in rupees.
It's much better for both the enterprise customer and ourselves, but there was a little disruption in the quarter, kind of moving everyone across, and we've got to kind of get them all fully, all the accounts fully there. The other thing is that we did have a change in the leadership of the enterprise division. Sean McMahon unfortunately left us, and we've already well into a process with a number of candidates. We, in fact, finished our final presentation and the final four in the last couple of days, and we aim to make an offer in the next 24, 48 hours to a candidate. I think we'll get a very solid step up a candidate there.
There's a few things, but, you know, some highlights in terms of the quarter, and where basically we executed MSA with a Fortune One Thousand IT consulting firm with 50,000 employees. Kicked off the first project, which is a fair large one. We continued our relationship, grow our relationship with a consulting giant with over 200,000 employees globally. They've been running several location-specific research projects for hundreds of freelancers on the platform on behalf of Fortune 50. We got 123% quarterly growth from a Fortune 20 technology client and strong forward bookings. We're running several large-scale projects to expand their central workforce program. We finalized an engagement model with a statewide government organization in APAC. This is actually quite interesting.
We're actually bringing them together with another enterprise client to do something pretty interesting. This is a very, very well-funded government organization that's looking to provide employment in their particular area to about 6 million people who are unemployed. We obviously won't be able to do a full 6 million, but the point is, they're very well resourced, and they've actually done a fair bit of work on the FTL platform, which is quite interesting, and we're going to dovetail into that. We also added to our government team with a new hire to lead the growth of our NASA Noise engagements. Trisha is quite a solid performer. She's got a master's from Caltech in geophysics. That's a great complement in terms of the science base for NASA.
We, in fact, have won three task orders in the last week for NASA, so some pretty interesting stuff up and coming, including. One of them is actually for the detection of micro debris in orbit. This is 1 millimeter - 10 millimeter particles in orbit. You know, figure out a way to detect and remediate that is a pretty interesting challenge, the space program, and we're pretty excited to be supporting that. We've got another one we've won, which is improving the GCC compiler, which is being used by the Orion space mission. There's another one on comfort in air taxis. Visual comfort in air taxis. When you're flying around in an air taxi in the future, how can you make sure that people are kind of, stomachs don't churn and so forth?
They're pretty interesting projects, you know, and task orders continue to come out and ramp. Obviously, the program's expanded from AUD 25 million - AUD 175 million. Let's talk about expanding it further. We'll see where that goes, ultimately, but the engagement is still going very, very strong and continues to get bigger and better. We're seeing the size of these task orders grow quite significantly. You know, obviously, with the NIH one that's live on the website for June, I just think that's, you know, one of the biggest ones to date, which is $6.7 million U.S. dollars. We also added to our enterprise sales team with us, using a hire with nine years of sales and marketing experience. Deloitte is now at 38,000 consultants in the platform.
We're powering through a whole bunch of product improvements that they wanna put in the platform. You know, My Gigs is, you know, is quite a revolutionary, I think, quite well leading platform for augmenting a physical workforce with a cloud workforce. You can post a project as a Deloitte consultant and have it go to 48,000 other Deloitte consultants or go to the, to the guys of the world. A couple of quarters ago, they doubled the engineering services component for us on that, and that will continue through the rest of the year. That's quite strong. We're really just heads down building product for them. The big focus is on the internal side of the My Gigs marketplace rather than the external.
You know, obviously, you've got to go slowly in the consulting world, and make sure you kind of, everything's kind of, you know, well-ordered and well-managed. You've got a lot of, you know, compliance and risk and other participants making sure that everything is done, you know, in a great, you know, well-ordered fashion. That just means things go slowly. On the other hand, we've built a really robust enterprise-grade product offering we can take to others. One advantage, however, of building the internal marketplace is that 60% of the internal projects are eligible for external, even if they're not all being allowed by the compliance team to go external. As you build that internal marketplace, it's also building potential external volume.
We're still, you know, plugging away at that and at scale, that's growing, that engagement. There's a lot of participants on the Deloitte side, we're working on that, it's powering along. That's going well. Our global fleet is going very well as well. We're now operative in five countries. We're in 28 cities. 25 of them, we've integrated directly into this customer's back-end support system. We've done 22,000 jobs now. This is 22,000 jobs for repairing computers and printers in five countries. That's gonna go pretty quickly to that sort of order of magnitude per month now because we're in the ramp-up phase. We've got some pretty interesting results. We've got a lot of parts usage per repair than the partners.
I don't know if that's the case. It, you know, we're just not sure, but that's an interesting statistic. We've got a service level agreement on par with the other partners, and we've got customer satisfaction for up to 22,000 jobs, which is a pretty solid achievement for our global fleet offering. Basically, what's happening here is, if you break your computer or in these particular regions, there's a chance that one of our friends is out there coming and repairing them. We are also going live and rolling out in other cities in these five regions. A couple of the cities in India. In Australia, we're going to Warrnambool and Alice Springs and so forth.
You can see that we're quite good, not just in city metro, but also regionally, actually thrive where it's very hard to get service coverage from either full-time staff or traditional providers. We're also being told, the bottom is going to ramp pretty significantly in Malaysia, and we're going to Singapore, and we're about to turn on overflow operations in the U.S., which is really the main game. This is really, you know, this is really where we want to go ultimately, is where the big volume and the high-value projects are. A couple months ago, we had a big kickoff meeting. I was on that call, and it was quite interesting going through all the various regions of the U.S. and so forth.
You know, where we will start is the same as in all the other countries. We'll be in, you know, Wichita and Arkansas and all these weird places, and we'll start with only overflow, but as we kind of prove ourselves, we'll get more and more and closer to the metros and so on. The last regions we'll probably get to ultimately will probably be New York and San Francisco and L.A. and so forth, and we'll have to prove ourselves just like we've proved everywhere else. We've done so quite well so far, and 22,000 jobs have been done to date, so that's pretty significant. In terms of mass, just completed a $1 million task order with the National Institute of Standards and Technology.
That's for building up the next generation heads-up display for first responders going into emergency situations. You know, this is, for example, going to an earthquake or a bombing, and the AR/VR system allows you to tap into the CCTV streaming count, vital signs of the first responders and so forth. And that was very successful. The four finalists, I believe, are all have commercial product or about to have it produce a commercial product. There's commercial outcomes from these innovation contests. Not only do the organizations that launch them get to solve very innovative, you know, moonshot sort of problems at the high end, there's also commercial outcomes that come from that, as well as a talent pool they can tap into to do further work.
There's a lot of advantages, and contests generally give you a 40x bang for buck in terms of spend to results, and that's out of Peter Diamandis' XPRIZE Foundation Book, Bold. We also I've talked about this $10 million task order engineering thing. I won't go again. We also completed an innovation contest for the U.S. Bureau of Reclamation in modeling of sedimentation in rivers. This required a speeding up a optimization, like a solver. We actually got very dramatic improvements in that. I can't remember what the original target was they wanted to get in speed up.
I think it was the solver can take up to a month to run, and I believe they were looking for a speed up around 30%-40% from memory, and we ended up getting it 80 x faster. They were pretty blown away with the results for that. That's quite a good achievement. We also built a next-generation precipitation measurement device. This was for water measurement in extreme environments that could be extreme, so from -40 to +50, you know, UV exposure, high wind loads, extreme precipitation events, and so forth. Probably something that could be quite useful for Australia as a product if it was productized. We also won a few other small projects.
As I mentioned in the last week, we won three either by ourselves or in partnership. In terms of escrow, the GPV in the second quarter, now this is quite interesting. It was AUD 220.6 million, down 21.7%, or AUD 147 million, down 27% on PCP. You can see the long-term uptrend. I mean, there's, you know, we're still in the uptrend, and in fact, you know, this is a top 10 result up until the last four weeks of the second quarter, we were actually very strong. On a rolling three-month basis, up until the end of May, the GP was actually at AUD 193 million, which is actually above the first quarter numbers.
It was really just in the last four weeks that it came off, and again, the volatility that you see in these numbers here comes from domain name transactions, in particular, mega transactions. There's about AUD 45 million of mega transactions that kind of rolled off in those last four weeks that weren't replaced by new transactions. I did actually expect that to continue. We had just come out of the NamesCon conference, which is the big domain conference of the year, which ends the June 3rd. After what we're told is that after the June 3rd, that, and the conference and the traveling, that a lot of the participants who are involved in these big transactions to leave rather than going to business. I was a bit surprised by that.
That was really just in the last four weeks. I do expect this to bounce back quite strongly, for the rest of the year. There is a bit of a boom happening in venture funding in certain segments such as AI, and I do think we'll see a bunch of mega transactions go through in the second half of this year. We already sold chat.com and prompt.com for quite significant numbers. I do expect a lot more of that to continue, and I do think that the funding for venture backed startups is shrinking a little bit after that, sort of, that valley of death in the second half of last year.
I was a little bit taken aback by that last four weeks, but I do think that's going to come back in the second half, and so on. You can see that, you know, if you, if you kind of go into July, you know, this is, again, very short-term data here. This is only a couple of weeks into July, but you can see the counts gonna bounce up a bit in July. With the confidence that this is gonna come back, and we actually will finish up the year actually pretty strong, and we'll continue that trend. In terms of product, we did do a major overhaul in the second quarter.
We've really ripped out the products, and we're a lot more modern. It was quite, it was quite dated. That was quite a bit of technical debt we paid down, and that will continue into the third quarter a little bit, but we'll finish that off. A quarter of this year, one, customer feedback about improvements. Obviously, customers always come first. Two is to improve the KYC to be best in class. The next is improve the friction of the transaction flow, is to get super slick. Really provide a great experience. You know, 95% of the volume that goes through Escrow now results in eBay Motors and eBay watches, and you have quite a common, and so on.
The real long-term plan with Escrow is to get that checkout experience super slick, and then, you know, kind of do. You know, try and do what Afterpay did, which is that great sales team going up to merchants and platforms. The second is having a partner activation team that when you do these platforms, you activate them really strongly, you really build that ecosystem and you get the volume. Other than the Bits around partners, major shopping cart. These shopping carts, we're really dedicating a lot of resources in the third quarter on that. I'll talk about that later, but it's early days, but we're one of the alternate payment methods. Shipping segment, this is going extremely well.
In the 2nd quarter, we saw significant upwards in the 1st quarter. We're not really reporting year-over-year because the merger of the bulletin board model and the marketplace really only happened in August of last year. We'll soon be able to talk about year-over-year numbers that make sense. Otherwise, you're kind of mostly talking about different business models. The GMV was up 53.7% quarter. We had an all-time record, the most quotes per day, which is up 52%-53% quarter-on-quarter. We had an all-time record on things that are up 35% on the quarter. Commercial drives up 72% in the quarter, using completed loads up 39%. Average completed load size is about AUD 3,600, flat average freight charge 8% up on the quarter. That's something that's from PCP on.
We've also got the break here. It's positive in May and slightly end of June. Basically, you know, all three businesses now did that, which is great, and I think we've got a big ability of the delivery, and each business is on its way, so the goal is to keep that going. In the second half, I think we'll be able to start publishing year-over-year because we'll lap the August. In terms of total loads posted, I mean, I do put these numbers out there, but they don't really. A bit of apples and oranges here.
When we compare the bulletin board model before, where it was free to post a project, and for AUD 79, the drivers did get the phone numbers, compared to now, we've got to put the money through the site, pay for the site, et cetera, and so on. It's removed a lot of the fluff and the other activity that's happening on the site. The loads posted in the second quarter are about 13,000, and the commerce were about almost AUD 18 million. These numbers are down a fair bit from the actual bulletin board model, they're not realistic because in the bulletin board model, the loads came off within 72 hours automatically. In the marketplace model, that's about 30 days. There's a lot of reposting.
It's, you know, in the original numbers, but I'm still reporting them for consistency's sake. It'll make more sense when we lap year-on-year. You'll be able to see a real understanding of kind of what's going on here. There's about 300 million of notional load volume from being posted onto the site, and the goal really is now to basically convert as much of that as possible to GMV and revenue. The composition of the freight machinery stays strong. It's about 29% of the week. General full load, full loads are up 6.5%. You can see here, there's lots of stuff we do. Freight that moves from locations to Australian location, anywhere in the country, Australia only.
The goal is to basically convert as much of that freight as possible. Currently, that's come from the merger of the two into the pay for the site and the marketplace model. After that, we have a couple of avenues, quite strongly in Australia, right? After that, go internationally. Product-wise, we did a few things, improving the flow and so forth. We, through Butn, which is, I think, a listed company, got payment terms, which is something that covers me in order to manage the cash flow. We've got better ways to run the trucks on the website, fuse and so forth. You can see a quite strong conversion to this model.
In the old days, one year ago, the funds just got handed out, and it was the Wild West, and no one quoted at all through the websites, so the quotes are zero. We've done quite a decent amount of work on getting that model going from one where it was literally, "Here's the money, and off you go," to going through a website, quoting through the platform, or paying through the platform, you know, putting reviews through the platform and so on. I can successfully say that the model has worked. It's well and truly away. It's ramping very strongly, and there's a huge number of benefits now being part of the marketplace model. Shippers now can look at reviews of truck drivers. In the past, they couldn't do that because the reviews were not weighted by dollars, right?
If you're not taking the payment, you don't know if the review is real or not. That's happening now. There's a lot of trust and safety that's been built in the marketplace. We've got rid of a lot of the bad actors that happened under the old bulletin board Wild West model. These are drivers that would take a deposit and then not show up. Also, if you have poor performance, you get a bad review, and so on. This is growing quite strongly. There's a lot of growth to come in terms of the revenue, the GPV here, and you can see the award rate is climbing very, very strongly. The question is really, how far can we push this?
I believe that we should be able to get probably about 60% for Loadshift. If you're getting a quotation on an excavator, moving from one side of the country to the other, it's a little bit different from, you know, do I wanna be an entrepreneur and get a website built and so forth. I think the award rate here will be significantly higher than on general freelance marketplaces. We've got a segmentation here in terms of the freight, that's very consistent with previous quarters. Overall, the operating costs are about 9% lower than peak and cost efficiencies across all expense categories. Second quarter were pretty flat on the first quarter.
We're effectively breaking even the half on operating EBITDA basis, and we've reached the cost base, and we're, you know, I think, in a good place for the whole group now. Operating cash flow of AUD 1.3 million for the first half, and cash and cash equivalents 1.8% on the first quarter. What I might do now is I'll open to Q&A. You may now ask a question to myself or a question to any of the executives in the room. I might remind you, I've got Neil Katz, the Chief Financial Officer, in the room. I've got Adam Byrnes, the VP of Product and Growth. I've got Barak, who's running Escrow, and Habib on the Loadshift side. You can address questions to myself or anyone else in the room.
I'll open up for the Q&A, just to put your question in the chat and we'll address it. Alex, I might get you to read them out because I can't actually see with this view, with where I'm sharing my screen, what the chats are. If a question comes into the chat, please read it out and I'll address it. Any questions coming in?
There we are. One just came through. Okay. There's a question from Emma Hayes: In escrow, where do you see the most potential growth outside of domain names?
Well, there's one... Yes, and you see, the answer is that, yes, I kind of think I alluded to the account I'm most excited about actually in, at the very end of the commentary.
We have one, a shopping cart, where one of the payment methods going into visible in that shopping cart. There are quite a number of payment methods in that shopping cart, but we are a unique differentiator, and this has got a lot of volume. It's many billions of dollars per annum. We believe we offer a very differentiated payment system. For payments from $0 to, say, $10,000, they are very well served by existing payment methods. For example, you've got credit cards, and you've got your PayPals of the world, you've got your Afterpays, you've got Zelle, and so on.
The issue is that all those payment systems that exist, all the digital payment systems you hear about every day, they're all built up on top of debit and credit cards. For example, Apple Pay, you put your debit or credit card to your phone and, you know, and that wallet is just built on that. PayPal's the same. You put a debit or credit card in. It's built on top of that. Afterpay is effectively built on top of debit or credit card, and so on the like. All those payment methods that are, those electronic wallets, the entire volume of digital payments from those wallets that are on, that are built on cards, are only 3% of U.S. domestic payment value. Not by count, but by value, because the they're all low-value payments.
Apple Pay, the average payment size is $23, PayPal is about $60, Mastercard and Visa is about $64 and so forth, right? And then you take all the card volume in the U.S., and you look at that by value, in domestic payments, that's only 8% of domestic card volume, and domestic payment volume, sorry, in the U.S. by value. All those digital wallets are 3% of 8% of U.S. payment volume. The reason why is because large value payments are not served by any of those payment methods. We fairly, uniquely, and I say uniquely because there's a very intensive regulatory environment around large value payments or escrow payments.
We have a complete licensing footprint in the U.S. except the territories. There's 50 states in the U.S., four don't require licensing, and there's six territories. We're licensed across all the U.S. states to provide escrow, and escrow allows you to take a payment from, say, $1,000 to $100 million or potentially more. By going into things like shopping carts and payment aggregators, we think we can unlock transactions that otherwise couldn't occur. For example, you go into a shopping cart at the moment, they can very much sell, you know, things below $10,000 very well with a whole range of different payment methods. You might get, you know, 30 or 40 different payment methods you can select from.
If you want to sell something like a yacht or you want to sell fine art, or jewelry, or song royalties, or yeah, businesses, or, you know, high-end websites or whatever it may be, you can't do that because above, you know, AUD 10,000, these payment systems will break down. I think we add a unique differentiator to shopping carts, and I think we add a unique differentiator to the payment aggregators. I think that's where you know, we I think once we go into this particular shopping cart, I think we've got to tip a lot of the industry, and I think we've got to tip a lot of the payment aggregators because we offer a unique differentiator to a really tier one mucky part.
I think the volumes we could see there could be quite significant. I mean, when I say significant, I mean, you know, it could be a multiple of our entire volume for the company. You know, I'm running that number off, just averaging the number of the total volume that they put through their alternative payment methods just looking at that, but not taking into account the fact that we do the high-end, high-value payment payments, where you'd expect to have a larger share than average. I think that those sort of businesses will have quite a large materiality in terms of the effect, from going into them.
I also think we can get a fairly large volume from the automotive space. As you know, we're in eBay Motors. We power eBay Motors. That's a relatively smaller player, but we have demonstrated that we can do those payments very, very well. I think that there's a whole bunch of automotive and well, ancillary marketplaces that we're in various levels of engagement with, or in some circumstances, you know, they're in development. In some circumstances, they've just signed an agreement, and they kind of, you know, haven't started development or are prospecting. There is a challenge in automotive in that you have to have. To unlock a lot of volume, you've got to provide the financing part.
We built a financing marketplace with Blinker for Autotrader. Unfortunately, that particular engagement was not successful, we are working with Blinker to go into other marketplaces. I think, you know, we'll ultimately crack automotive because I believe the time is right, that people do want to pay for a car over the internet. We've shown that with eBay Motors. I think we'll get into the other automotive marketplaces, we have to just keep plucking our way to... I think we've got to have a strong financing solution, which I think we're kind of, you know, at least mostly there with Blinker. I also think the M&A as well will be a fairly material segment.
We're in Acquire.com, used to be called MicroAcquire, so you can buy businesses and startups from a bunch of other sort of similar competitors, but there's a lot of volume in that particular space. I think that will be fairly material. In fact, in the slide deck, which is attached to the releases that went out today on the ASX, there are quite a number of segments we've identified that I think, you know, we'll be fairly successful in. Simply because at the moment, they're not well served at all by payments, and in fact, the payments are very much offline, or they're done through asset purchase agreements, or they're done through wire transfers to effectively an intermediary in a very clunky way.
I think we can-- I think basically those segments are going to be quite strong. In summary, shopping carts, payment aggregators, automotive and M&A are really, really my focus sectors. Ultimately, we want to get to real estate. Oh, we won't get to real estate directly. We can do half the states, done a very extensive work on what we can do and what we can't do at this point in time, given the state-based regulation around title transfer and so on. I do think things like construction and taking deposits to purchase a house, I think will... That will be our initial inroads there. How we build operating leverage? I mean, simply keep the costs down.
I mean, these businesses are all marketplace models. There's a fairly high gross margins, about 84% across the businesses, plus or minus a couple percent across all the businesses. You keep costs down in terms of the expense side, and then you get operating leverage as you grow the revenue, right? I believe that we'll get a lot of leverage out of the improvements we're making in the core marketplace. I think we'll get it out of, you know, any one of these enterprise customers on the Freelancer side really could really take off, and we're starting to see signs of that in a couple of the accounts.
On the Escrow side, again, it's just, you know, one customer could do significant amount of volume. On the Loadshift side, we've got just eating into that, the Bull and Boar we acquired, that's going extremely well. We're demonstrating that with the results from Loadshift.
Ray asked a question: Can you comment on the implications of recent AI developments?
I've actually written... If you want, Ray, if you want a full detailed thing about this, I'm going to be publishing an essay on this in about a week to two weeks. I've written quite an extensive essay on this particular topic. I've actually been interviewed by MacroVoices, which is kind of a prominent macroeconomics podcast in the world.
They're going to do a two-part special in August, late August, for the summer holidays on this. Effectively, in the short term, AI is turbocharging the freelancers because now you can be moderately skilled or relatively unskilled as a freelancer, and you're now using these tools, whether it's ChatGPT or Midjourney for design, you can now design at the elite level. I believe, and we're seeing it now, the adoption of the AI tools by the freelancers is pretty quick. Like, as soon as the tools came out, the freelancers were on them. I think we've done some done a media a couple of media hits around a couple of contests we ran.
One was a Harry Potter, you know, reimagine Harry Potter in a different scene. If you just click through the entries in that particular contest, you'll see that the freelancers all over the AI tools, all the entries were incorporating AI and fairly advanced uses of that. I think what it does is it lifts the skilled labor in our marketplace. You know, the winners here, I think, with AI, in particular in the short term, are, you know, the relatively new workers, the relatively unskilled workers who are now, their skills are lifted very, very significantly in terms of their abilities and talent.
The elite freelancers at the very top, they're fine because they'll figure out how to incorporate into their business model, and, you know, they've got relationships and they've got a track record. I think the Western-style traditional service provider will be challenged because, for example, if you're an illustrator and you're in the Western world, and you're charging 20 - 40 hours worth of work to design something, I think that the appetite in the future from companies to want to pay for 20- 40 hours of work with something you can tap out in 10 seconds from Midjourney is going to be challenged. I think that, you know, I did an article, actually, it's kind of came out last night in India.
I did a article, an interview with The Indian Express, where I think I said India is going to be a big winner of AI because they've got a once-in-a-lifetime net influx of about 14 million people into the workforce. Those graduates are relatively unskilled, and now using AI tools, and whether it's, you know, in copywriting or design and shortly to be software development, they're gonna be now superpowered, which is great. You know, I think it's fantastic for us. I think it gives the skilled labor. It lets the freelancers deliver at a lower unit cost and quicker. It lets businesses get things done cheaper and to a higher level of quality. I think for some time, you, I mean, you're going to need freelancers driving the tools.
I mean, you know, my mother is not gonna be able to interface with, you know, any of the AI tools anytime soon to completely grow her business. She might be able to get little bits and pieces done, but you're gonna need someone running the tooling for now. Where it's gonna end up in five, 10 years, I don't know, right? I mean, things are advancing pretty quickly. But at least for now and for the foreseeable future, it's I think it's fantastic for us, having the largest low-cost workforce in the world. They're all now superpowered.
Being that Freelancer platform operates globally, this from Brad, what is required to grow revenue? Is it enterprise SMB, consumer, startups, customers?
I think it's all of the above, right? We're executing in all those segments. You know, the... Right now, the biggest contributor of revenue by value is the core marketplace. It's really moving the needle in the consumer core marketplace. You can do that in really three ways. One is you can acquire more customers, either organically or paid. Paid is obviously quite intensive to do so. The second is you can convert the funnel better, so as traffic comes into projects to awarded, to accepted, to completed, to paid in full, you can convert that better. The third way is you can retain your customers or grow your average project size, or grow the amount of work that they do on your platform. I think that's the strongest.
We're attacking all parts of that, with the exception of we're not growing the paid marketing. We're just making it more efficient. On the consumer side, we're attacking basically, you know, the challenge of how do you bring in more customers organically, whether it's by referral because someone's had a happy experience. Whether it's built into the platform, 'cause you wanna work on a project with a coworker, or a colleague, or your mom, or your friend, or your, you know, your startup, you know, co-founders. It's chipping away at that funnel in terms of conversion. We're getting a lot of improvements with AI, as I said before, and personalization.
Whether it's retention, which is a combination of things like collaborative tools to really make it easy for people to work together in teams online and really have it part, you know, for us to be an adjunct part of your business. On the enterprise side, you know, the way I characterize this is, you know, this is something new. There's no solution out there for enterprise right now that is at scale anywhere in the world for consuming talent. Universally, however, the chief, you know, HR officers and procurement heads, and so forth, and chief digital officers of all these enterprises, they have universally decided, and this is independent of anything we've done, that at some point in the future, some percent of the workforce will come from the cloud.
Whether it's 5%, 10%, 15%, 25%. Whether it's one year, three years, five years, 10 years away, they've all universally decided that, and COVID has just accelerated their plans. Now, we, I think, built a solution with what... We've, we're a number of other companies looking at that solution and adopting it. This is quite... You know, it does require some integration. We have to integrate with their Venmo, so that the payments flow nicely. You've got to integrate with their single sign-on system so that, you know, the users at that large enterprise can log in.
You've got to have the look and feel of the large enterprise, because you can't just have the CEO of Coca-Cola say, "Go and use Freelancer and have it work." It's got to look like Coca-Cola, jobs look like Coca-Cola. The branding and, you know, log into the same password and integrate into the systems. There's other things that you might need to integrate with, such as their, you know, time tracking systems and directories to slope in the profiles of the users and so on. There is some level of customization. As a result of that, it is, you know, it is a longer sales cycle than the consumer side, but I think the payoff is huge.
The latent demand in some of these large enterprises, you can unlock, you know, 10% or 5% even of the labor force budget of a Fortune 500 and get that really working at scale. It's very significant. We're plugging away at it, and I think, you know, we've really kind of shown a pretty world-leading product, and we're taking that to everyone else. You know, we're doing, you know, very similar to this computer and printer company. You know, it's not exactly the same solution, but it has many of the characteristics of that solution for solving a certain point problem. We just want to get one thing done, but we need it scaled down millions of times all around the world.
You know, we've shown that quite successfully. 22,000 projects being completed through that, and that's gonna ramp to about that number per month very quickly. You know, what we've got there is quite a number of moonshots and a lot of irons in the fire. We've probably signed, I think, 60 or 70 MSAs at enterprise, and we're really trying to figure out how do we effectively service and activate those customers. If you leave them alone, like some of our customers that sign MSAs, they kind of, you know, we help them with the managed service team, but for the most part, they're not integrated into our systems from a technology perspective.
You know, they grow strongly, but we want that big step leap up, where you kind of deploy across the enterprise globally. You know, trying to build a repeatable solution as possible, while still having that sort of lock-in, that you get from kind of integrating. That's on the Freelancer side. Ray,
"Thanks, Matt. Can you comment on the Amazon Web Services thing, please?
It's not a rebranded company." I mean, we have three businesses in the group, right? We're in labor, payments, freight. All those three services are things that businesses need, consumers need. Effectively, we're trying to build a mini Amazon of services.
You know, I mean, some of the largest companies in the world by market capitalization are global marketplaces of products. You've got your Amazons and your Alibabas and, to a lesser extent, your Ebays and your Etsys and even Shopifys and so forth. These are all product, you know, marketplace platforms in various form factors, and we're doing it in services, right? That's what we're doing. We're small, and we're relatively early stage in the space. No one has cracked the Amazon of services, not. You know, if you add up into the freelancing space, the total number of users on all platforms, you're probably under 100 million total. Across the space, we've got about 67 million on our platform, so but there's 5 billion people.
The whole space is very, and it's slow moving 'cause it's very complex services, you know, over the internet, amongst other things. Do we have any other questions from anyone in the audience? I think we've kind of tapped out the questions that were in the chat. Please, if you want to ask a question, put it in the chat.
Brandon asked on Loadshift, you mentioned, a great 20%-60% increase in the award rate given that we can't clearly see growth as the sale. You said with apples and oranges based on what you've seen so far, Loadshift model works.
Let me explain clear what's happening here. We had a platform which was called Freightlancer, which was basically taking the Freelancer code base for enterprise and customizing it for freight. We customized it to Deloitte for that particular use case, and we took that code base and customized it for freight. We had effectively a marketplace for freight. It had small volume in it. There's a bit of a history here. We acquired a business called Channel 40 some years ago and then merged it into the code base, and we ended up with Freightlancer. If you think about Loadshift, and you think the freight model, and you're thinking about the freelancing model, very similar. I mean, freight is really number of categories out of the whole world of work, right?
You post a job on Freelancer, people bid on the job, you award a job, you set the milestone payment. The platform holds milestone payment while the job happens. You release the milestone payment, you leave reviews and feedback, and off you go. On the freight side, it's the same. You post a load, drivers bid on the load, you through the bid list, you talk to the drivers, you award a driver, you put a payment in, the platform holds the payment until the load goes through. You then release the payment, then you leave feedback reviews. It's exactly the same flow from a business perspective. What we did was we effectively took the Freelancer code base, and we merged in a business years ago, Channel 40, to build Freightlancer.
We had effectively two platforms, Freelancer platform and a Freightlancer platform, operated very similarly. Every time you make an update of the Freelancer code, the freight gets it as a freight, and there's a little bit of customization. There's this whole team doing some customizations for freight. What we did then is Loadshift was an inflatable board for heavy load freight. It's about 8,000 or so GMV a year. On that load board, about AUD 3,600 average load size, and in the past it was the Wild West. Basically, choose a site for free and post their load, then drivers, they pay AUD 99 a month will get the freight. That was the complete involvement that Loadshift had.
If the load moved, they didn't know if it arrived, they didn't know if there was a problem, they didn't know if, you know, if the driver was legitimate or the ship was legitimate. It was just the Wild West. We came in, and we moved that to a marketplace model. We bought that, and it merged it into the Freightlancer code base, and it was the last year complete, and then we rebranded everything. We have at that point is we had about, roughly, notionally, about AUD 300 million of the freight being posted in that platform, which we started converting to a marketplace model.
What I mean by that is, getting the drivers to change from just calling to calling and quoting, to them basically having the shippers award through the site, then having the drivers accept the payments flow through the website. You know, a bunch of features to assist with making sure that all happened really nicely and smoothly, ultimately releasing the money and then leaving feedback for both parties, for both the shipper and the driver on the platform. That was a big change in the space, right? It's a very big change in terms of how things worked. It's ultimately better for the shippers, because the shippers now know which drivers are good or not.
They don't get smacked with 50 phone calls all of a sudden. They can do something very orderly fashion with the whole operations team that is really a managed service team that overlays on top and ensures that the load goes through reliably and can offer a whole bunch of other services, such as arranging pilots or permits or, you know, road, you know, road closures and a range of other things, depending on what's required. From a business perspective as well, because so that the shippers now, if they want to pay for the live company to move them out of Katie at Goldline for Newcrest, right?
I think for, they pay 30 days to day, 120 days, all the drivers hang on, you know, payments is everything. The drivers now can stand out on the platform with reviews. If you're a great driver, you deliver great service, and you wanna charge more for your services, you can do so now because you've got the reviews that make sense and are credible versus maybe a cowboy that's trying to do the work and it isn't properly, you know, it hasn't got proper certifications or this or the other. We vet drivers on the platform to make sure you've got the right certifications, we can actually. Before, in the Loadshift model, anyone could call up.
The thing we've got to do now is we've got to see how much of that freight we can actually convert. What is the theory, the practical maximum, sorry, of freight we can convert? We've been very successful in getting that award number up and awarded, then you've got accepted and funded. Those numbers are growing very strongly. The question is just how far can we push that? The way it happens is just heads down, working at operational support, we've got control. We will see very strong growth over the next 12 months in this business. Very strong growth in financial metrics. Okay, any other questions? Thank you. It's kind of funny.
Two research groups initiated coverage on us today. The ASX has changed the rules now, so you're unable to... ASX, tell your investors that research groups initially cover into this range. I would have thought they want both. Which is one of the biggest independent sort of research firm out there, initiated coverage this morning, as did Pitt Street Research. Reports are available, I think, on our website in the investor section, or you can go just search online, go to Edison's website and download them, or, and Pitt Street Research. They both initiated coverage and will cover us from here on. I think it's something that investors have asked for, some feedback I've had, that, you know, we used to have you getting Canaccord coverage consistently for a long time.
There's no interest for these. We don't do that. We have two firms, Edison and Pitt Street. We actually have interest from a third one as well. It's actually an investment bank to cover us, but that hasn't started yet. I think that will obviously provide value, I think, for investors to get a bit of an overview of the business and ask questions and so on. I'll give it a few more minutes. If you have any questions, I'll have the answer or you address the question in the room. Otherwise, if you want to do a one-on-one with the team, you can email me at Matt, M-A-T-T, @freightlancer.com or investor@freightlancer.com, and one-on-one with myself or any other executives. I'll leave now.
I'll see you in the next quarterly results for that talk, one-on-one, in between.