Airtasker Limited ($ART)

Okay. So it's about 90 seconds past 11, so we'll kick off today. Thanks, everyone, for joining today. This is the first of our quarterly trading update, which we've now introduced, we were relieved from doing 4 Cs because of the sustained positive cash flow. But we thought it was best to establish these quarterlies because it was just getting too long between drinks to only do updates twice per year. So looking forward to sharing one of these quarterlies with you in the March and September quarter to offset the -- in between months in between the half year and the full year reporting. So usual disclaimer. I wanted to start with an executive summary. Overall, we had a really, really great quarter where Australian GMV really picked up. So usually, this quarter would be if you look at our prior season, it would actually be down on the December quarter. We actually bucked that trend and we saw our GMV climb to a record $56.7 million just for the Australian business during the quarter, and that represented a 17.8% growth on pcp, which was a pretty sharp increase compared to last year, FY '25, which is about 5% GMV growth. And then in the first half of this year, we're at about 6.4% of GMV growth. So I've been really, really pleasantly surprised and feeling really good about that growth. Most of that GMV growth then flowed through to our revenue growth. So you noticed that in this half, actually, our GMV growth outpaced our revenue growth, and that was primarily because we didn't make any changes to pricing in this half. And so a full year has now passed since we last made any changes to the monetization rate. And so all of that revenue growth is really being driven by the top line GMV growth. I think this is a really good situation to be in because in the future, updating prices and updating monetization is something that we can do once we've got massive usage going through the GMV line. Overall, I think this is mainly a reflection of the brand salience work, which we kicked off in September of 2024. It takes about 18 months. We're seeing that the first time we did above-the-line marketing in Australia. We saw that in the U.K., where we launched our marketing in September '23, and it didn't really produce results until about March 2025. And we're seeing the same thing happen as we reignite above the line marketing Australia. So as a reminder, we were not doing any above the line marketing from about February 2022 to September 2024. And so that reintroduction of marketing alongside predominantly ARN and oOh!media takes some while to build up, but has really, really done well and showing results now and proving out the brand salient strategy this quarter. The U.K. and the U.S. are also tracking along really well. The U.K. hit a new high GMV ARR of $23.1 million. This is actually in our Jan, Feb, March season, which is not the peak season yet. I think as we go to the April, May, June season, we can reach much higher heights in both the U.K. signing that base of 23 and in the U.S., where we're kicking off from 6.7 and I think can make a big difference as we go into the peak season. Importantly, we also launched memberships. Memberships, as a reminder, is designed to -- is the product that we're offering is $89 paid for by the customer. And if the customer pays the $89 annual membership price, then there's actually no connection fees for the rest of the year. And so what that is designed to do is make sure that there's no -- there's a huge incentive for customers to increase their purchase frequency because there's absolutely no friction or cost in keeping all of their jobs on platform. So really, really excited to have launched this. The feedback was pretty incredible. We passed 1,000 subscribers in the first 4 weeks, and we're continuing to see some really, really positive momentum. So I'm excited actually about what we can deliver once we have a full quarter of memberships running with the trajectory and velocity that we've got. One big call out on that is that with the utilization of AI-based tools, we were able to get membership up to market in this space about 5 weeks. And the pace of iteration on the membership, both the product and the go-to-market strategy as being really, really immense. And we've got a robust balance sheet. So we've got over $26 million in cash still on the balance sheet. And that gives us a lot of flexibility to continue to invest into the U.S. and the U.K. as well as to settle media partnership notes, which mature in July of '26. On the back of -- this performance really pleased to reaffirm the guidance that we issued in Feb '26. And for ease of reference, we've included that guidance again in this brief update pack. And that guidance was essentially that Airtasker Australia is going to deliver double-digit revenue growth, and that's going to come even more from GMV growth, which I think we're already demonstrating. Our Oneflare, we're going to maximize the cash contribution of that business, we've made some important changes there to reduce the cost base of that business. We're going to make sure that it's bottom line and does the best possible job for Airtasker in terms of cash generation. We're going to see Australian marketplace cash flows after head office, continue to generate more and more cash this year compared to FY '25. FY '25, we generated about $15 million of cash, and we expect that to go up in FY '26. And then the U.K. and the U.S., we're going to make a discrete and intentional $5 million investment into those markets for the remainder of FY '26, and we think that can deliver some really, really good proof and traction points over there. And just as a reminder, the U.K. business now at a $23 million run rate as we head into the fourth quarter. So I think that's becoming a really significant business, which is really exciting to see. And quite close to moving towards that profitability mark. We've got a strong balance sheet. Obviously, at the Feb, we showed a $27 million cash bonds. We still have $26 million cash. So as you can see, that U.S. and U.K. investment is going to play out predominantly in the fourth quarter. In terms of just visualizing that growth, Airtasker marketplace is 17.3% revenue growth, which was great to see and are largely driven by a strong Australian market and the U.K. really coming on strong. Airtasker Australia 17.8% GMV growth became 14.4% revenue growth. And again, gives us a really solid base for the future. And our Airtasker Marketplaces revenue was $1.3 million. But really, as you can see there, the fourth quarter and the first quarter of next year is really where we tend to see those big jumps in the performance of those markets. So I won't go into all of the detail here, but we've laid out really what the reasons supporting that strong growth was. Overall, it's a significant increase in booked task volumes coming from our brand awareness, our campaigns, which drove about a 19% growth in book costs, whilst a really strong uplift in the average task price and this is something that is a real feature of the business model is that when the economy grows, we automatically grow with that part of the economy. So unlike sort of like an advertising website where you've got to keep up, increasing your advertising prices, with Airtasker, we just grow the Airtasker economy, we are going to earn a percentage of that growth and that value that we create for our users. We are also worth calling out the third bullet point there, which is as our brand salience work has delivered results. We're going to continue to invest into that strategy. And I think importantly, we announced Channel 9 as a media partner of ours over a 3-year period, meaning that we -- that notice to be repaid or settled in FY '29. And I think these agreements have just proven their value to us. We got immense benefits as a partner of these media partners but also, I think importantly, from a cost of capital point of view, we're getting the benefits of this media upfront. And we are establishing -- we are going to repay that over 3 years. That gives us a massive cost of capital advantage because the cost of capital for our business because we've got so many growth opportunities is very, very high. So I'm paying 4.9% interest on that convertible note is a great model for us. And yes, as I mentioned, our brand awareness, unprompted awareness was up 33% on pcp. So that is certainly working. In addition, our partnership with Mutinex is to do marketing media mix modeling. And that proves that we have about a 1.08x short-term robust on our marketing. So although we're investing in this long-term branding, we're also getting short-term immediate results. In terms of the U.K., hit a $23 million GMV run rate. But I think more importantly is we haven't even started for this year. The big lift really comes in April, May, June. So starting to already see some results on that and very much looking forward to sharing the update in June. And in the U.S., a worthwhile calling out that the FX is actually obviously with the Australian dollar being strong, the U.S. dollar being relatively weaker that does make the ARR, which we disclose in Australian dollars, that does provide a delta on that of about 10% or so, given the FX fluctuation. But actually, it's a net benefit for us operationally because we're currently in cash investment mode, i.e., we're spending more cash than we're bringing in. So the net impact of a strong Australian dollar is a net positive for us as a business. Really looking forward to again sharing the results as we hit the high season in April, May and June. One of the things that has been a feature of our executive leadership team conversations a lot over the past 2 or 3 years, but predominantly like we really picked up pace in the last 6 months is artificial intelligence and AI. So I did want to just touch on this point and give a brief update here. So the first thing I want to say that -- which is really fortunate aspect of our business model is that Airtasker value is predominantly in the physical and real world. And so physical and real-world skills, which makes up about 95% of the jobs on Airtasker, they're all physical and real world, they're not remote or virtual jobs that can be done through computers. They're things that are done physically with your hands in real life. And these skills are going to become the constraining factor in the economy. Like if you kind of think about what all of this productivity means in the white-collar universe, it means you're going to be able to do things faster, you're going to be able to come up with plans to go and build things faster. But the constraining factor is still going to be -- you're going to need a plumber to lay the pipes in the building, which you've now sped up the economy around. So Airtasker is really well positioned to benefit from that growth in the white-collar economy. If you have a look at how blue collar is in real life, our jobs are going to be impacted. You can see here that Anthropic, our research shows that it's really in software and legal and admin that you're going to see these massive disruptions due to AI. And I think in some cases, potential replacement of labor with software. I think that this is -- Anthropic research indicates this is not really the case in the real world. Now that's not to say there's humanoid robots and other methods of taking AI and bring it to the real world on the horizon. I think they are. I also think that it's worthwhile looking at how far into the future that may be. And also the fact that although some of the jobs at the lower end of our physical work may be replaced, that's already happening. Warehouses have automatic robotic storage systems and things like this. But humans always just going to get pushed up to that next level of creativity. There's probably going to be more podcast. There's more influencers, more people doing things that are in that higher order stack. So I think that human labor is still going to be a really, really important thing as we move forward. If we then look at AI disruption from like in marketplace or a platform level rather than at the services level, there's definitely a lot happening in the software area. And I think we can all probably agree we're going to the SaaSpocalypse at the moment where companies that sell software are definitely being disrupted. So I think it's worthwhile sort of calling out -- if you take a first principles approach, how does that apply to Airtasker. So you look at this and an author, Nicolas Bustamante put out a great article talking about the moats that exist in software businesses. And on the left, you can see at this table, you can see the kinds of moats that are generally not going to be very valuable into the future. And you can see these other moats that it can be really, really important. I think calling those out, it's almost a description of what Airtasker is. So the first thing being that Airtasker is predominantly built on network effects. So these network effects are -- what makes Airtasker work is that when you post a job, you get 3 people saying, I'm ready to do that job right now, close to you. All of that is driven by the fact that you've got liquidity on the supply side of the marketplace, which is driven by network effects. Now I'm working really hard to work out like what the disruption case for that is because we're trying to rent team ourselves to work out like how might that get disrupted. But as far as I can see, these network effects are very, very powerful and set up a pretty strong competitive advantage/defensive moat. The second thing is embedded transactions. So one of the things that Airtasker does that, say, Hipages and Thumbtack and Angie's List do not do is actually running the transaction. And I think that if you look at what OpenAI did, they kind of went into e-commerce and they sort of thought about and they actually back out of it. And I think what you can see there is that's not the easiest pickings for a horizontal-based AI company. I'm not going to go for businesses where you've actually got to manage the customer, manage the money and be accountable to a customer. And they're mainly interested, I think, right now in moving pixels around and having somebody else do all of that hard work. So that embedded transactions component becomes very important for us. And then I think the third is proprietary data. So one of the things that Airtasker has done is we have found these Taskers, and we are giving them the opportunity directly to build their brand on Airtasker. And as they build their brand through their ratings, their reviews, adding their passport, their ID, their verification to the platform, that becomes a proprietary data set that gives us essentially a unique inventory of work as it isn't available elsewhere. So I think all of these things -- I mean, at the moment, I'm sleeping pretty well at night with respect to the direct AI impacts, noting that in the big picture, we're very cognizant of, if you don't disrupt yourself, you're going to get disrupted. Overall, I would also say that Airtasker -- AI is actually a huge unlock for us as a value driver. And I'm happy to actually take any questions around this more directly because this is something where I think the collective wisdom of all of our investors and stakeholders is actually like really powerful to sort of question what we're doing and interrogate it. But the first thing I would say is that we're getting a massive software development efficiency. And one of the best things about our business model is that we use software to deliver value to our users, but we don't sell software. So as the cost of software gets cheaper/the ability to build software becomes more efficient and more productive, that's a net benefit for us. Basically, it's getting cheaper/we can do more of the software that unlocks the underlying value, which is access to our network, access to our marketplace Airtaskers, we can do that at a lower cost much faster than ever before. And I think rolling out Airtasker memberships was a massive example of this. And also, if you check out what we're doing in recurring and rebooking of tasks, the pace of iteration has just picked up massively. So really, really excited about that. It also helped us with content moderation and leakage. We've talked a lot about people using Airtasker, we're going to clean up the streets. We've got to make sure that we don't have a broken window theory. We want to make sure that the marketplace is the most trusted place to buy and sell your local services. And one inhibitor to that is all of the scans and the little comment and things that come up on the platform. And it's the vast minority of people, but those few people can ruin up for a lot of others. And so by having AI-based tools which can really learn about this unstructured data, this ambiguous data. We can actually solve these problems, which were previously somewhat unsolvable with deterministic software. And then lastly, we're getting great operational uplift. Our customer support, our query times have dropped massively. We're able to get back to customers faster than ever before. You're now able to talk in real time to our customer service team via chat. And a lot of that is because the back end, the AI-based systems at the back end have been really, really great. So we just sort of come back to like defensive moats, or competitive advantages. I think I'd summarize it in saying, network effects, the fact that we run the transactions, the reputation or the proprietary data that we have in our marketplace, and then finally, the regulatory responsibility, which we take on, for example, providing all the tax information of our users submitting that to local authorities in Australia, the U.K. and the U.S. annually, which is now a law, which makes a pretty costly to come in and be a competitor of -- to spin up another marketplace easily. So thank you so much, and I'm just going to leave it on the cover page, which is something was a bit of a bucket list moment for the company, which is our first billboard going up in Times Square, New York. So if you can recognize those red stairs, Airtasker did a very, very savvy deal to start doing some great outdoor in that market. So thank you again for attending. I can see there's a couple of questions in the chat. There are also some questions that were submitted pre the webinar. So I'm going to now go through those questions. If you have any other questions, feel free to just drop them in the chat, and I'll make sure that we address them today.

So firstly, some of the questions that came in from [ Shreyas Sen, Solomon Alton and G. Sladden ]. So I think the crux of these questions were that the share price being challenged? And what are we doing about that? And why is that happening? So I think share price has been challenged. And frankly, as a business leader, it has been frustrating, I would say, I think -- I've been -- given the wisdom time and time and time again, which is, keep your head down and focus on just delivering for the business. And in the long run, the share price will take care of itself. And I think ultimately, delivering a massive reacceleration in Australian GMV growth is a testament to that focus of our team. And at the U.S., the U.K. charging as well as new product launches like memberships and recurring tasks. These are all things that we can be in control of. With respect to the macro, I think we are suffering from the SaaSpocalypse, which is bundling in with software I think we are being impacted by the AI uncertainty, which the slides today are designed to address some of those concerns and give people the ground-up knowledge around AI. And then finally, I think we are being hit by a small cap ASX challenge, which I think has affected a lot of small caps with some externalities ranging from certain funds being liquidated and having a clear positions and things like that. So I wish I had a better answer on what are we doing about the share price. If we can solve that everything would be ultimately solved, but I guess what I would say is I'm open to collaboration anyone's specific ideas on what we can do to try and improve the share price on the back of what I think a really, really strong business results. The second was a great question about Formula One and what it's doing for us and how it's adding value to us. So it's been an incredible journey with Racing Bulls. So as a reminder, we -- our partnership is that we sponsor the team behind the dream, the people doing all of the tasks in the great world of Formula One. Overarchingly, the audience or Formula One has just got up to the absolute roof. I mean, everyone is talking about Formula One. There's a bit of a hiatus in this last 4 weeks due to the cancellation of 2 Middle Eastern races, which have been a bit of a gap. But the audience for this is absolutely enormous. We get quarterly reporting from the Racing Bulls team on our brand exposure and all that. And that's proven a significant ROI, 2 or 3x what we invested into the partnership. But I think even more importantly is if you check us out on Instagram and Twitter, some of the collaborations that we do with all of these different properties that we've invested in are delivering some real results there. So a massively improved engagement across the board on our social handles and the brand cache that we're building. So Formula One been a great investment for us and they're actually one of the most economical and high ROI brand investments that we've made. Okay. I'm going to now switch from -- again, thanks Solomon, G. Sladden, and Shrey for your questions. And please keep them coming. Now switch to some of the questions that have been asked in the chat. So from [ Michael Trott ], a strong result, good to see, I want to understand why there's no small uplift in FY '26 on the back of this. So we have typically kept FY -- our guidance to be fairly -- we don't want to make promises that we can't keep. And I would say that we really want to make sure that we follow through. Double-digit revenue in the context of not making any changes to yield means that we had to deliver on our top line GMV. And I think we have done that. But I think we're reticent to upgrade that result from what's in February. And we would rather be more on the cautious sign and make sure that we are over -- underpromise and overdeliver. In FY '25 was one profitable from [ Daniel Sims ], what level of profit could you squeeze on it with this renewed focus. So Oneflare is not a separate business entity. So it's an asset that's owned by the Airtasker Group. And on a direct cost basis, it was roughly cash flow neutral, meaning staff, marketing and software that was explicitly allocated to the Oneflare business versus Oneflare's revenues were roughly equal. So on a direct basis, cash neutral. That said, if you would peel back all of the indirect costs, some of those being more abstract and some being quite connectable indirect costs. For example, some of the software infrastructure layer that we have or the fact that some of our software is on a per seat basis, and that includes a bunch of staff. So a lot of these things were not being factored in. And so on that basis, Oneflare was costing us money. So with these changes that we've made to change the cost base of that business, we've been able to bring back to in-month profitability. And we are assessing what we're going to do with that next. But I would say that its current position is it's roughly direct cash neutral. If we -- and if we eliminated some of those indirect costs, it would actually probably generate cash for us. So we haven't made any firm decision on that, but we are going to make a decision on our next steps for Oneflare in this coming quarter. I think one good callout in terms of like future profits of this model is that what it demonstrates to us is that the lead generation model, which is sort of the Hipages and Oneflare model, it is challenging to scale this model because the margin profile is very different to Airtasker. It's driven by salespeople, and it's largely dependent on Google search marketing. And both of those things mean that you don't quite get the same operating leverage as Airtasker. So I think it is good to have the inside run on the benefits of the transactional marketplace model versus the lead generation model. From [ Brent ], the monetization rate in the international marketplace same as Australia, roughly '20 to '22. Yes, we are currently running on pretty similar monetization rates. It's slightly higher in Australia because lowering cancellation rates increased the monetization rate. So as a reminder, the way that we calculate the monetization rate is the revenue divided by the GMV. And the revenue is much higher on a completed task versus a canceled task, meaning that as you decrease your cancellation rate, you increase your reliability, you generally increase your monetization rate. Again, this is really great to align incentives towards -- we do better if more people are getting their tasks done. So that's really, really great. But it doesn't mean -- there's probably a 1 percentage point difference between the U.S. and U.K. versus Australia because we have high reliability rates in the Australian market. GMV growth fueled by advertising contra makes it easily cash flow positive, but what do the results look like on an EBITDA basis? I think the thing that is really, really -- that's a fair callout. we are doing about $5 million run rate of advertising media contra in the Australian marketplace per year. And we try to be really, really transparent about that. We did 2 deals in 2024 with ARN with oOh! $5 million each, and we're roughly utilizing them over 2 years. So it's $5 million per year. And so if you wanted to get an adjusted EBITDA or cash basis for the Australian business, I would knock $5 million of cash off the bottom. So if you look at that in FY '25, we produced about $15.2 million from the Australian business. On an adjusted basis, if we had paid for all of our marketing directly with cash in that year, you could knock off $5 million of that, so that will be $10 million or $11 million of positive cash flow coming out of Australia. And we think that this is absolutely still the right use of our balance sheet, that every year, we bring the media forward and we'll pay for it 3 years later or 2 years later in this case. Because at 4.9% interest, which is what we just summed up with Channel 9, it's an incredibly good use of balance sheet. From [ Steve S ], so jump from 6.4% GMV to 17.8% is a material inflection. Can you talk to the cadence within the quarter as April tracking and what's happening with the different media partners? So a couple of things. First thing I would say is that the quarter was -- we saw a really, really strong run into Christmas. Not at 17.8%, but at a -- picking up a pace into the Christmas season. And as we came back in Jan, Feb, we just sort of -- we're just tracking in a higher volume, and that's been consistently at this high growth rate over the prior year. And as mentioned, of the 17.8%, about 9% of it is volume growth, predominantly being driven by higher throughput rate of tasks getting completed. So partially from an increase in posted tasks, but also a great material improvement in the assignment rate, which is great to see. And then about 7%, 8% driven by the average task price, just the value of tasks going up. One thing to call out is when you put a 1% pricing change on a 20-ish percent monetization rate, just tacks on another 5% because increasing your revenue from '20 to '21 or from '22 to '23 or '23 to '24. It's like a 4% to 5% year-on-year revenue increase. And what we're excited by is that this is a leading indicator. If you can increase your GMV, you can go back and monetize that later and drive revenue growth out of that. And that has a material impact on cash when you're looking at sort of like a $200 million GMV run rate so -- or a $220 million run right now. So really important to say that the revenue lags the GMV and that's intentional. In terms of how we think it's going into the fourth quarter of the year, we've got great momentum. And I would say that the third quarter is sort of reflective of what would be a good outcome in the fourth quarter as well. So no major update that we want to provide there. That said, seen some really, really interesting leading indicators from our membership program. So in the first 4 weeks, acquired 1,000 users, that's now 4 weeks ago. So we are seeing some new numbers coming through, which look very, very promising. And I think when you look at the structure of our membership model, the structure of the model is that it self-incentivizes customers to use the product more frequently because as a customer -- once a customer buys a membership, it's about $89. You are generally getting value from that membership when you use Airtasker 3 times or more in a year. And so we have a significant percentage of our customers now saying, 'Yes, I want to commit to using Airtasker 3 times a year." And the financial mechanism is really you get the cash upfront. And then if you use Airtasker 3, 4, 5 times more per year, you miss out on the customer side revenues, but you're monetizing that for 2/3 of its incremental revenue for every task that happens there. So I would say that for the most part, with some extreme edge case exceptions excluded when someone signs up for membership, it's absolutely good news. It's bringing forward cash and it's essentially committing to frequency and its recurring revenue that just auto-renews every year, which is great. From [ Charlie S ], in these changing times, how do you think about staying agile? How significant do you expect token costs to become as part of your overall cost structure? Really, really great question. We do -- we are investing into AI. But I would say that I take a pretty moderated or balanced view on investing into artificial intelligence. I know that there are headlines in the last week or so that Uber has burned through their token spend in the first quarter that they thought they would use for the first year. And I do think there are wasteful incentives that are in play around AI. Like, for example, a lot of boards are instructing CEOs, "Yes, just go and invest $10 million into AI." And then he'll just go burn through it. We're taking much more of an intentional bottoms-up process for that. So we are rolling out Claude Cowork across the company. Claude Code is already giving us about a 3x velocity increase on software development. But we think that, that for us as a relatively small team of 200 people, is about 75 per year. Apologies there. Had a little bit of a technical glitch. So I've just missed the last question. [Technical Difficulty] low double-digit number or high double-digit number on AI in FY '27? We don't expect it to be a material impact. [Technical Difficulty] Is that a little bit clearer now? If you could just get a thumbs up in the chat, that would be amazing. Okay. Awesome. So Steve S, I think there were some questions from you. If you could just -- if you could just repeat those in the chat, apologies. I'll go to the questions that have reappeared in my chat. So from Michael Trott, are there any updates on exploring other AI partnerships further, i.e., agentic storefronts that increase engagement? Yes, so we have an excellent search and discovery team, which -- who are exploring going well beyond our traditional Google search into a lot of social search and then AI search. And if you go in and look at some of our web traffic and a chat traffic and engagement metrics versus Checkatrade, Thumbtack, Taskrabbit, we're really, really far ahead. And the reason for that is that underlyingly, Airtasker's engagement model has a lot more content than those marketplaces because we are an embedded transaction model, people are posting the task, they're talking about the tasks, they're getting ratings, they're getting reviews, all these content lives on our platform. And ultimately, I think that the distribution is going to be massively favoring the people that are carrying that proprietary data. So yes, really, really excited to see that play out and to be leveraging one of our core advantages. We are building a ChatGPT app and a building an OpenAI app, embedded app, which is a lot of companies are doing this now. I actually tried the Checkatrade one in the U.K., and I asked it, "Hey, I've got a roof leak, who should I use to fix my roof leak?" And the Checkatrade app told me to use Airtasker. So I think I wouldn't necessarily say that, that distribution is reliable, and I would come back more to the fundamentals. If you've got the right product fit and you've got proprietary data, which I think Airtasker has the advantage in then I think you're going to see benefits from that. From Steve, so talking about brand awareness in the U.K., how is it tracking? Are you seeing uptick coming into peak season? Yes. So I think we are due to get a new YouGov brand update shortly as well as a Mutinex MMM model update shortly. What I'm seeing in my dashboards, I'm short of those quarterly reports is that the data is all heading in the right direction. One of the things I think is interesting and we're postulating sharing this in some of our investor decks. But if you look at some of the web metrics, I think those are great for like real-time indicators of how the market share is moving around. And we're doing a really, really good job there on both the traffic acquisition, but also a user engagement level. So I'd encourage you to look at Google Trends, App Annie, these kind of platforms to talk about. And we will start sharing some of that data ahead of them, too. What's the strategy on LLM integration? Any MCP connectivity? So I think one of the things that is really important is setting up your platform to be discoverable by the changing nature of the search landscape. So I've [ struck ] to one level back. I think that Airtasker has a really powerful proposition because it's rooted in the physical real world at the core level. And then we have network effects, proprietary data and transaction embedded, meaning that we don't get replaced as a platform. But I do think the nature of search and discovery is going to change a lot. People are going to -- instead of going to Google, they're going to go to different ways of finding out what to do next. Now I think that's going to proliferate. Many more people are going to be able to find more things to go and do. Constraint is going to be the physical world. So I think this is all going to be good for us, it's going to grow our marketplace. But it does mean that you need to think about how you engage with all of these different search methods. One layer of that is embedded apps. So we are building embedded up in OpenAI. I'm not sure whether that is the right methodology. We've also built our own AI-based interface. So if you go to airtasker.com/post-task-ai, you can actually try out our version of an AI model, and we've got 2. One is a sort of tap, tap, tap, call it, the building blocks AI model. We've also got conversational AI models. So both of those are interesting. But what we found is most users do not want to engage with AI to get their plumbing done just yet. They want to engage with Airtasker. It's actually simple, just post the task rather than talking to a chatbot. And then you've got the agentic protocol layer, where you want to make sure that people can buy -- agents can purchase on Airtasker and we are looking to make a lot of our code available and complementary to agentic commerce. But I think more practical in the short term is primarily making it so that the search and discovery bots can easily find their way around Airtasker and get to all the data, but that data is still living on Airtasker, so you've got to come to us to ultimately get it. And in doing that, some examples of that is we're moving everything to service side UI rendering. For example, like you don't want to have native UI rendering because bots don't actually render UI. So you want all of the pages essentially to be getting formulated on the server side. So when the robots come, they can go, cool, I can see how this whole page works. I can read it, and I can bring that information back through Gemini or OpenAI to the user. So that's an example of where it's not a direct AI integration, but we are basically setting it up so that the AI can access our platform and serve us traffic. Plans to AI to be embedded in the product, not as just a distribution channel? Absolutely. So we have a team working on a bunch of features. Some obvious stuff like predictive text, budget suggestions and things like that. And our conclusion after putting a lot of effort into building an AI interface, a custom AI interface was that people don't want to -- people don't want to post a task through AI. That's not what they're thinking. They're thinking, how do I come to Airtasker? And how does AI and make it easier for me to get a task done? So it's more that Airtasker can take components of AI and make the Airtasker experience better. I will continue. That was from Charlie Song. Steve, we talked already about brand awareness in the U.S.A., U.K. From [ John B ], why did you go with 9 in the latest media deal of ARN and oOh! now delivered? So we have done partnerships with an outdoor company oOh!, and an audio company, ARN, and Channel 9 obviously brings us the third kind of broadcast media, which is online -- which is TV and streaming. They do actually also own QMS now, which is another outdoor brand, which is an interesting place to be in. But overall, we've had a great experience working with ARN and oOh!. Awesome. We've uplifted our brand. We wanted to add television into that. And moving into FY '27. We're in great discussions about extending partnerships with all of our media partners and really emphasizing the one that have delivered outsized results for us. And using the marketing mix modeling that we're doing alongside Mutinex, we're really able to see what's working versus not -- what's not working. So really, really exciting. U.K. close to profitability, looking at new markets. So I think one of the things I wanted to call out here is the benchmark that we set out, which is over 3 years, we can get to a $25 million GMV run rate generating $5 million in gross margin, and that's enough to keep that marketplace going and easily be profitable. And we're going to fly through that target, I think, in the fourth quarter. We're at $23 million in March, and we usually see a good run into the year. But I think that really what we have is a choice of do we continue to invest in that marketplace and double down on that growth? Or do we bring it back to cash neutral or do we actually start taking dividends out of it to invest into the next market? And I'd say with a very cash flow generation mindset, I think the best decision now is to start bringing Airtasker U.K. back towards profitability. So we are expecting a net burn to be significantly less in '27, FY '27 versus '26 for the U.K. but we will likely still be net cash investors into the U.K. market. So I wouldn't think of it yet as a cash generator, but I would think of it as less of a cash investment, meaning that we can dial up investment elsewhere. Predominantly, that focus right now is in key cities in the United States, which is going well. But we're getting so much inbound interest from media players all across the world that there really is an opportunity to take that effective franchise model and move it into new countries, leveraging our software platform, which we're building and investing into anyway. From Charlie S, would you agree that the barrier to entry has fallen for developing marketplaces? How do you think about growing retaining market share outside of your network effects and proprietary data? It's really interesting to think about whether the barriers entry have increased or decreased for marketplace. Some examples of where the barriers have actually increased would be, for example, regulatory lock-in. So in Australia, you have to provide -- even if you do even 1 transaction online, you need to provide tax information to the government. That system costs about $1.5 million to $2 million a year to run that system. It also adds a lot of friction to acquiring taskers onto the -- on the platform. They're going to put in all of this like verifiable, essentially similar to KYC AML type information, and we have like 100,000 of these users who have done that. Replicating that on another marketplace is really, really difficult, especially when you're starting from 1, 2, 3 transactions a year. We know this because we had to start from 1, 2, 3 transactions a year. So I don't think the moat of network effects and regulatory lock-in has gotten easy. In fact, I think it's probably gone a lot harder because the cost of marketing has gotten higher. And the cost of capital has most likely gotten harder for marketplaces as well because in 2010 to 2020, the general valuation multiple was something like 20x forward revenue for a marketplace. And now we're probably looking at -- in a good world that sum -- that number is probably less than 20x next year's revenues. So I would say the cost of capital is higher, the cost of marketing is higher. And the ability to do that regulatory stuff is harder. How do you plan to pay back media commitments in the near future? We have cash on balance sheet to pay back media commitments if we so choose to settle them in cash. And as a reminder, it's really our choice to settle them in cash or settle them in equity. But we have sufficient cash to settle them all with cash. With respect to future media commitments in the Australian market, we're very much taking the view that we do not want to be dependent on any kind of external funding. So essentially, what that means is that we are storing cash on our balance sheet ready to pay down Australian media notes. With respect to U.S. and U.K. media notes. We provided some analysis in the last presentation, which spoke to how those media deals get sorted out. And if you kind of look through them, a very, very, very high probability that we're going to settle them into equity, into the local subsidiaries. And then depending on what the outcome is settle those again with equity in ART. And again, as a reminder, all of those choices sit with Airtasker as to repay in cash or repay in equity. So we have a lot of flexibility to do that. Any plans to spin off Oneflare? I don't have any specific update on that today. But we are watching this business. We've made some significant changes to its cost base, and we're reviewing again this quarter, and we'll have a good update in June. Are you looking to expand -- from [ Ben A ], looking to expand into any other international markets? We are getting a lot of inbound interest from media companies all around the world going, "Wow, this model works." The model of turning media traffic into transactional models is really, really exciting. So we've got a lot of options there. The constraining factor, I would say, is the share price because to go into new markets, it is a net investment, including cash upfront. So the 2 choices are that we continue to just buckle down what we're doing now, continue to expand the profitability and cash generation of Australia and in the U.K. shortly as well. We then use that cash to expand. But of course, if there was a change in the cost of capital, we can put our foot down massively. And so we think that's a really exciting reason to be a listed company is that we have a couple of good quarters, we can really build up that momentum. And if that capital, again, changes, we can turbocharge our investments into new markets as well. And as a reminder, we're using the same piece of software, same piece of software, going into new markets, we're just replicating the marketing. So the economics are very exciting. G. Sladden, are the growing media deals having an impact on the ART share price? Yes, I think it's been quite positive. I think the media deals are being -- as people sort of realize that the cost of capital is 4.9% on the media deals. I think that by institutional investors being perceived very, very well. So we're going to continue to deliver results of them. That is obviously important, but it's also a consistent part of our strategy moving forward. From [ Hayden Ness ], with the cost of building and testing software falling dramatically, as you mentioned, I'm curious about what's in the product road map of projects and initiatives that perhaps weren't economically viable to test and rollout 12 to 18 months ago? This is what gets me excited. Things that were literally felt like they were 3 to 5 years away and now 3 to 5 months away. I think membership is a really, really exciting product vector, which we are charging down, and there's so many iterations and variations on this. I think our recurring tasks and rebooking was always something that was -- it was a pretty heavy lift on the software side and the pace of iteration on rebooking has been incredible. Gift cards, I think Airtasker as a gift is also a very, very interesting product vector, which I think you're going to see a lot of rapid iteration on. And then in the big picture, I think the financial ecosystem that Airtasker is building, it's a sort of $200 million, closing on $250 million GMV marketplace now. And a lot of that money just to sort of transient and sort of moved through Airtasker. Building on an ecosystem which retains some of that financial value moving around is very, very exciting to me. And so watch this space. I think you're going to see a lot of movement in that space. And we'll endeavor to share more on the product update side. And then lastly, other AI partnerships. I know we talked about AI a little bit, but yet every layer we are considering AI, but also being quite balanced. Because I think there are a lot of companies just investing for the sake of investing. And that's great for Anthropic. And I think it's kind of genius the way that Anthropic has kind of told you, "If you don't spend money with us, your business is over, so spend all your money with us," but packaged up in a very marketable and positive way. But I think it's also important as the investor of that money to be really responsible and intentional about how we spend it. So we're going to continue to do that. Thanks so much and glad that we actually got to go through all of the questions today. Apologies for the technical glitch earlier. And I hope everyone has a great day, and we can see some positive moves on the share price. And if we do, I think we can have some great compounding impact into the future. Thanks so much, everyone.