I think so. Hello and welcome to the Freelancer Limited 2025 half year financial results. My name is Matt Barrie. I'm the Chief Executive and Chairman of Freelancer Limited. With me today I have Neil Katz, the Chief Financial Officer of the business, Andrew Bateman, who's the Vice President of Product Management and Enterprise, Mas Mohammad, who is running Loadshift business, and August Piao, who's running the Escrow business. As always in the Q&A, you may direct your questions to myself, any one of the executives in the room, or follow up afterwards for a one on one by emailing investor@freelancer.com. At Freelancer we are building the Amazon of services. We are in the fields of labor, payments, and freight. We have three businesses that are all very strategic and groundbreaking in their own right and world leading.
Freelancer is the world's largest online crowdsourcing marketplace with over 82 million people, connecting people up from all around the world in any country where there's electricity and Internet that we're able to legally operate. Escrow is the world's largest online Escrow company that secured over $7.7 billion in payments for expensive and complicated items. Loadshift is an Australia-only business but the largest heavy haulage freight marketplace in the country. We will be at some point in the future expanding overseas with the goal of solving trillion dollar problems. Each of the fields of labor, payments, and freight are very large trillion dollar markets. In the field of Freelancer, 5 billion people need a better job, and intellectual capital is the battleground in the 21st century. As Mike Milton says, the defining characteristic of the 21st century will be the competition for intellectual capital. With Freelancer we make it real.
We turn your dreams into reality whether you're a small business right up to a very, very large business in terms of delivering products and service. In the future for Escrow in the payments field, trust closes deals, and in high stakes transactions, bulletproof Escrow isn't a luxury. We ensure that transactions happen, such as we've secured things like the sale of Grok.com to xAI, we secured the sale of meta.com to Facebook, we secured the sale of chat.com to HubSpot, which ultimately sold it onto OpenAI, the field of AI. That's just in the field of AI alone, in the field of domain names. When you transact a large or valuable item over the Internet, there's risk. Escrow powers global trade without borders by reducing that risk. The freight field with Loadshift, it's logistics born digital. It's basically Uber for freight.
It's the same code base as Freelancer, just running for a dedicated industry being Loadshift, and you know midweek we get about 300,000+ km of freight per day, which is on par with the distance of the Earth to the Moon and exceeding it some days of the week. We keep the world supply chains running. How we generate revenue, it's quite a simple business model. It's a consumption-based model. With Freelancer, we do drops from $10- $10 million with 80 million users. We collect the 3% fee from clients and 10% from freelancers once you award the job and the job is accepted, and we get some optional other revenue lines and premium add-ons. With Escrow, we take a very small commission for securing transactions. We've been averaging about 1.64% on each transaction as a whole, and basically repeat different deals. Monetized trust at scale.
At Loadshift, it's the same model as Freelancer, 3% + 10%, etc. Our competitive edge is monetizing critical business infrastructure across the digital economy. Now, getting into the half year results. The gross marketplace value of the business is $439 million. The revenue was $28 million, which is up about 8%. NPAT swung positive to an all-time record for the half year of $1.9 million. That's a big step up from last year where it was slightly negative. The operating cash flow grew very strongly to $6.8 million positive. That's up 209% in the half. The cash flow is also very strong at $3.3 million, up from a small negative to flat number of last year. The cash balances grew for the fourth quarter in a row.
This time it was up 12.3% from the most recent half year number, and that was also taking into account that we spent $733,000 increasing our second Loadshift to 64.2%. I think overall the high-level story here is an all-time record half year profit and quite robust cash flow, powering the growth into the second half. Looking across at a high level by segment, and I will do some commentary after I go through the first set of slides here just on the overall high-level numbers. The GMV was up a bit, $66.5 million. The revenue was up also 3% at $21 million for Freelancer. You know a big win we had in the first half. You'll see the video that we've played at the beginning of the call today is we were a joint winner of National Open Innovation Services 3. $475 million U.S. dollar contract.
You know, we've been working with NASA for about a decade. We were a winner of NOIS2, which started off at $25 million and jackpotted to $125 million, $175 million I think it was. $175 million. That contract actually still has some spare powder in it. That was a five-year contract. This new one is a ten-year contract. The original number of vendors in NOIS2 was 32 and that's been actually reduced to 25 in NOIS3. We are by far the largest platform in that universe and we very frequently win task orders or partner with other smaller entities with domain expertise to win. This is a great validation of the business. We obviously work with NASA on very high-end scientific and technical breakthroughs and moonshots.
As you saw in the introduction video, the sorts of things we've worked on in the past have included how to detect, track and remediate, and I mean shoot out of the sky, you know, small orbiting debris between 1 millimeter and 10 millimeter, which, you know, we came up with quite a number of different solutions and different approaches and different technologies. You know, we've done things recently such as refueling spacecraft in orbit, new techniques for doing that in microgravity environments, etc. We've got a big one that's actually been running just right now, which actually is a mass market, but it's had heavy support from the astronauts in educating kids with a zero gravity device replacement that goes up with the Island 2 mission into space. I'll talk about this later. It's not just space.
We do everything from product design to electrical engineering to software optimization and so forth. That is only going to increase over the next 10 years in terms of the diversity, complexity as well as the sophistication of this work. The providers that enter these contests are really tapping into 80 million minds for really true genius and a very large variety of approaches. Some of them unorthodox, some of them really delivering breakthroughs. We've landed those moonshots successfully one after the other through the entire history of the program. We're pretty excited about being selected and honored to be selected as part of those 25 vendors. You know, moving forward I think we'll play a very, very big role in that program for the next 10 years. In terms of other momentum, we added 2 million new users in the half, up 17.5%.
We have now extreme liquidity in the Freelancer Marketplace, 56 bids per project. In terms of contests, we actually now have 772 entries per contest, which is pretty insane when you know that contests can run anywhere from $10- $10 million+ Escrow. The revenue was $372.5 million. Bit choppy on the GMV number because we do have these very, very large transactions. We almost had a big $35 million transaction happen in the cycle; unfortunately, that didn't go through, and there's a couple of very large $10 million transactions which are partially funded or pushed open until the next quarter. That could have easily been quite a positive number on the GMV. The revenue is up 32.3%, $6.1 million. That's doing very, very well. This is our fifth year of profitability with Escrow.
We're out of tax credits now, so we will start paying tax for this business, which is a good first world problem to have. We also were live with Shopify with the beta. Our golden transaction went through, which is a full end-to-end test of the system and paid. That transaction was a $12,000 U.S. dollar transaction. You can imagine that using the traditional low value payment methods such as credit cards and PayPal and so forth, it would have been a little bit difficult to take a payment of that size. The big thing here is that you'll be able to use a Shopify site to take large value payments, for example, sell a house, sell a jet, jet parts, intellectual property, a business, etc. This is a brand new capability. We've been taking it very, very slow, but funds are flowing through now and again.
We would just be gently stepping us up because we need to make sure that we have everything in place with the business to be able to scale because obviously they're very, very high volume. Ecosystem domain volume is up a bit to $196.9 million, and it's got a bunch of positive interest from import export. We have a very, very, very strong integration pipeline with Escrow that Elliot and the team have been driving. Loadshift revenue is up to 9.1%. We twice hit an all-time high monthly revenue in the first half, beating the previous numbers by 12% and 8%. I'm pleased to say that we've rolled out in-app video and audio calling, which is the big thing that we've been waiting on to really drive up the award rate. The award rate is actually slightly higher than 27.8% at this point in time.
We actually pushed an even new update to the audio/video calling this morning. For those of you that use the site, update your apps and the calling is getting fairly robust and there's quite a bit of push going into it. I expect over the course of this quarter to be able to report that the award rate has been nudged up to a higher level and we'll see how high we get it to be. We're also going to be opening up the calling on the site in terms of how it's accessible on both Loadshift and on Freelancer. The important thing to remember here is that a win on one platform gets a win on both platforms. Freelancer has been progressively benefiting from calling being rolled out.
We've heavily restricted the calling on Freelancer because we have so many users and you want to avoid spam vectors and so forth. Now we've got a real-time data pipeline running on everything that happens through the sites. We can control that, so we will be massively opening up quite a number of features including calling on the platform. That was great. We had record half-year profit and cash flow with all the Ethereum engines sort of starting to get primed. The big thing now is just needing to focus on the revenue number. Just focus on that revenue number and get them all pumping. We've got one business at a pretty good rate. We'll get all the businesses at that level and we will be flying. This is what the cash flow looks like profile for the first half. You can see that was really good.
$6.8 million of operating cash flow under the new AASB 16. You've got to account for lease liabilities in such a manner on your books and there's basically de minimis CapEx in this business. The $733,000 is for increasing our stake in the Loadshift subsidiary. That gives you a rundown of where we ended up: $26 million cash in bank. You can see the NPAT swung quite positively. We went from a negative NPAT in the first half of last year to almost $2 million in the first half alone. It's the largest half-year profit to date in the company at all time. Net revenue is up at 8.5% powered by Escrow and steady marketplace growth. We did get the gross margin lifting just a little bit. The lapping number in the first half of 2024 was abnormally low for our gross margin. That's not actually accurate.
Previously our steady state gross margin has hovered around 83%, 84%, 4% roughly. In that particular first half we had some abnormals from, I think it's primarily fraud. Is that correct, Neil? A toll fraud attack on the two factor authentication that we're running with Twilio. That has been rectified. We do have a very strong gross margin historically. At the very beginning of the business I think it was about 85%, dropped down a little bit over the years and we managed to get a little back up. That's great. We've been tightly managing our costs. Admin is down 3%, occupancy down 8%. We're gently nudging up the marketing at the same time a little bit. You can see that quite a number of numbers swung positive here in the first half. Net profit before tax was $2.8 million for the first half, up from a negative $1.4 million.
You can see that it's about a $4.2 million swing in the net profit before tax number and overall operating profit before tax about $1.8 million. That gives you a good idea of where we are. Obviously, I mean, I've said in the past, about a year ago that my goal was to get to half a million dollars a month profitability or operating profit consistently every single month, consistently forever and then build a base and then grow the revenue number as much as I can and just gently nudge up the profit over time. I think we've done a pretty good run to kind of get to this point.
We always have a little bit more to go to get to the number that I was putting it back there, but it's pretty close and I think that will provide a good foundation and stability with the business in terms of the financials. We see there's an impact number coming through which is fantastic. The goal here is I won't be happy until I've got double digit revenue growth numbers across all businesses. We do have a little bit of cost savings still to come. There is still fat in the business in particular areas. I'll highlight that there is fat, for example, in hosting costs still and we have a plan to reduce that a little bit. There's a little bit of fat still in payments infrastructure because obviously we operate with so many currencies across the world, etc.
We've got plans to reduce that with better integration of select payment gateways, which will increase not just the acceptance rate of certain payment types in certain geographic areas, but also cuts in costs a little bit. We've got some more profitability savings to come in addition to just working on those revenue numbers. As we said before, the cash and cash equivalents were up again for the fourth quarter in a row. This is up 12.3% to $26 million. Receivables down a little bit reflects a shift in the mix of payment processes with shorter settlement times, which is better for us, and the right of use, lower depreciation in line with lease terms and reduced lease liabilities. Neil's done a very good job. Every time we've got a lease renewal, we've managed to chop some off from all the offices around the world.
Obviously, we will continue to do that as the leases kind of mature and so forth. Congratulations Neil for kind of making sure that cost control has been well applied. The reserves are now 14% simply because we increased our stake in Loadshift, and so the non-controlling interest reduced due to acquisition of some minority shareholders in Loadshift. Some high level vision, strategic vision, and 2025 outlook. I think we've got, we got the, you know, operating profitability. The operating leverage is starting to show through with the profitability. We're focusing on AI leadership. We'll talk about a bit more in the commentary. The second half focus is really productizing our gen tools for customers and partners. We've also got a big engagement doing some stuff with foundational models that we're being paid for right now with integration. We have a bigger thing coming as well in that particular area.
I'll save that for the commentary area. In terms of customer acquisition, we did have a bit of a pullback in the cash coming in from new customers in the first 30 days. In the Freelancer business, I personally wanted the number to be significantly higher. Had a bit of issues over the Easter period where we were a little bit thinly staffed and we had a few bug rollouts. Not entirely our fault. We actually had an issue with the Chrome browser rollout, which affected quite a number of websites. Bay data, it's trending back now but we had good user sign ups and we've also been improving our workflows and my goal is to get the cash coming from customers back up to over 20% year-on-year minimum. I know how we can get there. Big focus on retention, the video calling rollout.
It's going to be a big one. We've got a bunch of smarter matches happening right now in development, agentic framework improvements. It's pretty crazy the sort of things we could do now with AI agents on the side. Continue the cost discipline and if we can just keep chipping away on the cost discipline while spending the money where we want to spend it, if we spend it on good people in the company, nudge up the marketing a little bit and kind of trim the fat from waste in a few areas. There are some waste in a couple of areas. The catalysts for Freelancer are continued customer acquisition, AI-driven demand, NASA government wins amongst other things. There's a whole bunch of really good enterprise opportunities here.
Escrow, expanding the high value verticals, expanding the integrations, continuing the Shopify rollout, Loadshift, enterprise freight growth with the audio/video calling, GPS tracking, onboarding improvements, et cetera. We're currently lean, focused, and ready to ride the wave of AI and we want to get basically the revenue across all the business lines now to double digits and sustainable profitability. The cash has been building up and want to continue to keep building up. What I might do now is just go across to the commentary and is that coming through? No, it's not. Okay, here we are. I'll just go through the commentary in a bit of an overview. I won't go through it in huge amount of detail but I'll leave to exercise the reader but then I'll open up the Q&A and again you may address questions to myself or anyone in the room.
We talked about the high level numbers for Freelancer and you know I've talked about the users being up 17.5%. The average project size continued to grow. It got to $365 which is up 36% on PCP which is including all enterprise customers in the Loadshift division, et cetera, just did across all projects it's the same stack. Liquidity is pretty insane. 56 bids per project up 30. The 722 entries per contest up 113%. A lot of that liquidity is driven by sort of AI assistance in bidding. People getting GPT write their bids and so forth. We have some work we're doing this cycle, which we run six times a year, two-month cycle for product development, to crack down a little bit on this liquidity. It's very, very good to have liquidity, but it's a bit too much, particularly in the contest side of the business.
It can be a little overwhelming getting so many contest entries submitted. We're going to crack down a little bit in the next couple of months. This is what the average project size looks like. You can see that really since 2020, it's kind of really getting, taking off. It's a mix of the ability to do higher level sophisticated jobs on the platform. There would be a little bit of inflation in there, but that's one of the components, and it's really one lever we can really move in this business is the average project size. For other businesses it might be the average order size. $350 today probably gets you about $7,000 worth of Western work. It really is, that has that sort of leverage.
Still, $350 is not a lot even for a small business who, if you run the U.K., U.S.A, Canada, Australia, etc., you probably have someone on staff of $45,000 a year. We're talking, this is a number that could easily figure out how to put a zero on the end of. We have some pretty exciting things coming that I think will put a zero on the end of this number. It is a leap that can really, really strongly move and far more than the effect of what you can do with funnel optimization and so forth. We're working on the funnel optimization and the retention as well. This is probably my most negative thing for the half, which was the acquisition. This is cash coming in from new customers in the first 30 days or 28 days.
You can see there, really up until Easter we had the number peaking above 25% year-on-year. Of course, that just flows through the business and ultimately hits revenue and ultimately it's GMV. Literally, I got in the plane to go do an investor tour around Easter and the number went backwards. There's a couple things that went on there. There was a Chrome, I think it was version, what, 135? Yeah, it's one of those. They had a Chrome update which reduced a set interval timer in the Chrome browser to basically not have a minimum trigger delay. It causes a bit of a locking up on a number of websites around the Internet. You can Google it or you can ask me for more information. That took a little while to debug and we fixed it.
This was over Easter where you do have a little bit of a pullback. When you've got a little bit thinly staffed, there's that. A couple other things. We obviously can see that the number recovered a bit after Easter above 10% year-on-year. It's come down a little bit again. It's trending up; in the last couple of days it's been trending up quite strongly. We will get this number back up again to higher than 20% year-on-year. This is the thing that if we had that number up there at 20% year-on-year, the NPAT would have had a much larger number on it. We're getting that. We will get it back there. This is the GMV. However, you can see the GMV is quite healthy in terms of the year-on-year numbers here.
You can see that we've been in green for quite a bit of time here and here's the delta. That's for new customers, that top graph. This is for all customers through the business. You can see that we're still hitting the greens but the key is really to get those acquisition numbers happening again in combination with retention. The three levers are customer acquisition and getting that firing again, which we will. The retention is inching up, quarter-by-quarter by percent or so at a time. I think we did notice it by about 5% last year. We just need to kind of keep that going, get the acquisition numbers moving and then keep pushing the lever on the average order size and we'll get to some good places. AI type of jobs going too. The business is going on growing really strongly.
We're actually going to issue a Fast 50 jobs report this week which will get into a lot more detail about that. You can see that the AI jobs are going up. I fully expect this to go 10x at one point. I have been very bullish on what you know, I can see what we do with AI agents on the platform. I'm seeing what customers are doing with AI agents and I guarantee you this is going to be the new web design. Every business in the world very soon will—the aha moment will happen. They'll realize how they can get it done and their phones will be answered by AI to take an order, process a credit card, make a booking and they'll come to Freelancer to get it done, the same place you get websites done and you get your apps built.
Pretty excited about where that's going. We've done some overhaul with brand marketing that will continue to improve. We continue to maintain a really good Trustpilot score, actually with the best on the planet onto the major marketplaces, and continue to maintain that. We've started doing the integration work, which is being paid for. The largest generative AI models in the world is 175,000 franchises that have at one point in time worked on this. Once we get this integration done, the ability for us to marshal users, issue them credentials, get them working, get them paid, will be significantly easier than what we've been doing up to today, which is very manual. There's quite a large number of projects we've been working on with one particular partner across not just the hyperscaler foundational models, but also quite a number of smaller engagements.
What we're seeing in the market is we're seeing that just about every major business in the world that's in the Fortune 2000 or G2000, what have you, that has proprietary data, are kind of building their own foundational models to basically provide a way to meter access to them. For example, if you're Bloomberg and you have access to all this financial data, you don't want it scraped by Google, you don't want it scraped by Meta, you don't want it scraped by Anthropic, you want to produce your own models and have some way of charging a tariff for access to that data. We've just opened an office in Bangalore. It's a small office at the moment because a lot of this work is happening through the BPO industry in India.
We did a six city tour in the half in India and that convinced us to open an office because there's almost infinite demand for this sort of work. Literally every BPO has this worker balance from the small scale up to the very, very large scale, and it's pretty exciting. We're very bullish in all of that for the field services as well. India is moving up the S curve and as by doing so everyone's buying air conditioners, alarm systems, computer equipment, networking is rolling out, satellite dishes, et cetera. There almost seems to be infinite amount of work for global coming out of India. That's another major reason why we opened office in Bangalore. Even this week we're operating on a very, very large job that wouldn't start till 2026. It's pretty exciting with one of the most exciting companies in the world. It's at scale.
We'll see if we win that. I'm very, very, very bullish and Mas, who's in the room with me here, is leading all of that. Let me tell you, get him on the phone with customers. He's a professional. He used to work on NBN and Telstra and what have you. He knows the ins and outs of field services. He knows how to get someone, anyone, anywhere with any skill set on a roof with all the certifications. You name it, he knows how to do it, just on a marshall it, and you do it very, very quickly. That's fantastic. I've talked about NASA and what we've done there. I'm pretty excited and bullish about that. I do think you can see some great things coming. We're already seeing a lot of task orders. I think Tricia last week applied for six or seven of them.
They got sprayed out in one week. There's a lot of activity happening on that front. As I said before, NASA is now representing the whole of U.S. government. We're getting task orders from every agency you can possibly think of in the U.S. government. This is obviously a massive upsize. We also poached the Head of Sales from the number two task order winner from NOIS2. That's Ed Wong, who has joined us in the last two weeks and runs our enterprise sales team now out of Vancouver, reporting into Andrew. I think we'll have a lot of firepower there on this front moving forward. Here are examples of certain things going out. We obviously continue to award prizes and have handed out $2.5 million U.S. for a gene editing task order, which I'm not sure the value of it now.
It was $6 million U.S., but I understand it's jackpoted a little bit. They've decided to award some additional prizes based on the quality of interest. It's pretty amazing. If I think back to when I started this business back in 2009, you had 20 categories of work. You put $50 in, you got a $50 website, and it looked like a $50 website. Now you work with the top, top end service providers. You've got everyone from Harvard, Yale, Princeton, et cetera. You've got the research labs, you've got professional scientists, you've got grad students, you've got CRCs, you've got hobbyists, retirees, you name it, hackers. You know, contributing to these prizes, it really is accessing 80 million minds of genius to solve and deliver moonshots, science, and technical problems.
You know, to think that back then we would be, you know, not just delivering on gene entering in the sense gene editing on the central instance of humans, but to do so in a way where they've jackpotted the by, by I think, I think it's $1.5 million. I have to confirm that. You know, a $6 million U.S. dollar challenge for gene editing because the quality of the work is so good they want to award more winners. It's phenomenal validation of what we do. We do plan now on opening up this capability to Fortune 2000 equivalent business globally. There's some exciting news to come on that shortly, and you know, working on all sorts of things, sustainable business models, etc. We're also working with government to help solve unemployment in countries around the world. We've done it in a few countries now.
Bahrain is going very well. 97% approval rate of the delivery of that, that particular program, Escrow. As I mentioned before about the numbers, revenue is doing really good. The GMV, GPB, and GMV kind of bounce around a little bit. It was almost a blowout quarter. If we got one of these deals across the line, it would have been a blow up quarter. As I said before, there was a big deal that kind of got missed and then there's a couple of the one got partially funded. We've got some amount of $10 million and there's a few other big $10 million transactions being set up. That number could easily bounce back up in a big way very shortly. We talked about Shopify, we're very excited about that. We do have a bunch of shopping carts lined up at the moment. There's a very strong pipeline there.
The big thing is we are a high friction payment business in that we have to KYC people. We have a lot of obligations as part of our 55 licensing jurisdictions that we operate in to ensure that we do all the checks around sanctions and politically exposed persons and check that the product is selling for an appropriate price and do modeling of the pricing in some circumstances and fuzzy matching and you name it. It is a high friction payment system. When you're taking a high friction payment system into a traditionally low friction environment like mass, mass commerce, you want to make sure that you're just doing the slickest possible, as modern as possible, and that support operations, compliance, and everything else will work, operate hand in hand to make it work, to give it the product experience.
We're chipping away at that, and we will continue to report on that. We're just going slowly because we want to make sure it's all good, that we have done our first golden transaction, and there's quite a number of transactions now that I'm seeing on the dash being set up at various stages. It's starting to happen. We also have a very strong pipeline as I mentioned before. Elliot and the team of integrations and August, maybe you can jump in and help me here because my memory, there's automotive. We found a great new use case for car dealerships. There's machinery, there's agricultural machinery, there's fuel credits marketplace in Europe. What else we got?
We've got some M&A marketplaces.
M&A has been particularly strong.
Yep, yeah. M&A in particular, numerous marketplaces.
Some of them domain names, some of.
Them heavy machinery, automotive. I think that's something more.
Yeah, so that's going very, very well. I think that's going to be a pretty good focus for us for the next. We've got three things we're kind of doing with Escrow right now in terms of product. One thing is we're doing as much automation as possible, preparing for potentially high volume coming from places like these shopping carts. We come from a relatively low number of transactions a day. That's why you kind of get this choppy GMV, GBV. We're looking to pull them out and diversify that and make that, making it less choppy by going into more mainstream. At the same time, allowing the shopping carts to sell high value items which they could never do before because there's no other payment system to do that.
At the same time we're doing that, we're also producing features for each of these different verticals so we can cater really well on these. Whether it's M&A, whether it's automotive, whether it's machinery, whether it's domains or IP addresses or what have you, there's a whole bunch of features that are needed to really make those categories shine. The third thing is really just an overall modernization of the infrastructure. They're really the three categories. Escrow is really busy and it's got. I think it's going to really start hitting its stride in a big way. There's a bit more chat about that in terms of the verticals and half the overall domain volume is up, down a little bit in Q2 because you get your Easter, what have you in there.
Overall we're just talking about the things we're doing in terms of improving the service and we will get it to 24/7 support very, very, very soon. I think it's this quarter we'll have it 24/7. In terms of Loadshift, which is just Freelancer for heavy haulage freight, this is the sort of stuff we move. That's a barge. These things can be quite complex in terms of the movement. You may need permits, you may need pilot, may need a police escort, you may need to take power lines down, etc. We move this sort of thing all the time, whether it's a barge or it's crane or whether it's heavy machinery, etc., and we really excel. The harder it is, the more we excel at it. We're just really consolidating on that and gradually going down the stack in terms of operational performance.
We get two all-time highs for the month in the first half, and you know, order act chipped up a little bit, which we plan on chipping up a fair bit more now we've got coin deploying. Water jobs tripped up a bit, delivered loads tripped up a bit as well. In the first hour, we got an average of about 2.2 quotes per job, and I think overall we get 8 for the business. Yeah. Really now it's, you know, the big focus is getting that award rate up, and the big focus on that is the calling because the use case is a little bit different for Freelancer. The people are on the truck driving, so they've got on hands-free on the phone rather than in a chat environment, which they do on Freelancer.
We really, you know, focusing on this, just getting it fully opened up away, and that obviously allows us to increase the order quite dramatically. You can check it out as live in the App Store. Just make sure you got today's version for the latest updates. It looks pretty good, and it's got a Zoom back end and so forth, and there's a bunch of other features we're going between now and the end of the year. We'll have, you have the ability at the moment to put a GPS tracker in, but we'll have inbuilt GPS tracking. We'll have a better carrier onboarding experience. We'll have more robust features around the unloading, loading of things on the vehicle and so forth in the half as well. We now also have managed to launch an ADR through Deutsche Bank in the U.S. The ticker is FRCY.
It's 100: 1 ratio to the assets as listed FLN. We did have previously the FLN CF ticker, which is still up and live, but FRCY is actually a proper U.S. security now. It's getting visibility in the U.S. share trading platforms, which was not available under the F ticker, which is an Australian share with a U.S. ticker. It was considered as a foreign share, and even though we thought it might get visibility in the platforms, it didn't. Now we actually are appearing in the U.S. share trading platforms. Obviously, it's trading at 100: 1, so it's around, I think, $16 a share U.S. at the moment, something like that. I think, you know, initial good start in terms of the trading ticked up a bit, and I think now our 80 million customers will be able to buy the stock.
We get queries every second day. Particularly hard if you're someone like India, you can't buy an ASX listed share. We know because we called every single major broker in India to ask them. It's very, very difficult and we think now we'll start to get visibility around the world, not just in the U.S. and not just with our customers and also with U.S. investors. In fact, in my last trip to Europe to talk to investors, there was commentary that would like to be able to trade your stock in the U.S. time zone, which now is available. You talk about, you know, the big story is about profitability. The first half impact of $1.9 million versus negative $1 million last half, revenues are 8% improved gross margins. You know the operating leverage is really coming through now.
Disciplined cost management across the board, pretty much just nudging up the costs where it matters. That drives the business. In marketing, we're nudging a little bit in terms of the staffing. There was a bit of an FX unrealized gain in that, but they tend to net out quarter and we obviously delivered a strong turnaround in cash generation. Positive operating cash flow, positive cash flow of $3.3 million versus an outflow in the previous, and operating cash flow tripled to $6.8 million, up 209%, driven by strong RE growth in the structurally leaner cost base. Cash outflows turned $3.4 million primary lease payments. We increased our share on Loadshift to 64%, spending $733,000, and cash went up despite that, spinning by 12% fourth quarter in a row and provides a strong platform for growth.
Looking forward, enhancing the customer acquisition by that cash coming from customers in the first 30 days, particularly on the finance marketplace for retention and the average order size, the average project size. Those three levers in combination, just chipping away at them, accelerating all the things we're doing in AI and we really are at the center of AI. Our freelancers are using AI and their skills are dramatically increasing as a result of that. If you're an average copywriter, you know, exceptional. If you're an average illustrator, you're now exceptional and so forth. We're working on some of the largest foundational models in the world and we currently do an integration with one of them to deliver RLHF and other services to that foundational model faster.
There's a bunch of other things in the pipeline with various customers around AI, plus all the AI features that are going into our marketplace. If you use our sites recently, you'll see that we've got AI helping you write your project description. We've got all sorts of neat little things that are on the site like helping you to get things done and increasing your productivity. They're being all very well received. At the same time, expanding our financial service offerings. There's a lot of cost savings as well as improved ability to take a payment, which we're working on and really drive the operational excellence of the business, platform reliability, calls, year performance, etc. As I said before, consistently achieve over $500,000 a month of operating profit, which the jaws are now starting to open. I will now open up to questions.
You may ask a question to myself or to anyone obviously in the room here. I'll remind you, I've got Neil Katz, the Chief Financial Officer of the business. We've got Andrew Bateman, who's the Vice President of Product Management and Enterprise. I've got Mas Mohammad from Loadshift, from Escrow, so if you could please open up. Oscar, if you could let me know of any questions. I know it takes a little time sometimes for people to ask because it might be a bit shy. Any questions you may also arrange at any time. One on ones you can reach out directly to myself at matt@freelancer.com or investor@freelancer.com to arrange them if you would like to do something privately. We trade in the assets, obviously as FLN and over in the U.S. as FRCY. Now any questions come in.
Yeah, we have a long motion from Ray. Ray says the demand for electrical power from large AI companies is becoming an issue. COP30, where the largest international commitments to climate change actions will be announced, is likely increased scrutiny. A shift to renewable energy could have a positive impact on Freelancer's brand. Does Freelancer have any plans to source.
Is power from renewables produced? We have no plans to source our power from renewables. No. Power is kind of de minimis in our expenses, to be honest. We just use, you know, in terms of our biggest power consumption, it's just whatever data centers that Amazon uses. That's basically it. Any other questions? I'll keep it open for a little bit longer because sometimes it takes a while for questions to come in. We also, in the half, won our 13th Webby Award, which is the picture on the front of the deck. The Webby is the equivalent of a Grammy or an Emmy for the Internet. We're very pleased that we've now got 13 of these over the lifetime of the business. We've been operating from 2009 to now, so that's 16 years and 13 Webbys. That's a pretty good track record.
Simon asks, can we expect the share price to pick up in the short and midterm?
The share price has been picking up over the short to midterm and I would expect, you know, what do you need for a share price to move? Fundamentally you need good financials. We've been focusing on the profit and we did an all-time record profit for the half, which was a big swing positive. I think that provides a good foundation for showing the numbers can flow through and, you know, now we just have to focus continually on moving the revenue numbers across all lines of this business. You can see there that we've got a decent profit and I want to maintain that profit and continue the profit. With good numbers, you know, the stock market in the long term is a weighing machine that depends on the financials. That's what we're doing, focusing on that.
We have a question from Greg Ward from Trafalgar Capital. This is a question for both August and you, Matt. Can you provide more detail? Re Escrow revenue up strongly 32% but GMV down 3.3%. How much is attributed to price changes you made last year versus mixed yield affected by verticals?
Yeah, the very high end GMV stuff we do, we do transactions up to $50 million. That's the record we've done to date. Some of those deals have been done at very, very skinny margins. I think the lowest we did it was 0.25%.
A big list in that ballpark.
In that ballpark, yeah. Some of those big GMV spikes you see up and down have been done and we won't be doing the 0.25%. Anyone's listening trying to get that, that will not happen again. That was a rare circumstance from a trillion dollar company wanting to do something. As a result, you can get a big mismatch between the GPV and GMV versus, you know, when the marketplace as a whole is doing 1.64%, you can get those numbers bouncing around a little bit. It is true that March of last year, I think it was, that we did make a few pricing changes. We had a pricing table that was going maxed out at $25,000. The way it worked was $0- $5,000, it was 3.25%, $5,000- $20,000, the intermediate is like 1.8%+ something or other, and then $25,000 and above it was 0.89%.
The problem with that pricing table, which existed until March of last year, was that we were giving account managers the ability to negotiate higher end value transactions. Back 10 years ago when we acquired the business, you didn't get many transactions above $25,000. Then we started regularly getting $50,000 transactions, $500,000 transactions, $5 million transactions, $50 million transactions. We blew by several orders of magnitude off the end of the pricing table, and quite a lot of random pricing was being issued which was just stupid, non-sustainable. That was fixed up at March last year. In addition, we just tightened up the collections from fees that are already collected properly.
We had things sitting around, for example, in our custodial account that hadn't had fees paid in long and we were still performing a service, a service that needed to be remunerated because it was us holding custody of an asset, for example, and so on. Overall, there has been some pricing in there, but that's well over a year now. We now, I think, we've got a very, very, very good pipeline of marketplaces, platforms, etc., which will start ramping up, start diversifying and hopefully reduce the volatility from the GPV numbers. Any other questions?
One more question.
Yep.
Matt, you are very confident about getting acquisitions back to 20%. Can you elaborate on this? Any changes in strategies or area of focus?
No, it's just sticking to our knitting, you know, ensuring that product quality is consistently high and improving. Andrew Bateman's now in charge of product quality. He's doing an excellent job in terms of a number of missions and just getting that on track because I think the drop backs happen in really two top two times, right. One is when there are holidays and you kind of have like these periods where it's Ramadan or it's Eid or it's Easter or it's holidays or you have drop backs due to liquidity effects. The other is when you have some product issues where you ship something and it's a very complex business in that you have two sentient beings on the client side and on the Freelancer side trying to maximize their value equation. It's a bit different from selling a book on Amazon where the book is not sentient.
You do have the second order and third order effects. You have to be very, very, very careful when you make changes. The reason why I'm confident is I know if we just stick to our knitting and just really focus on product quality, these numbers just rise naturally and they flow through. With all the efforts we're doing today, with all the sticking to our knitting, continually looking for opportunities in terms of where we can acquire new customers, continually improving features, improving convertibility, improving size and stuff. I know we can get the numbers there because I see how the business reacts and I know what causes when the numbers go backwards.
Really it's just a pathological focus on product quality, I believe is really the overall arching thing we have to do at the same time as pushing out new features, but doing so in a way which is not move fast and break things, which I think was probably one of the most damaging statements for Zuckerberg to ever say to the tech industry. Be very careful and make sure the quality is exceptional and you're pathological about the quality of things you ship and I think that's been our bugbear since 2017, 2018 to be honest. Great.
We have another question for Matt and August regarding Escrow from Greg Ward. He says how long will it take for you to get best in breed checkout for Escrow, and will we hold off on onboarding new Shopify customers until we reach, say, 90% targeted best practice checkout?
Do you want to comment on that, August, or?
I think there's sort of a lot of opportunity to improve our checkout experience, and we are carefully designing a flow that we think is going to be world class. Exactly how long that will take, I think is a matter of iteration. I'd say optimistically we'd like to get a version out, let's say in.
The next, at least a new duration.
Out the next six months and iterate from there. In terms of exact, you know, bearing around percentage of quality we're looking for, I think that's something that's hard to pinpoint. We'll have to wait and see. We'll look at the market reaction, the customer reaction, and gauge from there. I'd say, you know, an 80% ballpark sounds pretty good, but again, you know, we may accelerate that.
The funnel issue is that we have to do KYC and that's the problem. Right. When you use your credit card online, people forget that they KYC once by going into the bank and setting up a credit card, which is actually quite painful in many circumstances, and actually having calculated your IDs and said the other. With a transaction, you have to potentially get the buyer and the seller, potentially third party brokers, all to KYC at the same time depending what you're selling, and that experience people don't like. We've got to get that as slick as possible. There are things like electronic KYC, which we have in the system, but that does introduce new classes of fraud where people can buy databases of people's Social Security numbers and so forth and they will, they'll work. You've got to be careful.
We've got to make sure that we're not fully 24/7 just yet on the support team. This quarter we'll get to 24/7, but we're not there yet. We've gone to 24 by 5 and a bit at the moment, almost 6. We have to get that running. We have to get the full automation of quite a number of things across things like compliance payments, because we're funding the business of being paid quickly. We do have enhancements, a part of a big banking effort that we did recently where we do now in some circumstances have the ability to take instant payments. I think that's a big one which will enhance that.
When August is six months, that's not just one iteration and the big dump and you've got a new product update, it is just chipping away at operations, compliance, automation, products, customer experience, merchant onboarding experience as well. I think the biggest model to look at is Afterpay. They had two parts of the business. They had one which was an aggressive sales and marketing team going after the merchants, and then you have a second thing which is they had an activation team that really onboarded those merchants and actually it became a bit over the top but almost made it like a celebration that you were part of the Afterpay community, which is a bit strange for something where you're borrowing money, so you know, effectively lay-by an item. We need to have all that infrastructure built and operating well.
We've got a new Head of Account Management out of Vancouver now, Antonio Yan. He is building that onboarding team as we speak. It is more than just product, it is operations, it is support, it is compliance, it is account management, it is onboarding, it is the product itself, customer experience, and it is happening and it is just slowly nudging forward. You only really got one shot at a business of this size and you want to make sure it is as good as possible. The end-to-end production, flow of funds has started. It is now just careful ramp up, you do not want to hit it and then kind of have the transaction volume go up, you know, 10x and not be unexpected for that. We are just being really cautious.
Got a question from DL just now on the AI category. How are they affecting new client growth, volume growth from existing clients, and the average project size?
AI categories started obviously at zero originally. I don't think they are a major contributor at this point to average project size or overall mix. I do expect them to get to a very large number. By way of comparison, software as a category as a whole is about 30-something, low 30s, percentage of the business that we have. I expect AI as a category to get to probably at least that over time. I think when it starts really moving, it will move very, very rapidly. I do think, if you think about the transformation of businesses thanks to the Internet, you've had really three waves. The first wave is 1994. Only the geeks had an email address. 1995 was the year that your grandmother had an email address. That led to an explosion of businesses trying to reach customers over the Internet.
They did that through websites and that led to web development as a category which became a very big category on Freelancer. The second big transformation that happened was smartphone rollout, Android versus iPhone. You reach customers through mobile. That didn't become as big as website development because we figured out ultimately how to make websites responsive, and so you didn't need to have a separate app. That's a very, very big category on the site today. The third big category I think will come is AI development, which is basically just, even just starting with AI agents that will answer the phones, take a credit card, process an order, make a booking. I think every business in the world, large and small, will have that happen.
You're starting to see it creep out with things like everything from, like McDonald's going through the drive-through sometimes now in locations they've got a multimodal AI agent taking your order. You're starting to see it creep out over phone calls. We've got it in trial, even doing outbound calls to customers in production environments, talking to customers to do things over voice modality. We have a multimodal experience coming out right now in the next quarter for the post project, post experience. I think it's going to be slowly, slowly then all at once in terms of the volume coming through with AI. It is building very, very strongly. It's building from a very low base. I do anticipate it to be a very, very large category. Great.
Final question. Are the average project sizes from AI projects currently higher or lower than that?
I will have to take that on notice. If you want to send me an email at matt@freelancer.com, I actually get the actual data on that. We do have a Fast 50 report coming out this week. Maybe I'll tell Brent to include that number in the Fast 50 report. Robert. Sam.
Okay, great. Final question from Greg Ward. If Shopify and potentially some other marketplaces embrace Escrow as payments platform, is it conceivable that Escrow's revenue will surpass Freelancer? If this does crystallize, how is this likely to influence the board's strategic thinking of the group? Regards to CapEx allocation, ownership structure, etc.
Yeah, yeah. I mean we routinely talk to customers who will quote us volumes for Escrow where they could see us ramping up to $100 million, $200 million, $300 million of volume with them. When you look at a business like Shopify and then you realize that while there are 100 payment methods in Shopify, none are for large value payments. This is actually a capability to Shopify to do things like sell real estate or whatever it may be. You know, I've always said this, I could easily foresee a business that would, you know, one customer that could potentially do more volume than we're doing now in aggregate across all of Escrow. In fact, in the past for a very large trillion dollar marketplace, I've been asked to quote pricing for volumes that have been larger than the entire marketplace as a whole.
We will see where this goes. That having been said, we're also quoting on very large things for Freelancer as well and also we've got a large quote for Loadshift that we're putting to the right and have quoted on some very, very big business for Loadshift. We will see. Each business ultimately may have a somewhat more independent future, and certainly I have been pitched many times, very early stage, non-affirm random possibilities for the subsidiaries and we'll continue to be pitching them in the future. Who knows what the future is of the subsidiaries in the future and potentially it could be something transformational but there's nothing currently on the books for the end of this. We're constantly thinking about that.
Yep, Greg says thank you very much. I think as we're now past 10:00 A.M., if there's no final questions, we'll close it.
Thank you. Thank you. You may follow up with myself at matt@freelancer.com or the investor team at investor@freelancer.com to arrange one on one. Thanks for joining and I'll see you otherwise in next quarter's call. Thank you.