Thank you. Hello and welcome to the Freelancer Limited first quarter of 2026 business update. My name is Matt Barrie, and I'm the Chief Executive. Here with me today, I have Neil Katz, the Chief Financial Officer, Andrew Bateman, the Vice President of Product, Tony Yang, the Director of Operations, and Mas Mohammad, Director of Loadshift. As always, you may direct questions to either myself or any of the management team in the Q&A. Moving on to the first quarter business update. Group GMV was up 14% to AUD 263.3 million in the first quarter. Freelancer was down 14.8% at AUD 28.4 million. Escrow was AUD 234.9 million, up 18.9%. Group revenue was down at 9.4% versus the first quarter. Freelancer was AUD 8.7 million, down 19%.
Escrow was AUD 3.5 million, up 18.9%. Operating profit excluding FX translation, which is a reflection of the operation of the business, was breakeven for the quarter, slightly positive, and Escrow achieved and quite a healthy all-time record operating profit before FX. About 10%-11% was a drag on the business, due to the Australian dollar versus the U.S. dollar hitting a three-year high on a very quick move, so if you look through that lens. Our operating cash flow was neutral versus last year. Cash flow itself was AUD -1.7 million. There are a number of one-offs in there, including the second large payment for a buyback of 10% of Loadshift. We have one payment remaining in the next quarter, and that's it, which takes our stake to, I think, 73.4%.
We had a couple of one-offs related to the closure of our Buenos Aires office and move to our new premises here in Manila, including deposits, get out, and so forth. I'm here at the moment with Tony and Mas in that new premises. The unaudited net profit before tax was positive for the quarter. The cash was AUD 20.8 million at the end. Again, noting that we had a buyback of Loadshift shares and a number of one-offs and so forth. The Australian dollar's appreciation to a three-year high against the U.S. dollar of 10.7% created a material translation headwind for companies predominantly USD-denominated revenue. Underlying cash flow is expected to improve in coming quarters, benefiting from the non-recurrence of one-off items related to the Buenos Aires office closure and a minor lease transition alongside the recent cost initiatives.
We're working across the group to build our own services. We connect over 90 million registered users across the globe with a seamless ecosystem of labor payments and freight solutions through our core business. Freelancer is the largest crowd workforce in the world and the most liquid. Escrow is the world's largest online escrow company, which facilitates and secures large value payments, and the only multi-jurisdictional online escrow company that's licensed in many jurisdictions outside of real estate. Loadshift, Australia's largest heavy vehicle load rate marketplace, and these services meet the everyday needs of consumers with large and profitable scale. On Freelancer, the first quarter was a challenging quarter for us. The segment ex-Loadshift delivered a modest operating profit for the quarter on a reported basis, supported by a favorable FX translation and underlying result reflecting revenue pressure, partly offset by continued cost discipline and optimization to the business.
The platform added 1.3 million new users, 150,000 new projects in the quarter. Several quality indicators improved meaningfully. The average project size grew 20.4% to $414. Bids per project grew 4.1% to 51, so on average, 51 people bidding on it, and the contest entries grew 72.4% to quite an incredible 862 entries per contest, which is quite an incredible experience if you try contests. That's driving up the average project size over time. Now, in terms of our paid acquisition channels, they grew 5.5% of the pcp, which is CAC coming in from new customers the first 28 days. Return on investment remained stable in that channel in a slightly more competitive landscape, just demonstrating the underlying demand intent remains strong.
I threw a graph in here this time around the award rates and the milestone rate over time, and you can see that they've been lifting over the year as we continue to roll out product improvements and all the things we've talked about in previous quality updates. However, we had a group of key factors weighing on performance. In particular, the first factor was resolved for I think over 70% of the issue, and that is we dramatically looked at the security of the platform in the fourth quarter of last year. We basically enhanced bot prevention controls and expanded the role of two-factor authentication. That's materially reduced fraudulent account activity, and pretty much stops it in its tracks and strengthened platform integrity, but introduced additional login friction that's weighed in on activity for returning users, most notably long-tenured cohorts with legacy account configurations.
We believe that this has been the biggest source of drag in the core marketplace this quarter, and it also affected us in the fourth quarter of last year. We've been iterating on these controls rapidly. We're quite on top of them, I think, and refining the authentication experience. Security is a bit of a trade-off with usability, and you can see here, I've put a graph in here, the success rate of two-factor authentication from users returning with the same browser cookies. That excludes bad actors, for example. You can see that when we did roll that out, that there has been an impact in the success rate of users kind of returning to the site. It does have a little bit of a cumulative effect as well.
You can see there as we've iterated, that we have managed to get the accounts with larger tenures back onto the site. We've iterated with various whitelisting and device tokens and other forms of convenience, et cetera, but you can see there that there has been a dip in the successful logins from accounts. Now, we believe it accounts for about 75% of the issue. We can talk about it later if there are questions in the Q&A. We also had a little bit of a challenging time with the organic search. It has been a little bit of a headwind since late 2025. Substantially all the site traffic now is being migrated to a content delivery network caching infrastructure because our site gets crawled quite dramatically, particularly from AI training scraping agents.
There's a benchmark known as SWE-bench, which these AI agents are all trying to get their software to outperform in doing work, et cetera. The scraping has over, particularly over the last year or so, caused us some issues. We have done some enhancements to how we deliver, updated all the outlook, our pages, and we have pushed it to Fastly and so forth and various caching layers. We believe we're pretty much over the hill with that now and quite substantially. As part of this remediation, randomly, what was applied to Google crawler traffic, which inadvertently affected legitimate indexation. Basically, we were 429ing Google because they were crawling too quickly, and they used the same crawler for their AI scraping as they do for their service, which contributed to a decline in indexed pages and overall organic referred traffic.
Potentially some impact as well from AI-generated search summaries. Now, we do have an exceptionally strong domain footprint. Our domain rating is 86, supported by 67 million backlinks and 37,000 referring domains. We're in the top 0.01% global domains, and we have been working on technical fixes and they've delivered incremental recovery, but not back to where we were in the previous year. Consistent with our previous SEO strategy to reduce reliance on organic search and really focus on things like AI search and so on. We're responding with three work streams. First, to improve the technical crawlability, indexation, schema markup, and site architecture, including remediation of low-value auto-generated URLs. A content program targeted at high-intent long-tail project categories to capture commercial search demand and invest in answer engine optimization, generative engine optimization to surface us on LLM discovery platforms such as ChatGPT, Claude, et cetera.
We are starting to get traffic from those sites as these grow as an alternative to traditional search. There was a minor transient impact from Recruiter. We moved the Recruiter team, which is a somewhat global team, but the headquarters primarily for that from Buenos Aires to Manila during the period. The team have successfully lifted the award rates on the projects that they manage, and in fact, they're doing very well on that. What they have done as part of that is they've had a temporary reduction in the complimentary upgrade to Recruiter, that gets applied based on various project parameters. We're missing a bit of the free volume, which has a little bit of a minor GMV impact here, but we're pretty comfortable gaining that back in the second quarter as we build the team out.
In addition, to accelerate execution against these priorities and a broader product roadmap, we've appointed Andrew Bateman to lead product for the group late in the quarter. He brings extensive experience leading product organizations or driving space for platform innovation across the marketplace. Primarily, it's not the best quarter. We are on top of what we think the primary issue is. The overwhelming majority of the issue is we are iterating on it. You can see in this graph there's been improvements, et cetera. We will continue to work on that moving forward, but we're fully engaged on the issue. On product. Our product investment in first quarter was directed at the areas most likely to drive recovery as volume returns, particularly improving how clients and freelancers find each other, communicate, and transact. A major focus this quarter was improving the quality of the bidding experience.
This is a long-term feedback we've had about the site is while we've got lots of users. In fact, we're the biggest crowd workforce in the world, but it can be sometimes a bit confusing. Sometimes the users are using AI to generate the bids, et cetera, and it can be hard to get the signal from the noise. What we have done is implemented a new bid-ranking algorithm, delivering a lift in overall award rate and accept rate. I believe it's at 6%, and especially strong results for new clients. This is a pillar of a broad initiative to leverage LLM search capabilities and annotation and so forth across the marketplace.
It's an example here where now what we will do is we'll go through all the past activity, and particularly paper reviews is a very strong signal, and we'll surface work that's relevant to your job, and we'll annotate. That experience is actually quite phenomenal. I encourage everyone else to use the site to give it a go now. I think you'll note that it is a big uplift, and we've got a number of enhancements coming through in the next quarter. We have also implemented policy changes to limit the bidding pool.
I think one comment we've had back previously as well is, "Well, I know 50 people bid on your site, on project on average, but I don't really want 50 I'd be happy if I got 25 that were great." What we did was we put in some barriers for poor quality or low-skill bidders to appear on the bid list. It had no negative impacts, and in fact, lifted the award rate overall. Again, back to that first graph that I showed up here, Figure 2, you can see there the award rates and the milestone rates have been lifting over the last period. In particular, this year, they've lifted quite well. Those initiatives are actually delivering for us, and we get the traffic lifted at the front end, at the back we should be in a very good place. Additionally, we continue to expand with our AI capabilities. We have a full LLM-powered whiteboard called Prototyper. You can actually write full software there as if you were just writing on a whiteboard.
We had a bit of a competition internally in the last couple of weeks. Someone built Minecraft on a whiteboard, and you can hit a button, make it real, and end up with a fully playable game just by writing on the whiteboard. That's great, and they actually won an Xbox Series X award for innovation in artificial intelligence. Something we also just launched is Freelancer manufacturing. Anything that you design on Freelancer , you can now hook up with one of our team who will organize manufacturing for it, connect you with various manufacturers, get you quotes, get you samples, and also head into full ongoing manufacturing with an exciting new product that's being launched by Astro. Automation of support functions is actually going to accelerate in the second quarter. We're pretty excited here with the initiatives that we have.
We've already got an agentic framework on the platform that's doing support operations and sales. That platform now will be extended in the next quarter to tier one support, which is the first line of general support for users. We believe over 50% of engagements will be handled by tier one. Now, the AI, it will escalate it to tier two human support, et cetera, if the problem's complex or it needs to. We're pretty excited about that, and it's allowing our staff here in the Manila operation to be elevated into a strategy and operations team to do a number of other initiatives, which I'll talk about later. Furthermore, we got a pretty phenomenal breakthrough this quarter with our in-house autonomous coding agent called Diffuty. This is an autonomous junior software engineer.
It's now capable of processing increasingly complex tasks one shot, so things like small UI changes, bug fixes, creation of landing pages, et cetera. It's pretty phenomenal. This will allow us to increase our throughput significantly, allow us to burn down our backlog of low-value tickets quite significantly, focus our engineering on the high-value delivery, be able to deal with customer feedback or customer issues very, very quickly in terms of little UI issues or so forth. It really should allow us to increase the velocity by which we can improve quality on the platform primarily, because we have a big backlog of little issues to improve all the time. That's always very hard to get on the OKRs for engineering teams when they want to work on the big shiny new things.
That combined with quite a robust and extensive quality program that Andrew Bateman's already put in place, that we're rolling out progressively through the teams. The remainder of 2026, looking to see continued product focus on improving and streamlining the core experience of the marketplace. We've got a pretty special contest experience, this quarter that will be delivered that I think I'll be pretty keen to show you in the next update, which I believe will be better than AI in terms of getting various contest deliverables to you. I think you'll be pretty excited about seeing what we're doing there. We have new product leadership in place, and we'll be directing efforts towards high-value revenue opportunities. We continue to rank number one in the world as the most trusted platform for work with 4.4 excellent Trustpilot, 4.7 on Sitejabber.
That's from a total of about 40,000 reviews between the two of them. Our AI-enabled division delivered commercial momentum in the first quarter as well, converting pipelines into signed engagements and expanding client base across AI data services, field operations, and managed workforce solutions. We also progressed high-value opportunities with market accounts across digital platforms, customer experience, AI data services, contracts, and pilots at various stages. For our global media provider, you'll see that if you go to our social media or you log into the site, we've commenced sourcing for over 1 million images globally, using our workforce, going out to take photographs for training for the visual system for one of the large AIs. We're recruiting at scale for that right now, and you'll see a bunch of stuff on our socials.
For an AI data training data specialist, we're executing multiple AI data collection projects spanning healthcare, conversational AI, armaments projects, and computer vision across 60 countries. For an emerging AI data provider, we're executing a speech data collection project to rapidly expand across additional data modalities. For a European headquartered telecommunications infrastructure provider, we're onboarding at the moment with project commencing shortly. In addition, in the media division, we have the strongest pipeline to date. We've got two field services engagements for payment systems, ability to field sales teams at scale, multiple AI data service engagements into advanced commercial discussions with several large cap enterprise accounts. Near-term priorities are converting the digital payments pilots. There's two now, progressing six and AI data service opportunities into signed contracts and scale of delivery volumes across active engagements.
The Freelancer Moonshot Innovation Program, which delivers scientific and technological breakthroughs for Fortune 500s and equivalent organizations, we've now commercialized that and we launched to corporate clients. We had a landmark moment in the first quarter where NASA's Artemis II crewed lunar mission flew with the fifth astronaut, which was Rise, which we sourced, which was the zero-gravity indicator, which we sourced for them through a global contest across 50 countries. The purpose of this is to teach kids about space. It was designed by an eight-year-old, Lucas Ye of Mountain View, and Rise carried a microSD card with the names of every freelancer who registered on humanity's first crewed mission to the moon in over 50 years as the fifth crew member. They got a lot of coverage for us, which you can see. That's the breaking news with the freelancers that powered NASA.
The program will continue to expand its footprint across high-impact missions. We worked with the United Nations Development Programme to crowdsource tools for the affordable detection of underwater explosives. It's a global humanitarian challenge looking to clear out mines, which will be quite useful at the moment in the Strait of Hormuz. We also had our first commercial partner launch a physics-based AI modeling challenge to model the distribution of particles from explosions, could be asteroid impact, structural collapses, volcanic activity, et cetera. Also the National Institutes of Health, we continued with a major engagement around genome editing and health research data sharing. Trisha , our Director of Innovation, was also invited to present on live innovation by the UK Space Agency at the United Nations Development Programme. Quite eyebrow-raising sort of work.
Again, if anyone works for a Fortune 500 equivalent organization, give us your hardest scientific and technological breakthrough. We'll solve it for you. If we don't, you don't pay for the price. It's quite a compelling offering. In FY 2026, the key focus of Freelancer will be to obviously get on top of the transit issues we're dealing with right now, which we talked about at the beginning, enhance the marketplace engagement, continue improvements to the user experience and matching capabilities to attract, activate, and retain high-quality freelancers and clients. Let me tell you, in the next quarter, it's going to be a pretty amazing product. Desktop and mobile, projects, contests are going to be very slick and I think we will have a very exciting platform to show you with all the work we're doing.
We're also going to accelerate AI-driven innovation, expand integration, advanced payment solutions across product services, enabling efficiency, automation, and new opportunities for enterprise growth. Expand our financial service offerings, broaden and streamline our payment methods, our financial structure, and improving transaction needs, security, growth, scalability, and drive operational excellence to try and achieve at least AUD 500,000 per month operating profit consistently on an ongoing basis. Now, the product Escrow. Escrow had a pretty decent quarter. Its gross payment volume was AUD 238.4 million, up 18.9%, driven by strong year-on-year growth from general merchandise, domain name, and also IP address, IPv4 blocks. The revenue for the quarter was AUD 3.5 million, up 18.9%. Escrow enters its sixth year of profitability, and the first quarter represented our second highest revenue quarter since inception. We hit a milestone of surpassing AUD 8 billion in cumulative transactions secured.
I can say as of today, on a rolling three-month basis, second quarter looks like it will be better than the first quarter. Escrow is committed to bringing digitalization and scalability to processes originally designed around manual human-led workflows, building on the strong foundation, financial foundation of the business's decision to make meaningful investments in product and services to leverage scalability through the remainder of 2026. You can see here there's a 26-year history of the business, and you can see there's a strong trend line there. As mentioned, it was a good quarter for us, and it will be a better quarter, I think, at this point, it looks like it in the second quarter, and I think we'll have a great year.
As part of our platform innovation efforts and digitalization and scalability efforts to continue to strengthen our payment infrastructure to deliver meaningful gains and accelerate payment processing and real-time payment support. In the second quarter, we're launching separate bank account numbers for inbound payments. Just like Wise, you get automatic matching based upon that, separate account ledgers for separate users. We'll have real-time payment support. There's RTGS, which is a real-time payment system in the U.S., for instant crediting, et cetera. We will have flexible payments, flexible transaction structures to be able to set up transactions that may have milestone payments, balloon structures, what have you. It can be quite complicated in terms of what you can do. It's going to be a pretty phenomenal platform, as well as a UI overhaul.
During the third quarter, we completed several new integrations with e-commerce partners and achieved execution of our go-to-market strategies from partners to deliver successful technical implementations to drive sustainable adoption across their user base. Our activity continues to build a pipeline of net new e-commerce merchant partnerships. In fact, both our high-value account managed pipeline, and Tony manages that here in front of me, is the best we've ever had. In addition, our sales pipeline is also the best we've ever had in terms of platforms. The execution is very well and cohesive in the team as a whole. To support the expansion, the merchant acquisition team is being expanded with a dedicated go-to-marketing team responsible for pipeline development, merchant activation, and post-activation to drive sustained transaction volume from a partner network.
We're trying to do what Afterpay did, which there's two parts of Afterpay. There was the direct sales team going after merchants and platforms, and then there was an activation team that once they came in, you had the developer support, marketing support, and so forth. That's what we're building now, that second piece. In terms of new verticals, we continued to deepen our position in high-value B2B transaction categories during the quarter with meaningful progress across global trade, secondary electronics, and IPv4. In global trade, we've got quite a number of things happening. We've extended our platform to support complex cross-border transactions. We have an exciting, I can mention it because it's somewhere out there now, a digital letter of credit product, to supplant or replace traditional letters of credit, which are clunky, bank-driven instruments which have a 60%-70% rejection rate when their trade gets settled.
We have a digital offering that we've launched, and we already have our first big partner with a live transaction signed on that, which is a substantial transaction, I will note. Substantial. It offers a far more accessible alternative to traditional trade finance instruments. A newly signed partnership with a global shipping and logistics company will deploy this capability across a series of international trade deals. In B2B electronics, the vertical continues to mature. We've signed new partnerships with TradingBlue, BrokerCypher, and The Broker Forum, all recording live transactions on the platform during the quarter. Most notably BrokerBin, which is the world's largest B2B electronics trading platform, does $100 million a month of volume, has finalized technical integration with Escrow that jointly announced, already released to market. Stage two of the launch is scheduled for BrokerBin's annual member meeting in Croatia.
It's happening as we speak, and we are there at that event, where Escrow will be introduced as a partner solution to BrokerBin's network of over 10,000 member companies, positioning Escrow to capture a significant share of the secondary electronics transaction volume globally. Yeah, we are in Croatia today, with BrokerBin, working through at that event. We've had a good period with IPv4, and we've done blocks in the past up to $50 million in size. We've had a pretty decent period for IPv4. We have a partnership with a leading APAC brokerage bringing seven-figure deal flows to the platform. The relationship with existing partners also deepened further with establishing Escrow as the trusted payments over IP address trade. We'll take quite a number of precedents.
We announced the Master of Domains Award, a prestigious award that's recognized as the industry award, which ranks the top brokers across the entire space. This year's winner was Andrew Miller from Hilco. We've had a long, and quite a fruitful relationship with Andrew. He is now the number one domain broker in the world, breaking a seven-year streak by the previous title holder. He also came public with a few large transactions just recently, for example, clock.com for $10 million, that went through us, and several others on our platform with Andrew Miller there. We also continue to engage with the media and key go-to-market initiatives while tracking industry trends through our proprietary domain investing index.
In that index, the .ai domains are starting to take off. Reached a new quarterly high of AUD 10.3 million in total transaction volume through Escrow in the fourth quarter of 2025. The emerging niche represents an area of strength and meaningful growth opportunity for Escrow in the broader industry. In addition to domains, we're also going to publish imminently an IP address index, so that's going out as we speak. The index has been written. We're just doing some video collateral to go with that. That's about to go out the door. We also attend a number of major industry conferences, ICANN, NANOG, et cetera, really the center of these industries because we're where the transactions get done. We had a decent quarter for domain name volume, up 19.6% year-on-year. Valuations continue to rise.
There's obviously been some marquee valuations in the space recently where these domains have sold for tens of millions of dollars as they are the commercial property of the internet. Beyond, we aim to deliver new avenues of value to the domain space with expanded offerings and streamlined service delivery with regards to customized features around that. In terms of the customer experience for Escrow, we implemented enhancers to automate backend payments, customer success workflows. Tony's got the team running 24/7 now, which is fantastic. It's one of the reasons why he's here working with the teams. He's also working night shift.
Continuing the front-end migration to the Freelancer technology stack that go live in the second quarter, which allows us to get all the benefits of chats and a modern web front end and AI agents and all the other things we've got, which are great, as well as the mobility of the engineering between the teams, as well as Diffuty autonomous software engineering, which I talked about before, which is absolutely phenomenal in terms of what we can do with that in bypassing engineering without even the QA teams directly run autonomous engineering. Heading into 2026, the operations team is building enhanced internal systems, and external transaction support channels based upon proven VIP client practices to benefit all users.
Payment processing improved with new banking partners to improve automation, simplify funds transfer, and new compliance initiatives to leverage AI and new licensing advantages across 55 jurisdictions to simplify experience for brand new system partners. Very quick on Loadshift. Loadshift is Australia's largest heavy haulage freight marketplace, delivering a single digital platform that connects freight runners directly with a nationwide network of verified carriers. Matt's in the room with me. He's the guy who runs that if you want to ask me questions about it. The technology behind them secures the right capacity at competitive rates for everything from palletized goods to oversized industrial equipment, eliminating broker margins and legacy inefficiencies.
The platform provides end-to-end visibility with real-time tracking, integrated communications, and seamless transaction management, cutting costs and boosting reliability for shippers, while giving carriers a steady flow of quality loads, high asset utilization, and faster payments on metropolitan, regional, and remote routes. We started Q1 strongly with Loadshift, with March delivering the second highest revenue month on record. It was just short of an all-time high by how much?
AUD 1,700.
AUD 1,700. We aim to hit record highs in the rest of the quarters this year. The quarter finishing well ahead on the year-on-year revenue and GMV both grew year-on-year to provide marketplace improvements. In the first quarter, our revenue was up 6.8%, GMV 5.7%, March was up 11.3%. I believe that will accelerate over the course of this year, and we'll talk about in a second why. The platform continues to demonstrate operational improvements. Job postings increasing quarter-over-quarter up 18.3%. Award rate softened slightly as fuel price fluctuations drove quoting levels higher, resulting in increased job repostings. The award rate is 26.8%. Total jobs are up 9.6%. Delivered loads up 11.6%. Loadshift enjoyed its most successful quarter for earned media exposure. If you're in Australia, we're pretty much in the news every second day or so. We've become the authority, I think, on this national crisis in fuel.
We even had a hit from the BBC worldwide on it. We appeared in more than 35 articles across major domestic and international outlets. Most of them unique journalist-driven coverage rather than press release pickups. As Australia's diesel prices deepen, exposure for Australian haulage newsrooms including Daily Mail, Sky News, ABC Australia, Nine News, The West Australian, 2GB, BBC, and others turned to Loadshift for on-the-ground insight and impact on freight operators. We also had press in the U.S., U.K., China, and New Zealand. Major Australian international newsrooms now turn to Loadshift as an authority on freight and logistics and what's happening on the ground, and that's going to continue. In fact, we're about to launch a national hotline imminently, to support that.
We continue to invest in platform capabilities for the first quarter, delivering enhancements to improve marketplace efficiency and transaction completion rates. In-app calling adoption continued to build, with March recording the highest monthly usage since the feature launched, up 40% month-over-month. Usability improvements delivered, including mini call windows and so forth. The back of the camera now works. Development is now underway, extending it beyond the app and bridging into the public telephone network, which is a critical step towards keeping all negotiations booking on Loadshift, and that should be ready by end of second quarter. A new quote on behalf of capability was launched allowing our ops team to create and post loads on behalf of shippers. If someone rings in the phone not signed up, we can get loads out there quickly, and that's what we lead to an acceleration in loads getting out.
It's designed to convert inquiries, et cetera, whether it's through social operations, whatever. Adoption's been encouraging, and we're getting a lot of loads already. We also have an AI-powered follow-up system, which is also doing pretty well. Our mobile app install rate continues to climb. It's at 60% now, up from 46% last year. Really looking ahead, our product and commercial roadmap is focused on three key areas. Carrier onboarding phase one, aiming at streamlining the sign-up experience for carriers. We should have a big explosion in carriers getting on the platform. We've also got a very value-enhancing new feature to help carriers find loads, which is about to be launched in combination with the National Hotline. We also have PSTN calling I've just talked about, which should really let us get that onboard rate right up from where it is.
That's a very, very big lever, and this is really the last piece to be able to get that lever full. On the enterprise side, really exciting what we're doing with Loadshift. We've got quite a number of very large enterprise engagements now. We've got a major equipment hire and storage provider where we're building a curated, approved carrier network that will allow them to move things between their depots. Heavy haulage. We're about to go ahead with a trial starting in one of their states, and hopefully, we'll be able to roll it out nationwide, and that's a very, very big customer. Then there are a number of others around auctions. There's another equipment hire and also mining. Really excited about that. I think Loadshift's in a great place, and I think it's going to have a pretty mark year for us.
We reported a broadly breakeven operating cash flow for the quarter compared to strong operating cash flows in the prior corresponding period. Our results lower year-on-year is reflecting a softer quarter for the core Freelancer marketplace, partially offset by quite a strong performance from Escrow. Escrow had a record operating profitability before FX, as I mentioned before. They're in a very, very strong position there. Operating cash flow was also impacted by absence of paperwork and capital movements experienced in the prior corresponding period, including a reduction in user balance during the quarter. Cash and cash equivalents increased over the quarter, reflecting breakeven operating cash flows, ongoing lease-related outflows such as the shutdown of the Manila office, the make-good, deposits, and fit- out of the new office here in Manila. The third of the four installments, which is AUD 400,000, will be paid in the quarter.
The acquisition of 19% of Loadshift. We have one final payment left in the next quarter, and that's done. That's taken up to 73.4% of that business, which is, I think, great because I think Loadshift is going to be quite an incredible business in terms of its growth this year and in terms of its value to the group and in terms of where it's going to go. We've got some things we've been working on for a very long time that are now basically coming live. Full app mobile experience, bridging to the phone network, GPS tracking, the ability to know and route where the trucks are, et cetera. We're rolling out AI agent automation on a number of things to help logistics. That business is really going to start to shine. The great thing about Loadshift is it's just completely leveraged off the Freelancer stack.
It's a very lean and mean machine. Mas is a very lean and mean machine as well, I will say. There is quite extreme operating leverage, I think, in this business that's going to kick in soon. As I mentioned before, the cash outflows, there are quite a number of things that were non-recurring, relates to the office moves, fit outs, and one final payment left for Loadshift. The group remains focused on improving operating cash flow and maintaining disciplined cost management. I think we have a good track record of doing that. Cash flow expected to benefit from these non-recurrence, these one-off items and risk initiatives. As at 31st of March, we held AUD 20.8 million in cash and cash equivalents, remain debt-free, which is down 18% from the previous corresponding period. [audio distortion] .
In terms of group management, a few things here. We have appointed Mr. Dylan Carter as the Chief Financial Officer. He starts on the 18th of May, notionally, but he's been meeting with us every couple of days and coming into the office this, that, and the other all through his break from his previous employment. He's very engaged. I think he's going to be a phenomenal asset to the company. Mr. Carter succeeds Mr. Neil Katz, who will step down after more than 16 years of very distinguished service and remain in the company through to August 2026 to ensure a smooth transition. Mr. Carter brings over a decade of experience across ASX-listed digital payments and fintech businesses, most recently as the Head of Commercial Analysis at Tyro Payments.
I mean, effectively it was a 2IC to the CFO, where he played a key role in the company's IPO and subsequent capital markets activities over his six-year tenure. His earlier career spans structured finance and paper money and mergers and acquisitions at Ingenia Communities Group and treasury at Macquarie Bank. He holds a bachelor's of economics from the University of New South Wales, a member of the MAICD. I think he brings quite a deal of technical sophistication into the role, as well as very relevant experience, being effectively 2IC at Tyro, CFO. The finance product group, Andrew Bateman, who's in the call, you may address questions to him as well, has been appointed to lead the finance product group, succeeding Adam Byrnes.
Mr. Bateman brings significant experience leading product organizations to lead the next phase of platform innovation across the marketplace with immediate focus on accelerating, positioning in the product roadmap and high revenue opportunities. Mr. Byrnes moves to the role of Chief Scientist, leading the group's data science and analytics efforts. Tony Yang is here with me today, has been promoted to Director of Operations at Escrow, assuming the leadership of account management, support and payments functions. The promotion recognizes Mr. Yang's contribution to Escrow account operational performance in support of the business's continued investment, scalability, and product and service delivery. Tony is a hot hire. He interfaces with a lot of clients. He is rapidly organizing the teams to be coherent, cohesive, reduce internal friction, automate, and so on. Very pleased to see him in that role. That concludes the overview for the group.
If I can broadly characterize it, at a super high level, Escrow is doing extremely well and flying and spitting out cash, and it's in its sixth year of doing so. I think it's really going to accelerate this year. As I said before, the second quarter looks like it'll be better than the first quarter. Already, I can see that very clearly in the numbers on a rolling three-month basis. We've got some transactions we've put through, obviously, since the first quarter closed out. We've also got our digital letter of credit offering that we've launched, and we've got our first customer. I won't mention the number, but it's a very chunky first customer, and [audio distortion] . It's a partnership with DocShipper, which is a global import/export platform. We're going to announce that in a bigger way going forward.
It was for the import and export of ongoing monthly containers of certain food products, and it's an ongoing annual contract, and it's of substantial size. I think the partnership with DocShipper will prove to be quite fruitful, as will the partnerships with BrokerBin and all the other high volume global sourcing marketplaces in combination with our digital letter of credit product, and then in time, hopefully, with our financial manufacturing offering as well. There's a number of verticals they're building up and there's a marketplace launching in regulatory credits. I don't know if there's any others you want to mention, Tony. I'll tell you that there's quite a number, agricultural machinery, et cetera, and so on. The management, I think, are in a very good place, and I think we'd have a pretty standout year for that business.
Loadshift, I think, will also be a standout year. We've hit an all-time high in the first quarter. I think it's only going to continue this year. The key features we've been trying to get out the door have now come to fruition, an integrated audio-video calling experience, because obviously on Loadshift, people are driving a truck, so it's unlike Creator where it's more of a textbook chat. That'll be live. That's hooked into our full agentic system. It's hooked into our transcription system, our anti-fraud systems. GPS tracking, fully live. A bunch of enhanced product offerings and features to help drivers find loads. Along with that, some pretty amazing, and I think, what looks to be very promising enterprise integrations. On to Freelancer, it was a challenging quarter. We know what the issues are. It's primarily the friction with deploying enhanced security.
We did that because we noticed, we've got obviously 80-something million people on the site, and we do have every range of actor you could imagine interacting on Freelancer. It's a very well-known and well-trafficked business. We just needed to crack down on the bad actors on the site and enhance the security. Obviously, that comes with a trade-off in terms of friction, and we have been, as you can see, iterating on reducing that friction as fast as we can, but we have taken a bit of a hit on that. The underlying product experience you can see here in this graph has been improving, particularly this year. There's been some big improvements in the underlying put rate and underlying milestone rates.
It's just front of funnel work and there was, as I mentioned, in the fourth quarter, there's still a very strong pipeline of enhancements in the funnel, particularly the upper funnel. They've been trying to get out the door. They've just been delayed through some QA test issues, et cetera, and combination with this. We've got new product leadership that's going to jump on and really already has the bull by the horns and so forth. We have to work through that, and once we do that and we get the third leg firing as well, this group will be in a pretty amazing place. That's it for the update. I will now open it up for Q&A. Liam, if you may direct your questions again to myself or anyone in the management team.
You have Neil Katz, Chief Financial Officer, you have Andrew Bateman, the VP of Product on the call. You have Mas Mohammad from Escrow, and Tony Yang from Escrow. Liam, we may put questions in the chat.
Dean Fergie asked: How concerned are you that the demand for some of Freelancer's core project skills like design, coding, et cetera, will become redundant with the advances in AI?
Yeah. We've obviously got a lot of noise out there in the market from the larger foundational AI companies that AI is going to take everyone's job and we'll all be unemployed, and we'll be on universal basic income. They have to pitch these ridiculous views of the world because when you're raising money at a trillion-dollar pre-money valuation, you need to have a biz plan that will match that, to show that you're going to generate a return on those funds raised. They've generally trended into, well, we're going to take certain categories of work and all the world's graphic design or legal work or accounting is all going to be done by AI, and everyone's going to be out of a job, and the government's going to put them on Centrelink and give them universal basic income, and that's the future of the world.
That's complete hogwash. Now, if you want to know my in-depth, I've got a multi-part series podcast that I've done with MacroVoices, and I've also written a series of essays, which I'm happy to share with you, Dean, on this. The argument around AI is completely misrepresented by the foundational companies. It is a tremendous resource. It's great for personalization. It's great for funnel optimization. I talk about Diffuty, our autonomous software engineer, which can place the tickets out the door, et cetera. AI is a productivity tool, right? It's a productivity tool. It's a power tool, right? You need to be skilled in the art of the industry to really get the best out of these tools. Diffuty, our autonomous engineer, needs an engineer driving it.
In fact, what we've seen through our adoption of AI coding tools within the company is that the engineers that are using the AI coding tools the most are our top engineers. The top engineers are the ones that are using it the most. They're getting the most value out of it, and then you get a bit of a bell curve out the rest. My mom cannot go to Claude and build an app and get that app in the App Store. That is impossible at this point in time, right? Certainly, it's a skills boost across the whole basket, right? If you're an average copywriter, you can now be a great copywriter, et cetera, and so forth. Where the dislocation will occur, thanks to AI, will be in companies where you've got people doing the same role in workflows, where those workflows get automated.
It might be someone at a bank, and their job is to process paperwork and do certain things, and you can automate that workflow. Yes, you will have some dislocation. It's in the areas like call centers where you've got 10,000 people in a room. Maybe in the future, you won't have 10,000. You might have a few hundred. I know the rollout of AI in my support organization, which is 24/7, 365, it's actually just created more work rather than remove work. AI is now in the chat. It does tier one support, but it frees up the support staff to be the managers for the AI. There's a lot of work in QA-ing the conversation to get them authentic and in coming up with new roles. For example, there are roles like a freelancer induction specialist, where, "Hello, Mohammed, you come from Pakistan.
I see you're a web developer. I've looked through your profile. There's a few spelling mistakes in it. I've noticed that you say you've got Flutter experience. I don't see that in your portfolio. Maybe you should upload something. And by the way, do you want to give me a walkthrough of the site? Right? That role would never be economically viable to deploy with a human ever with 25,000 sign-ups a day, but you can now do that with AI. What it's doing is it's elevating people. Now, the elevation of skills, as I said before, the more skilled in the art you are, the more your elevation is, right? If you are a highly skilled designer, you can get really good things out of the design tools, right?
If you're a pretty basic designer and just punch in, "Get me an image to put in a brochure," the effect you get is what you've had over the last few years, where when Midjourney came out, every company around the world updated all their imagery to use Midjourney-style images. Six months later, everyone looks back and goes, "Oh, that looks like an iPod photo." Or Getty Images, right? Because everyone has the same style. It's only something with these core tools. There's only a couple of styles that a basic designer can really get out of the systems that easily. If you're highly skilled in design, you can get great outcomes. I think really the teams that have maybe 1,000 people might get reduced to 100.
If you've got junior lawyers drafting in a law firm and you've got 40 people in that junior legal team, maybe in the future you might have 13, right? The teams that have got highly paralyzed people doing workflows, et cetera, the teams will be reduced. If you've got one graphic designer in your company, you're not going to go and fire that graphic designer and go and replace them with someone else. The ultimate combination is humans with AI. It's skilled professionals with AI. We have the biggest marketplace of skilled professionals. Those skills have now been dramatically lifted. All our users use AI, right? Pretty much. They're all Claude Code powered. They're all Codex powered. They're all Midjourney powered, et cetera, and so forth. Also, we're seeing quite a number of rise in new AI jobs, it's about 5% of GMV at the moment.
I think it published last quarter on that. It remains about that level. New categories of work are being enabled, thanks to AI, including fixing all of the slot. All the people that sit there and try and buy the perks on the weekend, then realize how hard it is to actually turn that into a real app and get it into the App Store. They go to the freelancers to go fix it up. We've done some press on that as well. Look, I think AI is clearly transformational. I see, as I said before, it's a productivity tool. It's not a human replacer. There will always be work for humans. There's an infinite amount of work in the world. There's always things you get people to do. All technology in the past has pretty much created jobs rather than taken jobs on net basis.
Yes, there have been dislocations. I'm 52 years old. When I joined the workforce, it was on the cusp of coming to work, and there was no computer on your desk, and then you came to work, and there was a computer on your desk. I couldn't imagine coming to work today with no computer, no internet, no mobile phone, right? Try to remember, how would we do it over the telephones and run a business like this, right? Yes, there's a massive productivity speed up, and there's speed up in ways you couldn't even imagine, a priori. Yes, some jobs did change. The secretary at the time, typing pools went away, and they became executive assistants or what have you. There's always more work, and what we've discovered is the productivity gains from this AI are such that it does create a lot of expansion.
It does create a lot of work for humans. The other thing is that there's a fundamental issue with the AI as well in the pricing model. Now, I've talked about this in my latest podcast, an essay called "Pay to Pray," which is a lot of these tools were launched with freemium models or flat rate model, so you can go use Claude for free. You get a basic amount of token allocation, then you pay $20 a month, you get a bit more. If you pay $200 a month, then you get access to more tokens. But your cap is sitting at $20 or cap is sitting at $200. Well, as it turns out, nobody in the AI space is making any money at all other than NVIDIA and TSMC. NVIDIA makes $160 billion a year of revenue, $100 billion of that is EBIT.
TSMC being the foundry, again, makes good numbers. None of these AI companies are making any money, and the reason why inference is not profitable. That's why when you use Claude or you use ChatGPT, and you're a power user, instead of saying, "Put your credit card in and top your account up," in the past, they've been saying, "Sorry, timeout for seven hours or five hours," or what have you, and you're in the naughty corner. If there really was a profitable model behind the inference, they'd just say, "Stick your credit card in and top it up." Then you've got some issues in terms of data center capacity and the rollout, the energy prices rising, et cetera, and so on.
What it all leads is this, particularly the software development packages where mass token burning is happening, they're going to have to move to a pay per token model. What that means is as software developers use the tokens in core code, they're going to have to pay for the tokens as they go. The problem is that the nature of these AIs is that they hallucinate, right? They are liars. In a way, it's turning software development into a little bit of a slot machine. You put in a query, you pull the handle, maybe the AI will one-shot it and figure it out, and for very, very simple things, maybe it can. If it's something more complex, maybe it will come along. There's a lot of social media cases where people have said it's catastrophically polluted the code base, right?
That's just the nature of the hallucination or nature of the underlying technology and how it works. Now, sure, over time, they are reducing this with smart engineering, the error rates at which these hallucinations, so forth, occur. The complexity of things that people are trying to do with it is also increasing. While the coin flip to get an error may be getting a lower and lower probability of the error, you're doing a lot more coin flips to go. Basically where I'm aiming all this is, it's getting to a point now where they're going to have to pass through the cost of the inference. The inference is actually quite expensive.
Not everyone's going to be able to afford that inference, and when you're software developing using AI tooling, just by yourself, vibe coding, you're really pulling the handle on a slot machine, going, "Okay, hit me, hit me, hit me. Do I get my app? I get my app." Sometimes it can go in circles and this, that, and the other. I think everyone has had that experience where you're trying to get something out of it, and occasionally it will go in circles, and you get kind of frustrated, and you kind of give up. You compare that experience to going to a Freelancer. You pay a fixed price. You pay a fixed price, and the other thing is when humans get things wrong, so let's say you use a Freelancer and he hasn't quite got it, they tend to hill climb out of the problem quite well.
A human will ask their friends, ask other skilled professionals, get help, go read stuff, learn stuff, read the manuals, and they'll kind of hill climb their way out of a problem. An AI can tend to go in circles and circles and circles and circles, right? There are some fundamental issues with AI. There's obviously a lot of hype around it. Our freelancers are powered by AI. It's the best of both. You've got a skilled professional with a power tool. You remember a power tool, it's like a chainsaw. In the hands of a skilled professional, a chainsaw can do amazing things. In the hands of a novice, like my mom, God rest her soul, look what you end up with, right?
The answer is, if you look at technology, cameras didn't put painters out of work. Sure, the market for portrait painting probably declined quite significantly from the 1800s, but the global market for artists to express themselves and generate income has exploded exponentially in terms of the ability for creative outcomes and to earn money. I don't think it's going to be the Dario view of AI stole my job and everyone's going to be out of work and we're all going to be on Centrelink. I think that's ludicrous. I think the nature of work will change. We'll be highly productive. I think there'll be an explosion of entrepreneurship. I think the ability to start a company is going to get easier and easier. We've got something we're launching next quarter around that. We can launch a company on our site pretty simply. Some jobs will change.
The people who embrace technology and have initiative and get themselves skilled in the art of communicating with the AI in the industry, they're trying to get work done. You've got to be a skilled software engineer to use Diffuty's internal autonomous trading engine. My mom couldn't do it, et cetera and so on. I think we are in a phenomenal place in that a skilled marketplace are lifted. We're getting the benefit of all the AI. A lot of people are coming to us now for AI jobs. There's AI training work we're doing and AI data collection work we're doing, et cetera and so on. I don't think everyone's going to be unemployed. To this date, I don't know of anyone who's in graphic design who's been told, "Sorry, you're fired.
A machine's taking over." Yes, there's been layoffs in companies for various reasons. I think the layoffs in tech companies has more mainly been around being able to afford the AI coding tools as the pricing keeps going up, and so they cut a lot of the low-performing engineers in order to pay for the AI software development. I think it's primarily what's going on, or they're going to soften dramatically. Yeah, you will see dislocation in things like support, where you've got people doing effectively workflows, highly parallelized, and lots of people doing the same thing. That will also free people up to do creative work. Thanks for your question. It was a great question. Liam, can you move on to the question? I know Ray had a few while I was there.
Yeah. Ray's got a few questions. His first one, is Stablecoin an opportunity for or threat to escrow or a different space entirely?
Yeah. Stablecoin, for those that don't know, it's a crypto coin that's pegged to the U.S. dollar, 1:1. It's a cryptographic digital currency. We haven't had zero demand from any of our freelancers to be paid in Stablecoins. Not a single freelancer has reached out saying, "Can you pay me in Stablecoins?" None. We have seen some demand on Escrow to take Stablecoin payments. We explicitly reject that. We don't accept crypto for a few reasons. One is there's no licensing in the United States in money transmission or in crypto escrow around crypto whatsoever.
There's a fundamental problem with AML, anti-money laundering, with crypto in that when you go to create a bank account, you typically go to the bank and present your ID, and they identify you, and then your bank account is linked to your identity because the bank has seen you and verified you and it's given out a bank account. Then, of course, you do KYC and things like that. With crypto, the wallet creation is decentralized. I just go into my app and I create a wallet, and I create multiple wallets and shuffle money around, et cetera and so forth. It does present challenges with AML. Also, it attracts hackers and this, that, the other as well. There's a lot of hot money sloshing around.
The toll that's being paid to the Iranians going through the gates, the Straits of Hormuz, that $2 a barrel tariff is being paid with crypto, et cetera. I did a bit of research into cryptography. For 14 years, I started the program at the University of Sydney, and I was back as I taught it. There's a lot of, again, a lot of hype in that space. It was very, very big in 2021, et cetera. A lot of that money, I have a firm belief it's hot money. It's money that's bypassing proper OFAC screening, politically exposed persons checks, sanctions checks, et cetera. It's very risky to take that money on, either to receive it and also to send it out.
We do not accept it at all in any of our platforms, and in fact, we have enhanced security measures around things like crypto projects on Freelancer. You've got to be a verified freelancer that's actually gone through an extensive video interview process before you can work on a crypto project. Just by my background in crypto, I'm very negative on the actual cryptocurrency that's out there. Now, look, things may change. I know David Sacks, who's the crypto czar for Trump, et cetera. They are changing regulation for the banks who have custody of crypto in the U.S. In the future, should the legislation change, we could potentially look at that.
There's pretty catastrophic things that happen with crypto in that people have a browser extension loaded, and all of a sudden, their crypto wallet's drained, and you see these hacks where hundreds of millions of dollars disappear overnight, and there's no way of getting it back again. It's out of the banking system. We've explicitly stayed away from it. When AI.com sold for $70 million, we did quote on it. They came to me first for that particular transaction, where it was a slightly smaller valuation. I did quote on it. The seller ultimately didn't want to receive cryptocurrency for that, and that's why we didn't touch it. That's why I went through to the founder of Crypto.com, who bought it. Ironically, we sold him crypto.com, the domain name, and we sold him a few other big high-value things itself as well.
We have turned down transactions because of crypto, but there's just a lot of risks around the space. Should it become more regulated and more of an accepted thing with the banks and actually adopted a little bit more widely, we might consider it in the future, but at this point, we have no plans to do it. In fact, we have enhanced security to our crypto activity in terms of crypto purchase on Freelancer. The answer is no, I don't see it as a threat at all. I think the governments around the world are not going to lose control of the money supply, and I think in the future, there may very well be digital currencies everywhere around the world that governments issue, which would not be crypto. It will be digital, more than crypto. We would potentially support that, or most likely support that.
These unregulated, well, semi-regulated gray crypto sort of projects run by unknown Satoshis and so forth, we don't support. Next question.
Another one from Ray. Any developments in taking Loadshift offshore?
Yes. We aim to launch in Canada first, and we have a partner we're in discussions with around that, and that's probably all I can say at this point, and that it's progressing quite well.
All right. Final one from Ray. Any comment on the share price trajectory?
It needs to go up, quite significantly. I was pretty amazed. Obviously, there's been a meltdown across the software space, prices are down 75%, et cetera. I mean, the share price is ridiculous at this point in time. I've been buying, I will continue to buy. My job is to get it up. I've got to get the freelance number working at the same time. The Escrow teams are doing well, Loadshift team's doing well. I think I personally think the gift of these prices, I'll be buying vehemently. Yeah, I'll pretty much just see where the market takes from now.
That was the final question.
Any other questions? Okay. I would wait a little bit longer, but sometimes people take a little time to ask the question. As always, you may direct questions either here now or if you want to do it in a more private forum, you can reach out at investor@freelancer.com and reach the team and we'll arrange a one-on-one with you. Always happy to speak to any shareholder, no matter how big or how small. Any other questions from anyone? A few more seconds to leave it open. Okay. Thank you everyone for attending the first quarter of 2026 financial results, and I look forward to speaking to you in the next quarter. Thank you.