Hello, welcome to the Freelancer Limited first quarter of 2021 financial results presentation. My name is Matt Barrie. I'm the Chief Executive of the group. With me today is the Chief Financial Officer, Neil Katz, the VP of Product, Adam Byrnes, and the VP of Enterprise, Sarah Tang. After the commentary, you may address questions to any or all of us in the Q&A pursuant. Thank you for joining us today. We just finished up the first quarter of 2021. We operate in the calendar year, that's from January to March of this year is our first quarter. We delivered Gross Payment Volume in the first quarter with an all-time record of $192.9 million, which was actually up 39% on the previous year, or AUD 249.7 million, up 18.8%.
We're on track to do over $1 billion in payment volume through the bank accounts this year. Our GMV for Freelancer was up 23.6% to $25.9 million or AUD 33.6 million, up 5.1% on PCP. The Escrow gross payment volume had a big lift, again, following last quarter's results, an all-time record of $158.2 million, up 43.7% on PCP, or AUD 204.8 million, up 23% on PCP. We possibly stand a chance of the Escrow business itself doing over $1 billion this year with the growth that's happening at the moment. The GPV ex-China was an all-time record at $151.1 million, up 48.7% on PCP. Group net cash receipts for the first quarter was an all-time record of $12 million, up 32.1% on PCP, or AUD 15.6 million, up 12.4% on PCP.
The Freelancer cash receipts were an all-time record of $10.1 million, up 31.4% on PCP, or $13.1 million, up 11.8% on PCP. The Escrow cash receipts were an all-time record of $1.9 million, up 36% on PCP, or $2.4 million, up 15.8% on PCP. We obviously had a headwind in the Australian/U.S. dollar exchange rate in the quarter. That was negative 17.4%. We still got fairly decent results, even though we had that headwind. We think that's obviously going to turn now because in the last year, you remember the Australian dollar dipped down to $0.55 and kind of bounced back up again. It's been relatively steady in the last number of months. Approximately 72% of group revenue is in U.S. dollars and 4% is in AUD. We're very much a global business, U.S. dollar is the main metric we look at.
A positive net operating cash flow was $4.2 million. It was a big lift in the first quarter versus $0.47 million in the first quarter of 2020. We're breaking profitability. As I mentioned in the previous quarterly report, Escrow, Freelancer, and the group were both individually and together profitable in the second half of 2020. Escrow was profitable in the first half of 2021. Freelancer was just slightly under, but we've been kind of dipping in and out of profitability over the last number of quarters. In fact, Freelancer was profitable in Q2 of last year, Q3 of last year, and Q4 of last year as the marketplace business. The jaws of operating leverage are ready to open. We had a reasonable lift in cash and cash equivalents, up 8.2% to $37.8 million.
If you look at the Freelancer gross marketplace volume, it's actually growing at the fastest it's been growing since IPO. On a rolling 28 day basis as of a day or two ago, it was growing at 24.1%. You can see here this is a multi-year GMV graph for the last three years. These plots are actually on a rolling 365 day basis, but you can see the growth has had a big uplift in the last number of months. As I've said before, we're working through each of the different revenue lines to get them moving at speed and really get the growth happening, and some of those efforts are starting to pay off. We're ranked globally fairly high for web traffic. In the end of last quarter, we were about the 1,200th biggest website in the world. As of today, we're about the 565 and continue to rise.
We expect that to continue to rise over the next quarter. We actually get quite a big deal of traffic. Web traffic in the year last year was up around 100% in terms of page views year-on-year. We obviously see the continuation of this. We actually think that this number will actually get us into the low hundreds fairly soon. You can see here on the GPV charts for Escrow, another big candle. This follows up on the last quarter's results. We had a big uplift of about $50 million on a year-on-year basis. Escrow's GPV actually continues to grow even better than that as we move into April. I won't preempt anything from that. What I might do now is I might slip across to the commentary. I'll just stop my screen sharing and re-share another page.
Just give me one second. Here we go. Is that sharing? No. Hang on. I don't think it's sharing. Is it sharing now? Okay, great. Getting down to the segments. I mentioned before, Freelancer is growing at the fastest pace since IPO. I've charted that down here, where you can see here the GMV growth year-on-year. At the time of the IPO, we were a little bit over 20%. Then as we put a lot of investment into really the paying down technical debt, GMV growth slowed a little bit because we basically had to redirect resources from product development into actual infrastructure work. In 2019, I have said before, we did kick a few own goals. Am I showing the wrong screen? Sorry, one second. What screen am I showing? Second. Where's the meeting controls? Sorry, it's not.
Sorry, Zoom is not showing me the meeting controls. Apologies for that. Here we go. I'll share my screen again. Chris, you give me access, please. Give me access to the admin so I can share my screen. Now I have that. Sorry, it's disabled screen sharing. You have to make me an admin. I'll keep talking until the screen sharing is actually enabled. At the time of the IPO, the GMV growth year-on-year was a little over 20%, and at the moment is about 24.X%. Here we go. The screen sharing is enabled now. Again, apologies for this. I think you can see the screen now. You can see here the GMV growth is faster than it was at the beginning of the IPO.
2019, I said before, we did kick a few own goals, in terms of pushing the new front-end architecture a little bit too quickly. That did shave off a bunch of GMV, but we've been working pretty pathologically on each of the bits and pieces where we had incompatibilities with previous versions in terms of features, and we are getting a bit of lift now. One example of this, which I've talked about previously, but I think it's a good example of this, is the hourly funnel. The hourly tracked hours are now growing about twice as fast as they did in first quarter of 2019 on first quarter of 2018, up 23% on PCP. You can see here, in this figure four that during 2019, we pushed the front-end architecture.
We had some features that we're missing really from the hourly funnel. It really shouldn't have been pushed that quickly. It's a very, very complicated piece to actually get out there because effectively we were rebuilding the entire front end of the website. You can see that in 2019, there was a slowdown. You can see now in 2020, 2021, that it's growing faster than ever. That's also replicated with the fee graph. We can see there again, 2019, we had a bit of operational slowdown due to our own fault. 2020, you can see it's growing exponentially again. That's what we're doing in terms of working through each of the different revenue lines and GMV lines, and we're getting some great results.
In terms of supply, we are the largest marketplace in the world by a long way, by a number of users, 50 million users, I think 51.4 as of a few weeks ago. We talked about in the last couple of quarters that we did see a little bit of a drop in the average project size from AUD 216- AUD 161 in the middle of the COVID. That was due to an influx of a lot of new labor as well as a lot of rookie clients. I did say that it would increase. In the third quarter of last year, it rose to AUD 172, and as of the end of the first quarter, it's now $187.
Over time, the average completed project size does rise slowly due to the complexity and sophistication of jobs that you can do on Freelancer. When there are spikes of labor coming in, which only really happened once before, which is during the global financial crisis. You do see a temporary blip in these numbers, but I do expect them to continue to rise, which will be a contributor to GMV and to revenue. Mobile, I've talked about a number of times over the last number of quarters. I'm pleased to announce that not only have we ripped out the legacy mobile site and replaced it, which we did in the third quarter of last year, but we also have replaced now the iOS app.
Last year, in the second half, I reported that the mobile web fees grew 51.5% on PCP from the second half of 2020 on the second half of 2019. That's in the first quarter, is now up to 110.6%. This has paid off. We expect to see a strong rise in the iOS fees, similarly to how we did with mobile legacy. It's simply because now we have one code base. When we update the desktop website the entire code base for the mobile sites are updated at the same time. Android is imminent. It's a couple of weeks away. I know I've been saying this for some time, but these are complicated things.
Now that iOS is fully out, we expect it to happen really in the next four weeks. Right? It really shouldn't be any longer than that. Maybe even be sooner than that. That's the approximate timeframe. We look forward to seeing what will happen with the iOS revenue, and see if it's on par with mobile legacy. It's a pretty strong lift, and it's been a great project. Then we effectively reduced four code bases to one code base. Desktop, mobile web, iOS, and Android now are all the same code bases. One area that we need to put a little bit of work into is contests. As a product, contests is working fantastically well. I mean, we're doing contests for NASA.
We've got some big ones that have come out now in the $300,000-$400,000 , the biggest contests we've run before. We've run contests for the U.S. Department of Energy, the U.S. Bureau of Reclamation. This is one of the live ones now with $365,000 National Institutes of Health. This contest live for $475,000. There'll be plenty more coming as part of that $25 million joint tender that we won from NASA. IBM is doing a bunch of things. In fact, they're going to be a great partner. We were added actually last night to the IBM Cloud catalog. Internally at IBM now, globally, people can use Freelancer from IBM staff, Airbus, Novo Nordisk, et cetera. The platform is working fantastically well from a product perspective. 14.5 million contest entries submitted every year.
Contest entries in the first quarter are up 52% to 3.82 million entries. There is no other platform in the world as liquid as we are for contests. It's actually amazing. If you haven't used it before, just put AUD 10 in and try and get a business card redesigned, you'll get hundreds of entries. An average of 220 entries get submitted, average of AUD 88 per contest, 80% of contests receive an entry in the first hour. While contest is working fantastically well from a product perspective, the fees are basically lagging a little bit. We know why. There's a bunch of upgrades that are missing off the contests platform. When you post a contest, the upgrades are hidden away. While before they were just on the post contest page. This is just again from the port to the front end.
We just basically cut a few corners to get something out a bit live. We will get this to lift up strongly in the next couple of months. You see here in the green line, the number of contests in figure 10 is higher as it has been before. We just got to get the fee revenue up, which is primarily due to upgrades. Number of contests being completed is way up. You can see there. The red line in figure 11. We're getting more contests being completed and handled than ever before and so on. We need to fix the revenue one. Over to Enterprise. We've had a decent quarter for Enterprise. Maybe, Sarah, if you can take over for a few minutes.
Sure. Can you hear me?
Certainly can.
Yes. Perfect. Hello, everyone. Freelancer Enterprise has had a great quarter. We've been growing about 83% year-on-year on PCP. This has been driven mainly by strong execution of key accounts. Average spend of our key accounts has actually increased by two and a half, about 2.3 times versus the same period last year. More of the organizations that have come out of pilot last year have actually moved through the deployment status, and the spend on the platform is increasing. A few clients whom we did not have last year have actually just recently spent about six figures on the platform in the last four months. That's very exciting. We also closed some additional MSAs, including one with a AUD 17 billion global business processes outsourcing company. We, within a span of two weeks, have actually engaged 318 freelance hires.
The speed and ease at which we're able to do that beat their internal hiring benchmarks by a landslide. That's actually opened up the doors for conversations to hire the same type of freelancers with commitment of about $50,000. That's looking to be deployed within the next year. Now, we also signed a statement of work with a Fortune 50 technology business, where we'll be integrating some pre-vetted branded communities into their cloud platform. This is the one Matt mentioned earlier with IBM. We actually last night just integrated with their IBM Cloud catalog, which means the IBM customers can actually go through that IBM Cloud catalog when they want to deploy and migrate to the cloud to actually leverage our talent and our freelancers. This was a great way to start testing and improving network effects.
Now we are also building a landing page, and they are spending an additional six figures, and that is continuing to grow over the coming months. We also signed what we have managed to do in the last year is really shift from pilot to scaling for a lot of our customers, with many of them actually committing up to AUD 400,000 at a minimum of GMV spend. One of them is actually an IT company leveraging us to manage their contingent workers. Now we realize we are actually the only solution that can actually help these organizations seamlessly build a blended workforce of their entire talent cloud, whether that is through internal, and that is leveraging and commercializing the product that we built for Deloitte, My Gigs, to manage their internal workforce through their existing contingent labor spend.
Through our Bring Your Own Freelancers Program, and a lot of customers are really liking this program where they can bring their existing freelancers onto our platform, and we charge less fees than many of our competitors. Finally, to them finding new freelancers on our platform. Again, under the $25 million NASA joint tender that we won. The U.S. Bureau of Reclamation contest valued at $365,000 that we partnered with and was able to win due to our partnership with Arrow Electronics. As well as the U.S. National Institutes of Health one for $474,500 that we partnered with a local data science company and a logistics company, Adiona. They're going well, and they're being deployed onto our platform as we speak. There's additional tenders that's coming up that we have a really good chance of obtaining.
What we really invested this last year that enabled us to start achieving this level of scale is really improving the account representation within our key accounts and streamlining a lot of our internal systems and operating metrics. This has really been, I think, our efforts within streamlining the operational aspects with these large-scale customers has really been demonstrated by the growth in our enterprise GMV. I think, yeah, that's paying dividends and actually our GMV is projected to grow faster than what we currently have within the next six to 12 months. I'm really excited to announce that we invested heavily in product and released a number of enhancements. One of them was project tracking, talent management, and shared payment options, which are super important for large-scale customers.
Once we actually get some of these products refined and features refined, I think really help us scale these larger customers. I think in summary, we're in a really good state, and I'm really confident in the continuous GMV growth. We've managed to come out of last year being successful at a majority of our pilots. A majority of the customers have actually now committed to a much larger spend on our platform. I'm excited that we're working with Adam and the product team to really build out the scalability of the product so that these customers can manage AUD 1 million of spend at a time on our platform. Back to you, Matt Barrie.
Great. Thanks, Sarah. Another exciting thing we did in the quarter was we produced a proof of concept app, which is actually live now on the iOS and the Play stores using the Freelancer API. The API basically lets you access the full functionality of the Freelancer marketplace from your own website, app, or software. Really this allows software really for the first time in history to control humans and task them with instructions. We thought we'd build a demonstrator app. For some time I've said that if you look at Uber, the last time I checked, Uber had 7 or 8 million drivers. We've got 51 million freelancers. We probably have more drivers than Uber.
In theory, you could build an Uber on top of Freelancer, but the interface at the moment for the website has been built for the general case and not for the specific case of transportation. In the case of using Freelancer, it's unlikely you would post a job for picking me up from a certain location and drop me to another location. However, if you actually build a custom interface which is specialized for a particular application, we wanted to see what the experience was like. We thought, let's come up with an Uber for something, and we came up with an Uber for photography. We've built this app. You can download it yourself from the iOS and the Play Stores. They're available now. Effectively it just behaves like Uber. You type in an address, you've got a map of the world.
You place your order with a bit of some instructions, and anywhere in the world, photos will be sent back to you within 24- 48 hours on average. In urban centers, it can be a lot quicker. When I post a job in New York, provided that the sun is up or in Sydney I've had photos returned as soon as one and a half hours. It works anywhere in the world. Some of the locations where I've tried this I tried it worked in Moscow, worked in Budapest, worked in India, worked in Bali, worked in Sydney, New York. It's worked in remote countryside of Wyoming, the remote countryside of the United Kingdom, even Chernobyl, we got results back within 24 hours. Timbuktu in Mali, a train station in Uzbekistan and so forth. Really it's the proof of concept app.
We think there's a lot of applications for this. For example, if you're Procter & Gamble, you probably want to check to see your products on the shelves of Walmart and Kohl's look great. If you are in real estate, perhaps you want photos of a house in a certain location. I've seen the app being used by a fund manager to get caravan parks looked at. They invest in a bunch of properties around the world. In fact, some photos just got returned only a few minutes ago in New Zealand for a certain caravan park and so on. It's pretty interesting proof of concept. The step out from here to actually produce apps that do the equivalent of what multi-billion dollar companies do is actually quite straightforward. For example, many of you have probably heard of Gojek.
That's a multi-billion dollar company in I think four countries in Asia, where you type in an address of a shop, and you tell someone to go to that shop and buy something for you and bring it to you, right? That's in four countries, yet it's a multi-billion dollar company. I mean, for us to modify this app to basically allow you to go buy something and bring it to you is only really a month or two's worth of work. There's a bunch of crazy things we could do with this. It's an alternative way of expressing the marketplace as more of a niche narrow app. This initially is a proof of concept. You can try it for yourself. In this particular case, it costs $ 35. What happened, you get five photos. In fact, in some cases, they send you back dozens.
The amazing thing about this is it's just worked everywhere. What is happening behind the scenes is when you click place order, it posts a project onto Freelancer with a milestone payment of $30. The app handles all the bidding. When a freelancer bids on your project, et cetera, handles the backs and forwards communication, it'll automatically award a certain freelancer and then the freelancer is then tasked to take the photos. When the photos arrive you hit the button to review them, and then you hit a button to accept them, and then it pays them and releases the milestone. All it's doing is just using the basic freelancer app, marketplace and the API at the back end. Nothing special there.
It handles the smarts around who to select and what happens if they don't respond back to you and select another Freelancer and so forth. Ultimately it's a great expression of the power of the platform, and I encourage you to all try it. Just download it, try and get some photos taken. It's pretty amazing what you can get done. That's it for Freelancer, and now I'll probably talk about Escrow for a little bit before I open up to Q&A. We had another fantastic quarter for Escrow, $158 million, up 43.7% on PCP. That was after another big step-up from the last quarter. You can see here in figure 13. These are sort of $50 million GMV absolute number step-ups in U.S . Dollars.
Ex China was up 48.7%, so $151.1 million. Cash receipts are up 36% U.S. dollars to $1.9 million or AUD 2.4 million, 15.8%. There's some pretty strong growth in the GPV, and that growth is actually faster moving into April than it was in the previous quarter. It's really moving in a big way. We hit a milestone in the first quarter of $5 billion all time in transactions secure. That's actually cash through the bank accounts. There's some pretty good volume going through this platform. The growth came from strong sales in digital asset sales, domain names and so forth. We sold domains such as nft.com and nfts.com in the quarter, crypto.org and so on. Lots of physical merchandise. That was actually up 87% year- on- year.
IPv4 contributed AUD 16 million in the quarter. Last year, I think we did $70 something million in IPv4 sales. These are basically IP addresses. This only two years ago was basically zero. This is an example of a segment where we have taken the segment from zero to possibly this year, AUD 70 million-AUD 100 million in volume in that segment alone. Motor vehicles were up 230% in PCP, contributed about AUD 12 million to GPV. The number's still fairly low, but growing fast. eBay Motors made a small contribution to this, but it was not the majority. We are in a bunch of different car marketplaces in addition to consumers using the website directly. We do, as I've said many times before, expect to tip the entire automotive industry. We are the payment system for automotive as automotive classified sites move towards becoming marketplaces.
It might surprise many of you, but the car sites around the world, being classified sites, don't actually know very much about the seller or the buyer. They make an assumption that the car has sold when the listing stops getting renewed. If you don't take the payment, you don't know if the buyer is reputable, and you don't know if the seller is reputable because you don't really collect reputation. It's a big problem for the automotive classified sites globally. We are the solution to that. We're in eBay Motors. We're in Best Car Finder. We're in a bunch of other different car marketplaces out there. In the quarter, we went live with Truck Trailer Tractor in more heavy machinery. We think we will tip the entire automotive marketplace. We think we are in a unique position to do that globally.
It's going to take a couple of years, but I think we will get there in a big way. There's a lot of volume in cars. EscrowPay is a streamlined version of the checkout experience. You can see a picture there in slide 13 or 14. We've really put a lot of work into making this really like a checkout. The old way you integrated Escrow, you had to log in, set a password, all this other stuff, and it was a bit clunky. With a lot of feedback from eBay, we streamlined this, so it now looks like a PayPal checkout experience. We're pleased to announce also that as of last night, we've just put PayPal and there's also credit card in this checkout experience.
We're in the process of rolling out this new funnel in the next month or two to all our partners as well as the main experience using Escrow. There's going to be some upside from that as well in terms of product improvement. It is really quite a slick experience. We also turned on ACH Debit in the quarter. ACH Debit is a bit like a wire transfer, but it's a bit more immediate. We obviously verify ID and get tax information and so forth at the same time. We've launched it first with eBay. Now it's in the process of being rolled out across all the funnels for all our partners as well as main consumer products. That's going to be happening in the next few months. We're in production with eBay Motors and in beta with eBay Watches for a small number of sellers.
In the last week, eBay has put a bunch of information up on their website about eBay Watches. We are imminently, within days— I won't mention which day because these things always take a bit longer sometimes. Even if that's the day it always kind of blows past the day. It's within days to maybe a week, we anticipate going into full production with watches. If you go to eBay's website in the U.S. and you browse around now, Escrow.com is mentioned on many different pages as to how it works and FAQs and so forth. That marketing material has only just started to go up online. Note that eBay, while has tremendous upside for us, not just in the categories we're in, but more categories over time, it is not the primary driver for the GPV growth in the last couple of quarters.
We do expect that eBay contribution to grow over time. There's a lot of upside still to come with eBay over and on top of what we've done already. I've talked about domain name sales, et cetera. We've gone to all sorts of marketplaces in the quarter. Just some we pulled out just for interest's sake. We added motorcycleshippers.com, Truck Trailer Tractor, Topmark, which is luxury car marketplace, MicroAcquire for M&A. In addition, we even went into an ostrich marketplace. We went into the United Ostrich Association, the largest association of ostrich breeders in the United States. That gives you the spectrum of sort of things that you can buy and sell with Escrow. Of course, the reason why you want to use Escrow is you want to make sure the ostrich arrives alive, and there's a lot of complication around shipping and so forth.
It's not something you would use a PayPal to check out with. The group had net operating cash flow of $4.2 million for the quarter. It is up substantially on the last quarter to PCP, which is $0.47. We end up with cash and cash equivalents, a big lift, $37.8 million, up 8.2%. Escrow, Freelancer and the group were all individually and collectively profitable in the fourth quarter and second half of last year. Escrow was profitable in the first quarter of this year. I mentioned there was a bit of a headwind for the currency, we anticipate that to fall back over the next couple of quarters. We also, in the quarter, started trading on OTCQX Best Markets under the ticker FLNCF. That's in the U.S.
That will increase our ability to be able to have our shareholders, as well have our investors in the U.S., our customers all around the world and so forth, be able to buy our stock. If you're somewhere like India, for example, you can't trade ASX shares very easily at all. You can do OTC U.S. stocks. We think that's going to open up really the ability for people to buy our shares more easily around the world and so that was good news. What I might do now is I might just open it up to the floor for question time. As I said before, you can address your questions to either myself, to Neil Katz, the CFO, to Sarah Tang, VP of Enterprise, or Adam Byrnes, who's the VP of Product. Chris, if you open up questions, please.
I'm not seeing many questions in the chat at this point in time.
Don't be shy. You can either put it into the chat. Just going to ask the question in a second. I'll note that the OTCQX share price actually has been trading a little bit higher than the Australian share price. It closed last night about AUD 0.8364. Surely someone wants to ask questions. Here we go.
Mike Chen asks. "Why are you so confident your web rankings dropped to the hundreds?"
The number that Alexa shows on the alexa.com website is actually a blended average over the last 90 days. In fact, if you download the top 1 million websites list for Alexa, it'll tell you the daily ranking. The daily ranking is substantially better than the 560 ranking that you see on the blended average. That's why we're confident we'll continue to go into the low hundreds.
I'd like to see the share price get over AUD 1. What do you think needs to happen to achieve this?
I think the company is just substantially undervalued. If you look at globally, you look at peers, you look at the recent IPO that happened in Australia. This year, we'll do about AUD 1 billion in GPV, and the Australian local equivalent that's only in Australia and very spotted exposure in overseas is doing AUD 30 million in GMV. That AUD 30 million of GMV includes chargebacks, refunds, cancellations, sales tax. They've got a valuation higher than ours. I think the share price is going to be substantially higher than this very soon. We've got getting the growth going. I think Escrow in itself is probably worth more than the current market cap of the company. I just got an email this morning.
BizPay is listing in Australia. It's got a monthly payment volume of AUD 7 million. They're valuing the company on 1.5x-2x payment volume for AUD 400 million market cap. We're doing $180 million volume per quarter. We completely blow out the water BizPay in terms of volume by far. We've got phenomenal growth in that business. I think the market cap of Escrow alone is probably worth more than the market cap of Ascend right now. I think it's continual execution, continued results in the numbers, and I think there'll be a bit of wake up in the market. We're going to be a little bit unloved for a little while. I think that's going to change pretty rapidly.
Mick Chan says. "What's driving the improvement in the ranks?"
Simply web traffic. Web traffic relative to everyone else's web traffic. Everyone's working online. The new front-end architecture is performing very well. It's a lot more responsive, a lot quicker. We've made a number of improvements to speed on the site. Also just improving the product as a whole. If you look across all the metrics, if you look at time spent on site, page views, and bounce rate, all those metrics are going in the right direction. People are using the site more. There's more people going to the site, and they're staying out for longer. A question from Aaron.
"eBay integrate beyond watches and motors?"
The answer is yes. We're in discussions with a number of other categories right now. Nothing is committed as yet. There's been at least three or four other categories that we are in discussions with. We'll have to wait and see on that.
You can understand that eBay and other marketplaces have a big problem with fraud, right? If you thought traditionally, why would you buy a watch from eBay?
Well, yes, you get the nice new deals. It's going to be cheap, but are you worried about it being a stolen watch or fraudulent or a counterfeit watch or what have you? That's why there's been an emergence of a bunch of these different watch marketplaces like Chrono24, WatchPayer, WatchBuyer, et cetera, which have kind of picked the eyes out of eBay. eBay has decided to drop a hammer on that and to ensure there is no fraud. To do that, there's the combination of the authenticity guarantee, which means that no longer will a seller be able to send a watch to the buyer.
You'll have to go through a third-party authenticity network or authenticator network, where this is not us. This is a network they've set up themselves with third-party authenticators. They get the watch. They clean it. They open it. They check all the parts. They make sure it's all authentic. They then close it up, put a sticker on it, give you a certificate, and put it in a nice little box if it doesn't come with the original box. The watch gets shipped to you, the buyer. Only then do you release the funds after you've seen it all and a third-party authenticator has verified that everything is authentic and it's all great. You release the money. You can understand that that kills fraud. I mean, that drops the hammer on fraud.
There's a bunch of sections where eBay has big problems with counterfeits and so forth. If we're in watches and we're in cars, I think it's safe to assume we'll be in other areas. We're in discussions, and something might be coming in the future, but there's nothing committed at this point. I think it's a reasonable assumption that we will be into future areas of eBay as well as other similar marketplaces.
Next question is any comment on Shopify using Escrow?
Yes, they're using us for Shopify Exchange, which is the buying and selling of Shopify shops. I have told the sales team I would love to see soon rolling out in Shopify more generally. We have got literally 50 to 100 emails from people saying, "Can you please put Escrow on in Shopify?" There are a number of categories where Shopify has a problem, such as electronics and so forth, where a lot of fraud happens. I'm confident at some point we will be turning on Shopify, but there's nothing happening at this point right now. It is one of the top missions I've told the sales team to try and expand that relationship. We are expanding the sales team to get more bandwidth.
Next question is, do you think Freelancer Enterprise can sell the same stories, Infosys and TCS?
Well, what I will say is one of those names is one of our customers in Enterprise. We are helping power one of those businesses. The demand from them, they've told us is Well, we've got some reasonable numbers in terms of what they want in terms of staffing. We are actually helping one of those names actually grow. I'm not sure, Sarah Tang, if the second one is actually there as well. There's one of the competitors, but I'm not sure which it is. Maybe Sarah Tang can talk about that later on. Would a demerger of Escrow.com be beneficial to realized value? Look, at some point in the future, I've said this before, Escrow.com should be its own independent public company. There's a reason why.
It's a financial services business, you need to be able to have audited set of accounts that are publicly available that anyone can download. Globally, as you go to banks and open bank accounts and so forth, and you're dealing with so many regulators, I think we have regulators about 50 different jurisdictions. You just need to have it independently audited and run like a financial institution. At that point in time, we'd still retain management control and probably the majority of the shares in the company. At some point it will IPO. The question is, when is the good timing?
When we IPO'd Freelancer, it was doing about 18 million in forward revenue, et cetera. IPO'd about 200 million market cap, and it opened on the first day opposite $1.1 billion et cetera. The question is, what size and shape should Escrow look like to IPO it?
At the moment, it will look next year if it continues to grow at the current rate, very similar to Freelancer in terms of both the size of revenue and the growth rate that it did when we IPO'd Freelancer.
The question is, when would be a good time to do something like that? Should we do a private raise first or what have you? There's a lot of optionality.
Obviously, the market's hot at the moment. You're seeing companies IPO at substantially higher values with very little revenue to no revenue. I think that it would attract a very strong valuation at this point in time. It's something to consider. Certainly, the optionality around the corporate activity with Escrow is increasing. We have had some interest in that. We'll kind of see where that goes.
From Aaron, n ow you're onto OTC. Is the goal to go onto Nasdaq Small Cap or Nasdaq Main Exchange? It's obviously something we thought about for a long time. When would you partially float Escrow?
I've just answered that. The answer is yes, at the appropriate time. We're not sure when.
James says, "How do you think about Escrow's take rate over time, say in five years?"
The take rate has been coming down a little bit. Two years ago, it was about 1.45%. Now it's about 1.2%. Some of that's come down because we're doing some larger transactions now. Some of the larger transactions are obviously at a skinnier percentage. Really, for corporate transactions, commercial transactions with Escrow, it's about 89 basis points. That could go down to about 30, 35 basis points for something like an aircraft.
For consumer, the rack rate is 3.25%, going up to 6% if you use a reversal thing like a credit card. There's two things that have kind of driven the take rate down a little bit. One is these large size transactions that are going through in the tens of millions of dollar, $10 million+ . The other is some of the partner transactions we're doing is we're doing a bit skinnier because it's the likes of eBay and so forth. There are some ways we can lift this take rate. For example, if we started doing FX. Some percentage of our volume, for example, is a U.S. dollar transaction where the seller gets paid out in euros, right? At the moment, we're kind of giving up those FX fees to some other financial institution.
There are things like that where that would actually substantially lift the take rate. In some categories, we do offer value-added services. For example, in cars, we do title collection. In the U.S., when you sell a car, you've got to collect the titles and so forth. We do lien holder pay off. We do lease payoff arrangements, et cetera. There are some value-added services where we could lift that. We haven't really been focusing on that, there has been some interest, for example, the likes of eBay, to add some of those value-added services that are specialized into those niches. At the moment, it has been getting a little bit skinnier with the big lift in volume, I think there is some possibility to lift it. We just have to get the bandwidth up to do that.
Group GPV in first quarter of 2021 was AUD 102.9. Fourth quarter was AUD 101.8, so it's not a significant quarter-on-quarter. No, Richard, if you go back to the fourth quarter, you'll see that there's a AUD 50 million lift in that GPV number. If you look at the Escrow numbers for the fourth quarter, there was a massive lift in that fourth quarter. While quarter-on-quarter, the lift is small, year-on-year, the lift is massive. I think that's probably the wrong way of looking at it. As I said before, it's growing a bit faster now anyway, for Escrow, and also for Freelancer. There is seasonality in quarter-on-quarter as you look at numbers, but the seasonality is messed up a little bit by COVID.
The seasonality you have typically is Easter is usually negative and August is usually negative, which is the summer holidays in the U.S., North America, and then you've got a bit around New Year's Eve, et cetera, and then a little bit of Diwali and Ramadan and so forth. A lot of these seasonality numbers aren't behaving like that at all because people are in lockdown. Their habits have changed a little bit. Will we get a BNPL option in 2021 for Freelancer? We have thought about it, actually. We've had numerous discussions with all the BNPL guys. It's on the list of things to do. It's not on the immediate plans next month or two, it is actually on the list to potentially turn that on.
Kevin Chang says, "What's your view on the Freelancer market growth during COVID and post-COVID?"
I think we've got a permanent step change. I don't think the world will go back to fully working back in an office. I think we're some years away from COVID being solved. I don't think the vaccine solution right now is going to be the be-all and end-all solution. You've got a number of strains now, like the B.1.617, which has got antigen drift, which means it's resistant to many of the vaccines. You've got those breakouts happening in India, where the cases are growing by 10% per day. I think it was like 275,000 cases yesterday, up from 230,000 two days ago. I think we're some way away from solving the COVID problem.
I think even if COVID magically went away tomorrow, you're not going to have everyone just go back to the office. I think people like having a hybrid. I don't think people will be fully remote, but I don't think people will be fully back in the office. I think it's going to be a mix, certainly Pandora's box opens once you try using our website and you discover how easy it is to find people, how easy it is to marshal resources, how inexpensive it is. I think there's no going back. There's a question from Kevin Chang.
"What's your view on the freelance market?" from Ray. "Does the lack of share liquidity contribute to the undervalue of the company?"
I don't think so. I think it contributes to volatility, certainly. The share price either goes up a lot or down a lot, depending on the buy sells of activity. It also could equally help us. When there's a bit of momentum, the share price moves pretty quickly up. It is a volatile stock because it is low in terms of liquidity, but I think ultimately, it's going to be quite beneficial to us as we continue to print great results. I've answered Simon's question previously.
Mick Chan, g iven the slow pace to correct the issues in 2019, do you need to spend more on engineering so future issues are corrected faster? What do you think your income or cash flow issue are going to be the best returns? Good question.
Look, we went through a very complicated architectural change in 2019. We replaced I mean, Adam, maybe you can talk about this in a second. We replaced the entire front-end architecture, and we were page by page by page, effectively redoing the website. We did that because the previous architecture was based on four front-end stacks. It was very slow to load. Google increasingly is penalizing slow-loading websites. The user experience is not great when it's slow to load. We got to a point, it was around 2017, where we said, "Yes, we need to do something radically here to bring it up to Facebook level in terms of performance. What do we do?" We spent a long time architecting a solution, and then we implemented it.
The issue is it's a very complicated site. You might think of it as just a site for posting jobs, but effectively it's a country in software, and you have intelligent participants on both sides. It's not like selling books. It's a very complicated site to actually manage. When we went out in 2019 and we started pushing some of these pages, we started with the most important pages first. Really, the problem was that the product managers didn't have a perfect understanding of all the edge cases on those pages. We did push, for example, the project view page too early. You can see that in the Alexa graphs, how in 2019, the growth just kind of stopped because a lot of the features, the edge case features around Alexa about how it works, were not fully implemented.
Now, obviously, as a result of that affected the share price, and we're acutely aware of this. In hindsight, we'd probably have spent some more time really on the CI and the test suites and the QA. We were putting a lot of effort at the same time into the QA. We hired literally a new person to run it who wrote the textbook on software quality assurance, speaks at conferences, and is a Professor at University. He's an ex-Dolby. We were putting a lot of effort in parallel. Being a business that's not massively cash burning, we are constrained by our revenue in terms of the resources we allocate. I am not of the belief of going deeply cash flow negative in this business and losing $ tens of millions a year just to get growth at all costs.
I'm of the belief that we should reinvest everything and just try and break even, maybe marginally grow profit from here over time. I do think the jaws of operating leverage are really poised to open now. You can see with the cash flow of last year, which was AUD 7.8 million, the cash flow of first quarter of this year, AUD 4.7 million. The cash is coming in a big way. It won't take much to get those jaws wide open. Now, the second part of the question is a great question. Where will incremental cash flow go? If you look at our business, there's no CapEx. It's basically OpEx, and it goes into headcount, rent, and marketing. Marketing, we want to generate return quickly on the marketing. We're not a Silicon Valley company spending 200% of revenue on marketing.
We are constantly trying new things and constantly inventing. We've actually made some big improvements to reduce our cost per acquisition, per customer, in terms of how we've targeted our paid advertising. Now, you've got to remember, most of our traffic is not through paid advertising. 66% of all projects at any point in time are from repeat customers. Of the 33%, roughly, that come from new customers, a marginal amount comes from paid. Instead, it comes from things like SEO. It comes from just referrals, inbuilt virality in the product, and so forth. There is some paid. There's some paid in the form of Google AdWords and so forth. There's also some paid in the form of virality. Give, get. I get AUD 20, you get AUD 20 if you refer someone.
I think there will be a bit of a step-up in marketing, but not a lot, because we have tight parameters in terms of generating return. The rest will go into headcount. You can't hire. It doesn't take much from here. If you've got 20% growth, which is a pretty modest growth in revenue numbers of the business, that's a big step up in terms of numbers driving the bottom line, unless you put it into hiring. You really can't hire that fast. In terms of organic, it will be going to headcount, but we're going to have some cash left over, I think, pretty quickly.
Zane had a question, f reelancer was not really mentioned here. Anything happen here?
The answer is yes, and real soon. We'll be talking about something there, but I don't want to preempt that right now. Yes, stay tuned for something there of reasonable size. Next question from Adrian Lee.
What is the position on moving marketing spend to online rather than physical marketing on buses? Seems like the poster hasn't changed in years.
We spend nothing offline, really. You saw the buses in Australia. They were distressed inventory, because COVID had wiped out what have you. Adrian, 4% of our revenue comes from Australia. Four. The rest of it comes from the U.S. The vast majority of our marketing spend, it's seven digits a month, is global, and it's mainly in the U.S. The stuff you saw in Australia was de minimis distressed inventory. The bulk of the marketing is Physical marketing is less than 1% of our marketing spend. It's tiny. On an ongoing basis, tiny.
Sheba, a ny more acquisitions on the horizon?
Look, we're always looking for things, but nothing's happened in the last number of months. I won't talk to that.
Has Escrow, from Trevor, approached AutoTrader or Carvana as clients?
I don't know Carvana. AutoTrader, the answer is yes. There are three AutoTraders. There's the U.K., U.S., and Canada. We are well advanced with one of them. We have the other one in our pipeline in deep conversations. They are three separately run businesses, believe it or not, even though they have the same name. We are well down the track with one. In fact, I wish I could have said something about that six months ago. We will get AutoTrader. Don't worry about that. Adrian Lee,
What are the growth levers for Escrow in the next few years?
eBay is a major one that will grow, but just thinking about further growth. There's a number of key verticals we're basically attacking. There's about eight of them, which are key for us. There's domain names, IP addresses. There are automotive, cars, boats, airplanes. There is fine art, sculptures and so forth, jewelry, gemstones, diamonds. There is import, export, et cetera.
There are a number of these key verticals that we are working on. We're building out basically sales, and we're building out account managers and BDRs, et cetera, to go pursue them. In addition, there's product improvements, really making the product super slick. It's a unique experience. You've got to remember that if you look out there's not really many other Escrow companies in the world, and the ones that are operating have one license, and they're operating illegally. You've got to remember, Airbnb got a letter from the California Department of Business Oversight saying that you hold money for the provision of a service, staying in a vacation rental, and you need to get an Escrow license. Airbnb could go and get an Escrow license. They haven't.
If they get an Escrow license in California, they need one in Texas, Arizona, North Carolina, Alabama, Hawaii, New York, et cetera. The question will be, do they really want to spend a decade chasing these licenses and all the compliance overhead that goes with that? Will they just pay us instead? Increasingly, the regulators are forming a view that marketplaces will no longer be able to take payments, and you need to put them through an Escrow provider, and we have the global online Escrow provider. I think it is an extremely strategic asset for any marketplace business in the future, particularly services-based marketplaces where you're holding a payment for the provision of the service. The regulators are forming a view that is Escrow.
The question is, what are the margins of Photo Anywhere? They seem as if it's actually higher.
It's $35 at the moment for the basic product, to get five photos. Of that, $5 is taken as a Photo Anywhere fee, and that's posted in the marketplace with a $30 milestone where the normal fees apply, which is 3% plus 10%. The margins are a bit higher for the Photo Anywhere. We are going to put into the product, Take a Video, which will be a higher price. We'll add in a checklist. Someone can go to a location, fill in a checklist, take photos or a video, and there'll be other things later on going into that. Okay, that's a good set of questions. Any other questions from anyone else? Adam, did you want to say anything about the product at this point while people think about questions?
You've covered quite a lot. The big headline release of the quarter was the iOS release. Android will be following very soon. Photo Anywhere was another major release. Like I said, I think you've largely covered most of the major product announcements.
Okay, great. I'll open up for a few more minutes just to see if anyone wants to ask any questions. We'll call to close. As always, you're free to organize one-on-ones with us or any of the management team, or do a call at any point in time. We'll come visit.
Peter question, t he update in Freelancer Growth appears to have a lot of potential given the current delivery supply side disruptions.
Well, there's a bunch of stuff we're going to talk about soon with this business. Probably in the next eight weeks, there'll be a big talk about this. I think there's a lot of potential. The average project size in Freelancer is about $4,500. It's about 25 times larger than Freelancer. It is primarily at the moment an Australian business. We do plan on taking it global. There's a lot to say, but not right now about this business. Any other questions? Okay. My call to a close then. Thank you very much for attending the first quarter of 2021 financial results presentation. I look forward to speaking to you, if not in between then, in a one-on-one at next quarter's results. Thank you.