Welcome to the Freelancer Limited first half of 2024 financial results presentation. My name is Matt Barrie. I'm the Chief Executive and Chairman of Freelancer Limited. Today we have a full court. We have with us Neil Katz, Chief Financial Officer of the group. We have Shaun McMeeken, who's the Head of Sales, Enterprise Sales for the business. We have Andrew Bateman, who's 2IC and Products and in charge of Enterprise Products. Adam Byrnes, VP of Products, August Piao from Escrow.com, and Cohen Wisniewski from Loadshift. As always, you may address the questions at the end of the presentation to any of the executives in the room or to myself. In the first half of 2024, gross marketplace value for the group was AUD 449.8 million, which was down 18.2% on PCP. The Freelancer segment was flat at AUD 64.7 million, down 0.8%.
Escrow, which is fairly lumpy and has been in the past up and down, that was $385 million, which is down 20.5% on PCP. If you do look at the long-term trends for GPV for Escrow, it is trending up, and that trend remains intact. We lack a spike from the first half of 2023. I will also point out that we will have a good quarter this quarter coming up for Escrow. On Friday, we just had the buyer and seller sign for a $50 million transaction. So that's going to go through the next two weeks. So we'll have a big jump up in this quarter for the Escrow GMV. So that's going to be a bit up quarter. So we do have a bit of a lumpy profile on that, which goes up and down quarter-on-quarter.
I'll talk about it more when I get to the Escrow segment. The group net revenue was AUD 25 million, which is down 8% on PCP. Freelancer was AUD 20.4 million, down 9%. A contributor to that was engineering services, which was down a little bit in the first half. Although, as of yesterday, we've got agreement to proceed with an engineering services engagement on the generative AI work. And I think ultimately that's probably going to end up being a low seven-digit engagement over time. But the first phase of that got approved yesterday. And we also have a second engineering services proposal going out this week, which we'll see when we get to more on the field services side. So there is improvement coming in some of those numbers on the engineering services side. And I will talk about on Freelancer a bit more detail.
It's looking pretty good in the last six weeks or so in terms of growth for that business. But I'll get to that in a little bit. The Escrow net revenue is AUD 4.6 million, down 3.2% on PCP. Although if you look on a three-month basis as of today, year-over-year, it's actually up. And in fact, in the second quarter, you were actually up 10.5% on PCP. The profitability of business is basically flat at $1 million loss in the first half. Operating cash flow was up 57% to AUD 2.2 million. That's doing okay. And the group ended with cash and cash equivalents unchanged on FY2023 to the decimal point. Getting into the Freelancer segment for the business, for the GMV was flat. But I do think this is going to actually grow fairly substantially in the second half of this year.
In particular, we've got quite a large amount of work coming through on the generative AI side. I think we've made quite a number of announcements on the platform about that. We're currently sourcing probably well over 100,000 freelancers for work on training for one of the largest foundational AI models in the world. There will be actually an announcement about the partnership probably in about 60 days from now. We got an approval yesterday that we're going to put together a release on this to talk about the partnership. We have to go through a number of big company approvals before we can kind of get the names out there public. We are actively already working on this. We've already sent across many, many thousands of freelancers. We are actively engaged in recruiting of those freelancers.
And there are many, many, many projects involved with not just this particular one customer, but also expanding across the generative AI field. So this is quite a very lucrative segment for us. And I do think that by the end of the year, this work will contribute a seven-figure U.S. GMV contribution per month to the business. So it will be a fairly substantial contribution. And there's quite a lot of upside from there as well. So we're at the moment just actively sourcing freelancers and getting them across. And it's across AI training. It's across safety work. And there's also a lot of specialist generative AI work that we're doing as well, in particular niche areas of human endeavor. The marketplace added 1.74 million users in the second quarter. And in fact, we had quite a big win on the sign-up funnel for Freelancer.
That resulted in, as of, I think it was mid-June, was it, Adam?
A little earlier.
A little earlier?
Yeah.
A 40% uplift year-on-year in verified sign-ups to the platform. We actually got quite a big lift in sign-ups from a bit of a win that we did in the sign-up flow. The average project size was $268, which is effectively flat on PCP, down 2.5%. It was up 5.9% on the first quarter. It's basically hopping around that market at the moment. That represents the average project size across everything we do across the consumer side of the business, small size, small business side, as well as the enterprise side. We put it all in together. I think it gives a good feeling of kind of where we are on the average project size. Marketplace liquidity improved a little bit. Average bid per project is 45.
Average entries per contest is an incredible 340, which I think blows out of the water any other contest platform on the internet by far. Contests actually span all the way from very small little contests where you might want to do something simple like get a logo designed for $10. You have hundreds of entries come in with a very interactive format, right up to the high-end work we're doing. The largest contest we have running is currently AUD 10 million or about $6 million for gene editing and the central nervous system of humans. So the contest platform continues to amaze in terms of what we can do there. I'll talk about that in a second. In 2024, we're really focusing heavily on client retention and engagement.
We've got some great results there. We've had on a sort of 1- to 4-week basis that are trending up over time. We had quite a number of wins in the core marketplace in the last quarter across payments, across the user interface, across collaborative features in the product, and across design and so forth. We're chipping away at that. And I'll talk about in a second kind of what that means in terms of revenue and what that means in terms of GMV. The core product focus for 2024 is really four-fold on the core marketplace side. The first is to turn Freelancer from painkiller into a narcotic. I mean, people talk about you want to build a painkiller, not sell a vitamin. We want to take it one step further and make it something that's quite addictive and people get hooked, like using Uber versus catching a taxi.
If you stopped using Uber, you would have a bit of withdrawal and try and get back to the old system of how you got transport around the place. And really, what that means is it's getting the client to come back over and over again and have regular repeat ongoing work and to integrate Freelancer into the workstream of businesses, small and large, on an ongoing basis. I think we've got very, very, very strong retention on the Freelancer side. In fact, I think it's probably one of the strongest product platforms out there globally for retention on that side of the marketplace. We need to work on it on the client side. That's the key focus for the rest of 2024. The next is to reinvent the world of work in the world of AI.
I think we are in a very, very good place, really, on three fronts here. The first is that obviously our marketplace of 75 million people, which is the largest online workforce in the world now, the skills have been lifted dramatically through the AI tooling. So if you're an average copywriter now, you're an exceptional copywriter using GPT or Claude. If you're an average designer, you're now an exceptional designer now with Midjourney. This is starting to happen in software development with a bunch of AI tooling coming into that field. And effectively, every endeavor of white-collar work that man does will have AI-assisted tools. So the value of our marketplace has lifted dramatically from the skills improvement that they've gone through.
The next is that just inherently, the marketplace and how we interact in the marketplace and people chatting to each other and so forth is very well suited to the foundational models improving conversion throughout the flow. So that is, for example, if you post a project today, we have AI assist will help you write a project brief. We can then take that project brief and break it down into tasks and plan out a technical specification for that particular work. There's a lot of things we can do with assisting on the recruiting side with AI assistants. In fact, that's library in the marketplace now today and getting good results.
We just deployed this week an agentic framework to help clients basically activate on the platform, help them get projects posted, help them get the first steps into understanding how the platform and the product works into selecting a freelancer . I think ultimately we'll be able to do a lot of things around project management as well. It's a pretty amazing framework. In fact, I showed that framework yesterday to one of the largest BPOs in the world that runs a gigantic multi-billion dollar contact center business. There could be some potential we can take what we've done internally there and prioritize it for their contact centers. So I think we're really leading the way in kind of some of this agentic work we're doing here on AI.
The next thing is that we are about to, we're on the cusp of, I think, a great transformation with businesses around the world. Now, for those of you old enough in the call to remember 1994, that was the year that the geeks had email addresses. 1995 was the year your grandmother got an email address. Every business suddenly wanted to become an internet business and to do so, they needed to build a website. And that led to the dot-com boom and every business in the world basically throwing websites and so forth. Today, website development is a very large category of what we do in freelance. It's about 30% of the jobs that we do. Every business in the world needs a website in order to kind of reach customers and sell online and so forth.
Subsequently, you had the iPhone and Android revolution come out and people built apps. That wasn't as big as website design, but it's still a fairly substantial category of what we do. What will happen now, starting right now, is you'll get AI development happening across a range of different functional areas of businesses. You'll have AI-driven customer support. We have that live on the site right now for tier one support in a range of different functions. Immediately, you can see side by side when you compare human support to AI support, that the AI-driven support is a lot more empathetic. The answers are more thorough. The AI has infinite patience. And ultimately, you get better at customer outcomes. You'll start to see this in sales with things like lead generation and outbound ADR sequences or account development rep sequences being replaced by AI agents.
You'll start to see it in marketing. You'll start to see it in very simple functions. Every business in the world probably has someone on staff that answers a phone call, puts a booking into a calendar, and maybe takes a credit card, whether it's a hairdresser or a fruit or what have you. So you're going to see a huge explosion in development across these particular areas for small businesses, mid-sized businesses, large-sized businesses. And where are you going to go to get that done? You're going to go to the same place you get your web development done. You're going to go to the same place you get your app development done, which is going to come to freelancers to do so. Whether you hire a one-man band, a small agency, a large agency, or someone like Deloitte, you'll come to us to get that done.
So I think we're on the cusp of this. I think some demos will start to happen in the next weeks to months. In fact, tomorrow I'm doing a session with all the freelancers talking about our AI strategy, talking about the AI marketplace of services that we're putting live that the freelancers can show off, the services they offer to kind of deliver in this particular area. And I think this is going to be a massive, massive, massive industry. I think it probably is going to be bigger than web development because it touches on so many different areas of a business. So that's the second part of the product strategy. The third is rethinking client acquisition and well with Google. We're doing incredibly well on the acquisition front, which probably I'll talk about in a second.
We're seeing some good growth numbers on the early stage of the funnel and on the revenue side of acquisition and GMV side of acquisition and milestones. I'll talk about that in a second. The next is really to take the next leap in the evolution of the product quality and user interface and really start winning awards and get it world-leading. I think we're in a very good position here. Certainly, the feedback that we're getting across the board from our customers, particularly on the Freelancer side, is they're extremely happy with the improvements we've made to products. That's actually coming through with some awards we're starting to win. In terms of turning Freelancer from a painkiller to a narcotic, we are starting to see some pretty decent improvements.
In particular, what we're seeing at the moment is that gross marketplace value, which is payments to freelancers on a weekly basis, is starting to beat last year, year-over-year, and since June. We think in the second half of this year, it will start to beat 2022 as well. The 30-day GMV, so this is payments from new clients that have been acquired in the first 30 days, is already beating 2022 and has been entirely ahead of 2023 since mid-June. We're starting to see some growth numbers there. We're seeing a similar story with milestones created, which is physical cash that's been put into the system to pay for freelancers since late June, with the fixed price segment of that for fixed price projects, which is the main way you get work done, starting to beat 2022.
So we're really in the last six weeks or so, we're really kind of starting to beat 2023 consistently, particularly in the early stages of the funnel, and starting to touch on being 2022. So there is some growth there that's happening in the core marketplace, and it's turning around. And I think we're at a good place now to just keep doubling down on the product quality, the product improvements, the tooling improvements, the AI improvements, the collaborative improvements, and so forth. And I think we're at a pretty good place because on top of that, you've got all the enterprise work kind of layered in on top, which is going to start delivering some really chunky numbers, which I'm pretty excited about. It's been a long time coming.
In terms of client retention, really for the last year and a half, we've been seeing some really good lifts in the 7-day retention of clients. That is, client stickiness after 7 days of starting to use the product. It is coming through in the 14-day, in the 21-day, and the 28-day. You're starting to see lifts, just single-digit percentages up, trending up. That is great because every unit of retention is actually far more valuable, actually, than even an acquisition. And we'll continue to build on that with the collaborative tooling improvements that we're building in the product, which we'll talk about a little bit in a second. One example of which being project updates, which is a way in which clients and freelancers get reminded to provide updates about the projects to each other and get feedback.
One of the things that commonly goes wrong with projects on Freelancer is particularly because a lot of projects are one-on-one. You have one client talking to one Freelancer. So if one person doesn't show up or respond to a message very quickly, projects can fall apart simply because there's a lot of small teams. So we've built a system whereby freelancers are reminded on a daily or a weekly or whatever time basis that you like to set to provide a little update about the project, what the blockers are, what the next steps are, and maybe show a screenshot or provide a link to a particular asset for the client to see. And then provide a notification onto mobile. The client can then click a note, acknowledge it, or reply, etc. and so on.
Now, that's a pretty simple functionality, but those daily reminders form the basis of one of the most powerful forms of game mechanics in platforms today. So for example, Duolingo, for those people that use Duolingo, it's got the concept of a streak where every day you kind of check in and you kind of do your language lesson for a few minutes, and you stay engaged with a particular product. And that's been an incredibly powerful mechanism for Duolingo to become a very, very, very, very large business from a very, very simple game mechanic. And so we're kind of building this sort of functionality in to mean that clients and freelancers stay in touch more often and can acknowledge each other's work and progress with the project. So there's things like that that have been built in.
We also have a lot of work being done underway by Cohen's team on the audio and video calling and screen sharing. We've got a very robust solution, which is currently in demo on the iOS platform and soon to be fully launched in the next couple of weeks. And then following that will be Android. And that's going to give us a really, really robust system for people to make calls and communicate, which will work even when the phone's unlocked and the screen's turned off. So we've got all the background calling stuff figured out. So I think that's going to be a huge improvement on the Freelancer platform and even more powerful on the Loadshift platform where people are driving all day and need to interact by voice. Now, the AI work we're doing, I talked about that in a little bit. We see big uplifts.
Every time we do an A/B test and we apply effectively an LLM to help people fill out forms or fill out the next stage in the funnel or kind of proceed with their project or break it down or figure out the next plan, we're seeing really big uplifts. So we'll continue to double down on the higher tempo testing there and deploying that, in addition to building out the number of roles we have under our agentic framework to assist people on a managed service layer, particularly with the tier one support. So tier one support, tier one recruiter functions, tier one project management functions, and so forth. We're seeing some really, really amazing results. And in fact, we see much better results than you'd ever see from a human agent doing the tier one. Of course, human agents are always there. They're always there for escalations.
They're always there for tier two support. We're also seeing some pretty good results on the organic channels. So particularly with the rollout of the Google Search Generative Experience, I mean, Freelancer has a lot of content that's constantly being sucked into the various AI training models and the search engine and so forth for indexing. In fact, there was an article over the weekend, which I was interviewed by the Financial Times about Anthropic doing fairly aggressive scraping of our content. There is a good positive flip side to that is that the content we selectively decide to show is valuable content that lifts the prominence of Freelancer within these AI foundational models as well as the search engines.
And then we're starting to see that actually through quite a decent uplift in sort of SEO-driven traffic, which is up quite substantially from Google after the SGE push and rollout they did. So particularly on the acquisition front, we're seeing some pretty good numbers, and we're coupling that with some work we're doing on the retention front. So we're seeing some upticks there in the core marketplace. And I think the second half of this year is going to be core marketplace alone is going to be pretty damn good. This shows the uplift now in new user sign-ups, the verified sign-ups. We had a big uplift here from late May, 40% year-over-year in terms of sign-ups through improving the funnel to the sign-up flow. So that was a big win. And then we just continue to double down on great product and great design.
One of the things we've done is we've shipped dark mode into beta along with a number of other improvements, which is the profile pages, which has led to a small uplift in clients creating milestones and so on. The heads-down approach to really focusing on quality products and so forth is also paying off in terms of awards. We have the highest ranked of all the major freelancing platforms now for customer satisfaction, and anyone can independently search that for themselves on Trustpilot or on Sitejabber. We ranked 4.5 excellent on Trustpilot. In fact, our rating on Sitejabber was so high this year, we won an award because we had 96% positive feedback for the last year from clients. We ranked 4.7 out of 17,500 reviews.
In fact, one of our major competitors actually has pulled down their Trustpilot page because they've been review bombed because they've been raising their prices so frequently. The freelancers are very, very upset. So I think that's really paying off quite well. There's an interesting survey that Marco did on AI usage by the freelancers. Most freelancers are reporting that they are earning more or the same with AI tooling because their productivity has gone up, as well as most people are either see a bright future in terms of kind of where all of the stuff is going. I think the tooling is going to really take off in the next 12 months across a wide range of different industries. I think you're going to see some pretty amazing productivity lifts in many of the skill areas.
On the enterprise front, probably the most exciting thing we're doing by far is on the generative AI side of things. We've had an MSA in place for some time with a particular partner that has a lot of foundational work with one of the largest foundational models in the world that everyone would know about if I mentioned it. As I said, probably in the next 60 days, we'll announce the partnership publicly with all names. We're already sourcing many, many, many freelancers. In fact, there's over 80,000 freelancers now that have gone through the first stage of qualification. And then we're putting them through exams and then ultimately whitelisting them on the web platform to do work. There are three engagements with this particular foundational model. We're only doing the first engagement right now.
As I said before, this engagement should lead to, before the end of the year, 7 digits of GMV per month on an ongoing basis. And then there's a lot of upside from there with rollout into several other projects. It's RLHF training, safety training. It's looking at creativity and factuality of responses and so on. We can really excel at this work, this rating work. We have lots of freelancers available to do it. We can source on a sustained basis, somewhere around 20,000 freelancers a day if we're out there hitting the channels and pushing it and so forth. So we can assemble very, very large teams here, better, I think, than any of the competitors by far, simply because we have the range and the breadth of the users across all the different demographics.
As I said as well, we also won, as of yesterday, some engineering services work. So that's kicking off immediately to do a workflow integration to scale it up quicker. And I do think that engineering services engagement will ultimately be in the low seven digits in terms of the work because there's actually quite a number of things we're doing here, and it's also quite ambitious. We're actually rebuilding a whole platform for a particular customer. So that's good. And I also mentioned we've got a second engineering services proposal that we're sending out this week. It's actually in final form. After today's results, we'll approve that and we'll get that out. Quite a number of clients have also been acquired in the half. I think it's over 50 now that you've got actively engaged in various stages, Sean.
I mean, the key thing we have to do is I think we have probably the most stellar pipeline in terms of the brands and the quality of the companies in this pipeline. The trick is we have to figure out very quickly how to get them into a position where we can put a team full-time on that particular client. As for example, with the generative AI opportunity, I mean, this is going to be a very, very large client. In the longer term, I can see per annum the GMV for that particular client being eight digits in a fairly straightforward way. In fact, potentially even to eight digits per month, I think, with time. But the point is that each of these big customers have huge amounts of work. A Fortune 500 customer would spend probably at least $10 billion a year on wages.
So the ability for us to deliver significant GMV for each of these clients is there. It's just there's certain things you have to do to figure it out in order to be able to achieve that. The first thing is you've got to find deep pockets of repeatable work. So you've got to find pools of work where you can very quickly get in there and you can tackle it, where you can do this where jobs are being done over and over and over again many thousands of times. So, for example, with our field services, we have a large computer and printer company that needs thousands and thousands of laptop and server and desktop and printer repairs done every month.
With the generative AI work, it's rating work, and it's over and over and over again, and it's millions of hours' worth of work per year and so forth. You've got to find those pools of work. The second thing is you've got to integrate. You've got to be able to integrate the workflows so the payments flow really well all the way from that client ultimately down to the freelancers. You want to make sure that the more they consume your product, it doesn't become more of an overhead because you have manual invoices going backwards and forwards and so forth. You need to integrate into the vendor management systems. You need to integrate into their workflows.
You've got to make sure that if a job gets tasked from that particular client, there's a very, very smooth flow right through the customer acquisition, right through to the Freelancer delivery and to the payment. But we have clients in oil and gas. In fact, one of the largest oil and gas companies in the world at the moment, one of the largest e-commerce companies in the world, FMCG, healthcare, technology, and so forth. So there's a lot of work in Sean's division. We're building out on the delivery side for that as well.
The goal is really on a client-by-client basis to really figure out how can we get a team deployed, an integration happening, and a ramp up so that we get really a meaningful revenue stream because we do have an overabundance of clients, and we need to make sure that we kind of focus on building a scalable model for getting each of them to pay in a fairly substantial way. Our Global Fleet program continues to grow in terms of the volumes going through it. We've done, as I mentioned before, we've talked about the generative AI work. I probably won't get into that in more detail. But if you want to read it, here's an example of some of the marketing we're doing to the freelancers to get them on board. We send out an email like this.
We'll get about 20,000 people in a day signing up and going through the qualification, as I mentioned before. On the field services, we're now in 48 cities. 71,000 jobs are being completed globally. We have a proposal going in maybe today for some engineering services work to expand the capabilities of the platform and also the skill areas that the platform will handle. We're tentatively looking at 2 other markets now, 2 other countries that are fairly substantial and Western in nature to go into. There's some examples of kind of what the field might look like in terms of more enhanced functionality. In terms of the U.S. government work, we continue to win with this particular program. We just won in the last 2 weeks. We won 1 engagement. It just shows you kind of the sorts of things you can do on a Freelancer.
It's for navigation on the Lunar South Pole of the moon for the Artemis manned space mission, particularly when the astronauts go into craters and parts of the sky are obscured. They want a way to be able to navigate in that sort of environment. We also won another engagement for improving the simulator for a space robotics operating system framework and so forth. And we had quite a number of engagements where we awarded winners and prizes. We had one for detecting 1 mm-10 cm particles in space and tracking them and remediating them. We had another engagement which we awarded for refueling spaceships in microgravity. Another one around art and so forth. So, as well as ones for improving the visual disturbance in air taxis. We also did something with the Australian Space Agency around an arm for a lunar rover for them.
We also have an engagement separate to all this NASA work that we're bidding on at the moment with the Department of Energy in the U.S. as well, which could be a seven-figure engagement, which we're working on. We also have two major engagements which continue to happen. One is the $6 million work on gene editing with awarded about half the money out of that so far. And again, when you look at the quality of the entrants, the quality of the client, the quality of the work, I mean, it's the most high-intensity work in the world you can think of. We had MIT, Harvard, Stanford, Yale participate. We had biotech startups, research groups, individual researchers, and so forth. Incredibly skilled and talented service providers, incredibly high-end clients, incredibly high-value work. It really shows off the power of the platform.
No one comes close to kind of what we can do in terms of the scale of providing and delivering work. Then we also built a water precipitation device for the Bureau of Reclamation as well and delivered that. On the Escrow front, you can see here this is the long-term trend in growth payment volume. We were lapping sort of year-on-year, a bit of a spike. But as I mentioned on Friday, the buyer and the seller for one particular transaction signed. In the next two weeks, we'll have a $50 million transaction go through. So we'll have another big spike up in the next quarter. To get away from this lumpiness in the spikes, we're working on a checkout solution.
We're going into one of the largest shopping carts in the world in a fairly unique way because we'll be the only large-value payment system in that particular shopping cart. We're going into beta this quarter. For that, we actually have all the shopping carts pretty much lined up one after the other. So there's about four of them in a row, which we're going to deploy in. I think it's going to be pretty unique because at the moment, you can go to a shopping cart. You can set up a consumer electronics store or a clothing store or what have you, but you'll be able to, using Escrow, set up a place for selling cars. You've got to set up a real estate marketplace. You've got to set up a fine art store. You'll be able to set up a diamond store.
You'll be able to set up music rights, oil royalties. Whatever you want to do, you can do on these platforms. We're going to get into all the biggest names. That's all pretty exciting. That's going to be starting to go into beta as of this quarter with one of the platforms. We also updated our pricing for the very first time in 20 years on Escrow. Our price card used to cap out at 25,000. The feedback from the sales team was every single engagement they're going into because we're a large-value payment, you would go over that $25,000 price point. Then they'll kind of ask us, "Well, what happens to the rate card after that?" We've extended that rate card out to $10 million. We also listed our pricing across a few different areas.
Really, what we did here is we're looking to get the take rate up a little bit. When we bought the business with Escrow, the take rate was 1.45% blended. What we did, probably mistakenly, is we gave the account management team a bit of flexibility to negotiate pricing because we've got the top of that rate card way too frequently on deals. And so because there's a little bit of flexibility on that, that kind of, I think, got taken advantage of. So I think the take rate ended up last year being about 0.9% as opposed to 1.45%. So we've done a bit of a reset there in terms of the pricing. We've gone through all our major partners and talked to our partners and explained it, etc. And I think that's been fairly well received at the end of the day.
We should see a lift in the take rate for Escrow.com. The other thing we've done is we've got some margin improvement we've made in this business as well. We did some cost cutting in a few particular areas, particularly around an affiliate program that wasn't doing particularly well, which we replaced with a market development program. We should see a few points in improvement in the gross margin of this business as well. In the last quarter, we've done, I think, a fair bit of improvement in the fundamental take rate and the fundamental gross margin of the Escrow.com products. Of course, the most exciting thing is getting the shopping carts out live. August is working hard. We've got some pretty slick-looking designs and really getting it as smooth as possible to provide a really nice checkout experience.
Now, the one thing to remember is that with Escrow, we are high friction simply because we are high-value complex transaction payment methods. In particular, we KYC the vast majority of our transactions because they're above $3,000 in the US or above AUD 0 in Australia. And so we do do quite detailed KYC, AML, sanctions checking, politically exposed persons checks, negative news checks, etc. So the goal is to make that process as slick and as smooth as possible. But the flip side is going into these shopping cart platforms, this will be the very, very first time they'll have large-value payments. So a lot of these platforms, it's going to be a coming of age because no longer will they be selling sort of micro-priced clothing or retail products, but they'll be able to sell high-end, valuable items.
I think it's going to be, I think, a real huge change for the ability of these shopping carts to sell in the big end of town. E-commerce is a percentage of retail sales, but only around 11% in the U.S. and dropped a few points since COVID. It will never get to 50% of sales unless a large-value payment is handled properly. That's what we enable almost uniquely. Almost uniquely because effectively, we have in many ways a regulatory monopoly on large-value payments with Escrow licensing and the licensing that's in 55 jurisdictions now. To achieve that licensing took a long time. It took us a decade. It is not something we could do again from scratch.
If we had all of the best of the team here and we wanted to start Escrow all over again, we couldn't do it simply because this business was started by Fidelity in 1999, stock-backed funded at $30 million. They took about a decade to get their first licenses. Then we took a decade getting the rest of the licenses in the U.S., Canada, and Australia. To do that again is an incredibly complicated and difficult task. Airbnb started trying to get the licensing at the same time we did in 2015 and still haven't completed the U.S. program. It was a huge amount of effort. It is a very unique business. The other thing we're doing is we've got a couple of automotive engagements, which I think are quite substantial.
We submitted a proposal to a tender to a large U.S. automotive marketplace that's looking to turn payments on. There's another Canadian marketplace as well that we've put a proposal into. Additionally, there is an Australian government RFI that we responded to, which is quite substantial. That particular RFI is on pause mode, but it will continue after they've sorted out a few things they've got to figure out on their end. There is another substantial one in the wings there that ultimately that's going to be about AUD 13 billion of volume if that gets over the line. It's currently in pause mode while they figure out a few things because it's quite complex and quite a substantial change to how vehicles could be sold in that particular jurisdiction in Australia. The IPv4 space is doing pretty well.
That $50 million transaction we're doing in the next two weeks is an IPv4 transaction. We'll make an announcement on that as well when it happens. The domain name space ticked up a little bit. I did expect this to tick up a lot higher, a lot quicker this year. It's proportional to venture funding activity. The venture funding activity is not really back yet in a big way. It's happening at a macro scale with AI, like for example, xAI, which is Elon Musk's group, raised a huge amount of money earlier this year. We're not seeing it broad-based across startups yet. It's really these mega transactions. But we do think we'll tick up. I think there's going to be a big boom in AI-powered transformation of all these different industry segments. I think we'll be upset when that starts to happen. It's ticking up.
We have put some pretty marquee domains through in the last couple of weeks. But I do expect that to be a lot higher than what it is. Then on the Loadshift front, Loadshift had a great first half. GMV was up 43.6%. We had pretty good engagement with quotes, which were up triple digits, 105% year-on-year. Average quotes per job is now eight. So it's a fully functioning marketplace. We've successfully transformed the Loadshift acquisition, which was the bulletin board and the Freightlancer merger, into a fully functioning marketplace that Cohen and the team have done extremely well. We had a 77% increase in award loads year-on-year. Delivered loads doubled to 105% year-on-year. Project fees were up almost doubled and so forth. So the numbers are all up and to the right on the Loadshift division.
Now, the big thing to remember here is we've got now a fully functioning marketplace, but we have a lot more freight than we're actually converting and monetizing just because we're changing a business model here from one, which is a membership model, to a marketplace model. We've got the award rate now up to about 30%. You can see that it started at 0 a bit over a year and a half ago, but now it's at 30%. I personally think we can get us to 60%. The big thing now to get the next big uplift in the award rate is to get the calling on the platform rock solid. Most drivers are in the cab driving the truck all day. You don't want them sitting on their phone kind of texting and chatting while they're driving. So you've got to have a voice interface.
Cohen has been working with his team on getting that rock solid. It's audio calling, video calling, and screen sharing. The beta for iOS calling is live and currently in testing with users. The background calling, which allows the phone to ring when it's locked or the screen to be turned off, will be live when?
Couple of weeks.
Couple of weeks. And then about two months after that, we'll have Android fully delivered. So we're chipping away at that. But I think by the end of the next quarter, we'll be in a pretty good position. Of course, this goes live not just for Loadshift. It goes live for Freelancer because it's the same code base. And I think this is probably going to be the next big leg up on the award rate side, which will get us up a fairly decent number. So this business is going very strongly. We have a lot of freight on the platform that we just need to convert. We have a lot more freight in a greater network of assets that we've got that we've been picking up. And we're going to have a lot of deco with our booth.
We have a lot more freight we can draw into the platform that we haven't drawn in. And then we have some enterprise engagements as well, which will also lift the numbers fairly significantly with car auction yards, with auction houses, and machinery platforms as well, which we're in the. Yeah, we've been talking to it for some time, and we've just been a little bit hesitant to bring it on because we want to convert the freight that we have rather than bring on a lot more freight and bring the conversion rates down a little bit. So we're working away on that. I think operationally we've got that in a very good place. Product-wise, we've got it in a very good place. There's a bunch of features coming through shortly. I think the load tracking is in shadow right now and probably live this week.
So you'll have all the like a FedEx style interface. So at any point along the way, you'll know where the loads are and the trucks arriving now in 30 minutes and this that the other. So that's great functionality. And Drew has been working with a whole range of very large enterprise clients to bring on more freight, primarily in the construction, industrial, manufacturing sort of areas. So we've got quite a good haul for Loadshift. And there's a picture of sort of things we moved in the slide. It's all the ugly heavy stuff. And the machinery is the most valuable freight, and that's 30% of what we do, which is pretty good. So overall, probably it was kind of flat. There's a loss of AUD 1 million, which is a fairly small number.
I think a lot of upside in the core marketplace, a lot of upside in the enterprise side of the division. We've got all the stuff happening over in Escrow with the shopping carts. We've got a big transaction going through, which we'll do a big press release about because that'll be our biggest transaction ever to the platform happening in the next two weeks. And Loadshift is going pretty well. So that's where we are. I think we're going to have a better second half than we did in the first half. But I think we're in a pretty good place. We've got the cash flow up. The cash is flat in the quarter. And it's a very minor loss of AUD 1 million. And I think that's going to go quite positive quite quickly. So that's where we are.
So I think what we'll do now is we'll open up the Q&A with the audience. As I mentioned before, you can address questions to either myself, to Neil Katz, the Chief Financial Officer, to Shaun McMeeken, who's the head of sales, to Adam Byrnes , who's the head of product, Andrew Bateman, who's in charge of enterprise, August Piao on the Escrow side, and Cohen Wisniewski on the Loadshift side. So we'll open up now. So Alex, maybe if there's any questions in the chat, please read them out and who they're addressed to and who they're from, and we'll go through them.
Wolfgang asks, "Is there a plan to integrate Escrow with Loadshift?
Wolfgang asks, "Okay, is there a plan to integrate Escrow.com with Loadshift?" That's a very good question. The answer is yes. In fact, there's a couple of different ways we're going to do this. One is that we make about, as I mentioned before, it's a little bit under 1% take rate on the Escrow.com side. And on the freight side, we're making 11% now. And that's going to tick up to 13% because we're cycling off the membership discount that we provided through the transition period. So the freight division will make about 13% in maybe about the next six months. Should get to about 13%. So anything that sells that's physical through Escrow.com needs to be moved. So for example, eBay Motors in the U.S., we have the exclusive payment methods for them. Some of those cars need to be moved.
They need to be moved interstate or internationally or what have you. We sell a lot of machinery through Escrow. We sell a lot of physical stock. And so at some point, we will go international with load shift. It's not going to happen this year. It may happen towards the end of next year. Maybe we'll turn on somewhere first like Canada because we've got a lot of mining industry experience. And I think maybe we'll do something like that first. And then after that, we'll just go global. But from there, what we'd want to do is we want to turn on the ability to get freight for anything you buy or sell on Escrow. So that would lift the margin up quite significantly from 1.1% to an extra 13%, which would be great.
On the flip side, there are some payments that do want to use Escrow for the movement of freight. It all depends on the particular industry and the particular thing that's being moved and so forth. For example, in the past, we've had someone use this to move manganese on a railway line in Africa because they couldn't find insurable options that made sense. And we've got some tentative plans to merge the Escrow and the Freelancer code base to modernize the Escrow product fairly substantially and take advantage of all the features we've got on the Freelancer side of things. There is potentially a plan to move the Escrow stack, the front end of the Escrow stack on top of the Freelancer stack.
And that will gain us a modern front end, a design system, all the AI tooling that we've built around agents, a bit of a transactional system that we've got, real-time chat, calling, etc., and so forth. So there may be some integration work we do there in the midterm on the Escrow side with the actual product stack as well. Question? Feel free to ask questions. Put them in the chat. Send them through, and I can read them out. Any questions, Alex? Don't be shy. There's a few on the call. Sometimes it takes a minute or two for people to put their questions together. I've mentioned everything in my notes. Okay, well, maybe while people think of questions, I'll just go through the highlights again just at a very, very high level. So the Generative AI work , I think, is going to go extremely well.
I think that'll contribute by the end of the year to low seven digits at GMV per month. We've got an engineering service engagement we've got approved for on the Generative AI work as of yesterday. I think that will lead to a low seven digit engineering services engagement with a customer. There's a lot of work here. This customer is going to grow very, very strongly. We had a great actually did a full day with one of the senior executives today. That's fantastic. In the core marketplace, verified sign-ups are up 40% year-over-year, particularly on the acquisition front with GMV from 30-day customers, customers up 30 days and milestones created. We're beating since really since June, year-over-year 2023 numbers, and we're starting to touch 2022 numbers year-over-year in terms of bidding them.
We've had some A/B test wins in payments, the UI, collaborative tooling, and AI conversion. I think we're going to have a pretty decent for the rest of the year numbers on the GMV side on the core marketplace alone. That looks pretty good. The retention between one and four weeks is doing okay, and we'll continue to double down and build on that. On the Escrow front, we've got a very large transaction we've got approved as of Friday, which will happen next two weeks. So we'll have a very good quarter for GMV for Escrow coming up. We're doing some shopping carts. We've got a couple of automotive marketplaces working at etc. and a range of other customers. The Loadshift staff are going pretty well. The next big leg up in a big way will be from the calling work we're doing.
That's at a super high level of the summary. Does anyone put any other questions in?
Simon asks, "Share price has been quite depressed for some time. Where would you like this to be in 12 months? And what are the top two or three things that will lift this price from where it is?
Yeah, that's a great question. I'm not happy with the share price as being. It has been a bit of a flaw, but it's been the worst low share price ever. And that's on my head to improve. We are doing an investor day next week on the 6th of August. I do encourage anyone that's on this call to sign up and join on that, where we're going to go through in each of the business units. We're going to go through in a lot more detail what we do. So I invite you to join that. I think there's really two things we've got to do. One is we've got to deliver a financial catalyst. So we've got to deliver a very, very solid quarter and with an ongoing solid improvement in revenue and in profitability. So we're head down working on that.
I think we've got the profitability to break even now. We're just going to get it to reliably and lift over time. And we've got to get the revenue numbers up. So that's what we're working to do. And I think we've got a pathway to do that across the work we're doing in the core, the work we're doing with enterprise, particularly generative AI work, the field services work. I think we've got some improvements happening there as well shortly and so on. And then we've got to get up there and do investor relations. So we're starting to do that. We've got a note coming out from Edison Group, I think, in the next 24 hours, a note coming out from Pitt Street Research . We're doing the investor day. Marko Zitko is now in charge of investor relations.
We're doing investor meetings, one-on-ones in Sydney, Melbourne, and Brisbane after these results. We've already got people signed up for that, which is good. We just have to get out there and speak the story more. That's on myself to do with the team. So it's really get the financial numbers printing, which I think we'll see visibility of kind of where that's going to happen now. Then secondly, get out there and talk the story more. Then separately to that, actually, we have a lot of freelancers that want to buy stock. We're just trying to figure out a pathway to do that. We actually sent out one notification because we had a broker that was going to onboard our users from around the world. Now, obviously, they're from every market you can possibly think of.
We had 1,000 requests to open share trading accounts in 24 hours from one notification that went out. So we're working through the complexities of that because, obviously, with an Australian broker, you need to onboard people, get the idea across. They want to have an Australian bank account to be able to do clearing in and out of, etc. So there are a few things we're just trying to figure out there. If we can crack that as well, we'll have an ongoing stream of our customers wanting to buy stock. And there's now 75 million on Freelancer, 2 million on Escrow.com. I don't know. We're probably 78 million across all the group, roughly. So we want to get that moving as well. So that's one of Marko's sort of KPIs to figure out.
But yeah, I don't think the stock price will stay down here for very long. I know it's been here for a little while, but I think now is the time it's going to start moving.
Stephen Rogers asks, "Matt, what are you doing to attract institutional shareholders?
Well, I speak with institutional shareholders actually every week. We have a lot of inbound. We have an outbound program, as I mentioned, with a 1-on-1 signed up in 3 cities. We've got some analysts covering us now, which is, I think, an important part of getting institutional shareholders. I think also, as I said before, we've got to get some numbers that print where people will wake up. I don't think that we're too far away from being able to do that, where people will look at it and go, "Hang on a sec." Yeah, if you think about it, 3 marquee companies, all leaders in various ways in their various segments. It's the largest online workforce in the world. It's the largest online Escrow company in the world. It's licensed in 50 jurisdictions. I mean, those licenses alone are worth $1 million each.
And we've got Loadshift, the largest heavy haulage marketplace in the country. And all of that's a AUD 100 million market cap. It's ridiculous. So I mean, that will change, I think. And I think when it will change, the perspective on stock will change quite violently because it's a very tightly held stock as well, which hasn't helped on the way down, but on the way up, I think will help very well.
John Phipps asks, "Matt, any thoughts on why Patrick joined? Has he attended any board meetings yet?
Yep, he's attended one board meeting. Patrick is a fantastic operator, obviously, a Catcha Group. He's built billions of dollars of shareholder value. He had iProperty Group. He had iCar Asia, iflix, Frontier Digital Ventures . He just IPO'd an oil platform in the U.S. publicly. I mean, he's a real mover and shaker. He's a great guy. He's a great operator. In fact, I spoke to him yesterday about something. Yeah, we've had one board meeting since he joined. That's been pretty exciting. I'm glad to have him on board. I think it'd be a great help with connections to customers, help with improving the business across many different fronts, attracting great talent to the business, and so on. I'm very pleased that Patrick joined. Any other questions? Last point. Don't be afraid to ask questions.
We'll hold open for a couple more minutes if you want a couple more questions. Yeah, as always, I'm looking to bring great people to the team. I think the team actually, from a management perspective, is probably the most solid it's ever been. I think probably everyone in the room would agree that. We've got the people in the room here and a number of the executives overseas, I think, are executing quite well and quite solidly. So I'm very happy where we are with the company right now. And we just need to head down, double down on kind of where we are at the moment and really execute. I'll leave over 60 seconds more before we close it off. Get the last question in if you want to ask one.
DL asks, "When do you expect the AI work revenue to start contributing this quarter?
Okay, so the question is, when will the AI work start contributing? Well, there's a few different angles of AI work. The AI A/B testing work that we've been delivering on conversion of the funnel is already paying off in little increments as we kind of deploy it across the product. The agent work, I think we've made some good results this week because we've got a client activation agent that's been deployed. And I'm looking forward to seeing the numbers actually stay because it's been 24 hours since it's been live. And I think that will help on the acquisition front. On the Generative AI work, it's already starting to pay money. We've got, at the moment, probably about 6,500 freelancers on board doing work. I want to get that up to 60,000 that are whitelisted and doing work on that first project with this large foundational model.
Yeah, as I said previously, I think this will deliver low single-digit GMV per month US before the end of the year. There's a lot of upside from there. I want to kind of sandbag it a little bit. I don't want to overpromise and underdeliver, but I'm very comfortable that we should be able to get to $1 million a month US of GMV from that one particular project and one particular customer before the end of the year. Talking about the engineering services gains we've already started, we've got a small piece of work to start off with, but we've got a proposal that's going in maybe even today or tomorrow for probably about $1 million worth of engineering services built.
Stephen Rogers asks, "Matt, do you intend to help Patrick in becoming a shareholder?" [Tightly held stock.]
Look, we certainly had a conversation with Patrick about being a shareholder. In fact, probably people don't know. He wrote a term sheet for $40 million for Escrow back in 2021 to finance it as part of his Catcha Group. So at a very attractive valuation. So he's already written a very substantial term sheet for one of the subsidiaries. We didn't go ahead with that for various reasons. I had a U.S. investor and I had another Australian investor also submit term sheets to fund roughly about $50 million into Escrow. So he's well across the businesses and he has deep pockets, obviously, through his group, etc. We've had some recently. We've had a few conversations about him coming on board as a shareholder. I haven't pushed it because there's a few other things I'm working on at the same time.
So I'm sure he'll become a reasonable shareholder in time. It's not a recent discussion we've had, but he has written a term sheet for $40 million before for one of the subsidiaries. Any last question? Okay. We can always follow up separately with myself on matt@freelancer.com or investor@freelancer.com or to Neil or to anyone else in the room. I'm very happy to arrange conversations, management or one-on-ones, etc. That's one last question.
One last question. John Phipps , can you expand on what else you are working on?
Lots of things, John. All shall be revealed soon. I'm working on a few things. Yeah, I'll probably give it a few days. Okay. I'm happy to have your thoughts privately with people, do a one-on-one. And remember, the investor day is the 6th of August. So please sign up for that. There'll be a thing going out to the investor mailing list if not already on that. Plus, we're doing one-on-ones in Sydney, Melbourne, Brisbane as well. Just get in contact with investor@freelancer.com if you'd like to do a one-on-one face-to-face. Or if you want to do it over the phone, we can do that as well. Okay. Thanks, everyone, for attending the first half of 2024 financial results presentation. I look forward to speaking to you in the next quarter unless you reach out privately to me in the meantime. Thank you.