We have Louis Chen with Perfect Corp. The floor is yours.
Thank you for having me. Thank you for staying with me. I'm Louis Chen, the Chief Strategy Officer at Perfect Corp., and today I'd like to share with you about our company. We are a technology company running a SaaS business model, very much focused on the vertical for beauty and fashion. The company was created in 2015, and it went public in October 2022, so time flies. It has been near 18 months. We are now listed in NYSE under the ticker PERF. Our mission is to resolve the consumer pain point when they go through shopping cosmetic products or fashion accessory products: how can you virtually try on this product through your mobile phones or other devices?
So we help beauty brands on their e-commerce website, on their beauty boutique shops, on their social media channels to use technology, AI, and AR primarily to help them bridge that gap and offer that experience over to consumers. The company has grown since 2015. Now we are over 300 people. The development center headquarters is in Taipei, where I sit. But then the U.S. is our largest market. 50% of our revenue comes from North America. So that's one of the reasons that we choose to list the company here in the U.S., because this is our key market. If you see the chart here, it's not surprising that we are geographically located in really what I call the beauty capital of the world, because our clients mostly are beauty companies or fashion companies out of France or in Tokyo.
The development team is about 150 people, developers, all in-house in Taipei. Then the U.S. is about 30 professionals, mostly business developers, marketing, and project managers. We have more than 20 different products since we created the company, so it's becoming a little bit difficult to remember all of them. What we did is try to categorize that in four different pillars to help you understand our business line. The Skin AI is technology such as skin diagnosis, so using your smartphone, iPad, or camera to do a very quick three-second scan of the face and be able to use AI technology to diagnose up to 14 different skin concerns, right? The state of your wrinkles, the spot, the textures, acne, redness, and so forth.
So we developed this technology, and we provide that as a subscription service to beauty companies or skincare companies or med spas, aesthetic clinics. So there are more and more of these clients that you started to see these digital experiences come as part of their engagement with the consumers. The second pillar, the Beauty AI, was part of our core product, mainly for virtual try-on, and I'll show you some of the product use cases, demo videos, so you get to see that better in action. So if anything, for trying on as simple as different shades of lipstick, Foundation Finder to find the right shade for your foundations, eyeliner, eyeshadow, eyelashes, and so forth.
In the last two years, we expanded our market into the fashion accessories, especially for jewelries and watches, so you'll be able to virtually try on the latest collection in jewelries or watches just using your smartphone. Then for the last 12 months, the product team has been really, very busy on Generative AI, so how we are using that to help consumers create better selfies and better photos. And I'll show you some demo as well. To understand our business and monetization model, we did about $54 million in revenue last year. And coincidentally, it's about 50% for B2B and 50% for B2C. So it's a very unique mix because we are able to develop this core engine once. So it's not two different P&L. It's not two different sets of costs.
It's one engineering team who developed this engine, and we're able to monetize it through two different vehicles. On the B2C side, it's more the freemium model. So we operate about five main mobile apps that people can download for free from the App Store, and they get a seven-day free trial with all the premium features. At the end of the seven days, they have a choice to stick with the subscription. It's about $39 a year or $5 a month. Or if they don't like to pay for the app, that's fine. They'll roll back into a free version, and they'll have the basic functionality for that. The B2C actually is going pretty fast, and I'll show you some metrics in a moment. The B2B is where it's much more unique. It's our relationship with these enterprise beauty brands: Estée Lauder, Coty, Shiseido, Chanel, and so forth.
So we provide them a series of different SaaS modules that they subscribe through their omnichannel deployments. The revenue there, again, is mostly an annual recurring subscription model based on how many modules they subscribe, how many countries they are deploying these services, and how many SKUs they have on our platform. So in a nutshell, today we service more than 645 brands. We cover about 18 of the top 20 beauty groups, the large beauty conglomerate groups in the world. And interestingly, this platform itself is evolving and growing every single day. So it now has over 700,000 product SKUs that they can digitally try. So all these logos, every quarter, when they have a new collection for the summer, for the spring, they bring this product and they digitize it using our software and technology.
And then they'll be able to offer this as a virtual product for people to try, not only on their own dot-com properties but also on retailer channels. Just to show you a little bit of the history of growth there, the number of brands that we cover by quarter, that's one of the metrics we announced in our earnings. So up to the end of last year, it was 645. And you see here the patterns. It has been quite consistently every quarter growing 20, 30 new brands. Some investors asked me how far and how long you can keep going through that to give you a perspective from our research, what we have done. I think worldwide, I'm not counting the indie brands, the very small brands, just the logo, whether global brands or regional brands, there's about 2,500 brands in the world, right?
So we are now covering 645. I think there's still a good runway to go for the next few years. The other metric that we also keep track is the number of SKUs that we have. So I said there were 700,000 SKUs up to last quarter. I think that continues to grow really well because there's always new products. There's always new shades that will come into the market. On the B2C side, again, we operate this franchise called YouCam mobile apps, right? So most of those are young millennials or Gen Z, young female users between the makeover apps, the selfie apps, and the other apps. Again, the other metric that we track there is the number of paying subscribers to this app. I think that's where the revenue is growing. So up to the end of last year, it was paying active subscribers 879,000, right?
So these are, as I mentioned, about $39 per year. 90% of the paying subscribers, they opt in to pay for the annual plan instead of the monthly plan because, again, it's a high-frequency-used product. If they like it, they'll pay for it and end up sticking to the product month after month or year after year. A little bit, I'm not so sure how many of you are familiar with the beauty industry and the value proposition. When we came into the market in 2015, before technology intercepted that, the beauty industry heavily relied on physical sampling and people to test it, right? Because you want to make sure that the shade matches your skin tone. My wife, when she goes shopping, does that, but then she wouldn't dare to put it on her lip, right?
So she's always doing the scratch on the palm of her hand, as the picture shows here. So it's not really the same shade on your lip as your skin tone, right? And there's always sanitary concerns, so you don't want to really touch the sample. And I'm even talking about pre-pandemic. After pandemic, most of the sample has been removed from this and stuff. And also, most importantly, the physical trial takes a lot of time. And women don't like to do that, right? I often consult with my clients, like, "OK, you have 50 shades of different lipstick. And how many of them are consumers using?" And most of the time, they are using two shades or three because they just don't have time. It's the tedious thing. You have to clean it and put it back again.
I think technology helps solve a lot of those obstacles. Right now, you can try 30 shades in 30 seconds. You can try all the colors. So it's not limited to only those few that have samples available in the store, right? So it increases engagement and sales, and it's clean and efficient. But the technology evolution keeps contributing more into this space. So I think we're spending more on skin diagnosis, on fashion jewelries, personalization, and more. It's kind of trendy to talk about AI these days. And I think this is, if I go back in history, we have been infused AI in the beauty tech since day one, right? And that has been our key differentiator back in 2015.
We were not the very first company to do the virtual try-on, but we became the champion and the clear leader in this space because we were the very first company to use machine learning technology into AR, right? So be able to track the facial movement of your facial feature, the eyeline, the lip lines, and so forth, making our virtual try-on much more consistent, much more accurate, and true-to-real than anybody else. And now, along the year, every year or two, you see this upgrade in different AI models and different AI techniques. So we keep evolving this platform to now support how to use AI to detect the shades, to combat the different lighting conditions, the different color temperatures. And then more generative, I'll show you in a moment. So there's a lot of products in there, from virtual makeup to hair coloring to nail products.
So let me just maybe jump right there to see some of how these implementations are done. For example, this is one of our partners that we launched last year with Walmart, right? So when the consumer goes into the Walmart app and they say, "You're shopping for beauty categories," now you find this new entry, say, "Try it on." So here in the video, you see, for example, that product from NYX, and you sort of try it on, and then the camera turns on in your app, right? So now you're seeing your own face, and she's swiping left and right between the different shades, and you're seeing the color on your own face. So this is a very interactive engagement experience compared to just looking at celebrity videos or somebody's model photos because it doesn't look like you. So this is from day one.
When we launched this Walmart, immediately we launched 20 brands, over 1,000 products available for you to try. Then every quarter, we keep adding more brands if Walmart carries them. This is another example of how we are doing that with Amazon, right? If you go to the Amazon app and you try to shop for a manicure product, we use our AI technology to detect the movement of your hand, the fingers, and the nails, and be able to create artificial colors on top of that. It's a whole new way of doing this shopping experience. Here are just more examples of how we are doing that for makeup. Oh, sorry. OK, I'll jump that one. For example, here are also we are taking this AR and turning it into a more advertising format, right?
So if you're browsing the media and you see these Gucci ads, the ad is not just a video. The ads now turn on the camera, ask you for the permission to turn on the camera. And then you maybe start trying either that for the Cartier watches here in the middle. Or if you come in just a second. Yeah, here we go. Right? Or this is with the Sally Hansen experience as well. So I think this AR experience, we have seen this expanding in different use cases, not only in e-commerce for the direct but also potentially for advertising. So you get more reach beyond. So I want to key partnerships worth mentioning here. We've been doing this for 8, 9 years now. And so all these partners that you see here, the global tech giants, they actually all license our technology to do virtual makeups, right?
So this is one of the things that we value a lot to our brands because the brand, one thing that's very important for them is when they bring a new product digital, they need to ensure that the rendering looks exactly the same. It doesn't matter in which platform you are, right? Because that's the value proposition. You're promising the user that this shade and the texture look exactly the same as the real product. It'd be really bad for Chanel or for Louis Vuitton that, OK, if you are on Instagram, it looks one color. On Snapchat, it looks a different color. And if you're on Temu, it looks a different color. The only way to guarantee that consistency is all of them use the same engine, which comes from Perfect, right?
Previously, a few years ago, each of them had their own development in-house, tried to do that. They were not really successful because they were not really focused on the vertical beauty. Whenever they delivered that to the beauty brand, the beauty brand was not accepting it because the color was off. The texture was not right. Up to a point, that, and just for the record, Alibaba became our investor in the Series B, and Snap became our investor in the Series C. They did the follow-on when we did the IPO as well. Some of them became our shareholders in our cap table. Some of them are our strategic partners like Google and Meta and so forth. The value proposition, again, is what we call the Beauty 360 approach. When the brand works with us, we help them digitize this asset.
But because of that unique partnership we have with all these different online or offline retailers, we're able to take their SKUs and make it consistent across this omnichannel. So whenever a user is shopping online or offline, or I call it the shoppertainment channels, they all get consistent effects. And this also created a competitive advantage, really a moat, because it becomes extremely difficult for any brand to leave our platform because of all these different touchpoints that we have built over the years. If you have to replace that, you have to replace all of them at once, which is just impossible. Generative AI, which is really cool, and I've been really busy over the last 12 months working on this.
Just to give you a little glimpse about what we're doing on our mobile apps, anything from generating your hairstyle, hair extension, and wigs, for example, AI makeup transfer. This is something we just launched this month. It's something that the beauty brands all wanted for all this year. If I see a great look on Instagram or on a magazine or on a red carpet, can I try that on my face? The traditional way, they have to take weeks for the designer to create those looks. By the time that you see it, the fast maybe already passed. Now, using generative AI, we're able to take just 1 single image. So I do have 4 examples here. So a very complicated look, right? So I'm intentionally taking this with multiple layers of shades.
And I'll be able to transfer that look into my original selfie, which you see here on the left-hand side. So the effect is pretty good. And that's the first step. The second step, you'd be able to match to the database of products that we have. So you know exactly which brand does produce and sell which products. So you know where to go to buy it and which is the color code, which is the color product to buy. So that's an opportunity for both the B2B and the B2C side. This is another application we have in our app. And it's really super cool. We actually work with brands like TRESemmé on Unilever. So they do a lot of hairstyling products. So you just need to send one selfie into the app, and we create dozens of different hairstyles for you, right?
Again, this is using generative technology. So it's not the traditional way that you overlay an image. It's creating it at once. Then imagine the application for that. You can be in the salon. You can be in other places, right? And hair extension, it looks super natural, right? This is all created by Gen AI. The power of Gen AI is just amazing. Clothing, right? Very, very often, investors ask me, "OK, you guys are very dominant in beauty. The bigger opportunity is on the fashion." Certainly, we're working on that, right? We have been researching different ways of doing that. I think the previous generation technology, the reason we didn't really launch it is because it all looked gimmicky, right? You probably have seen it in the department store. If you were like a robot, that's not going to impress any consumer, right?
If you want to launch that, you have to do it just the same way we did it for jewelry and makeup. I think we saw a breakthrough in generative AI. This is our development we've done about three months ago, that with a single image, I can analyze that reference image and be able to understand, the AI will understand, what kind of fabrics and style you are wearing. So I don't need to get the entire collection of millions of dresses and go into a scanning process. It's going to take time and money. So the retailer is not going to do that. I just need the images that they already have on Amazon, on Zara, on Shein, that I can analyze that. And I'll be able to create, using generative AI, the same way that we're doing AI makeup transfer. I'm now transferring your outfit, right?
If you look closely at the result here, it's not the exact same piece. It's not about creating the SKU and you try to put that on the screen. It's understanding the style and then trying to create an AI model that will be close to. Again, this was about 3 months ago. I actually looked at the last generation just before I bought on the plane. It's getting even better and better. At some point, the AI is going to get so close to the original that you can't really see the difference. Just a little bit preview on what's next after that, what we are already working on, I saw that in the lab. It's amazing. I'm not only creating a photo. I want to create a 5-second video of you wearing that shirt, walking down the runway, right?
So this will be something that I think is going to revolutionize the way that e-commerce is done for shopping. I don't need to—I just need to upload a photo of myself. And my generative AI will create a 10-second video reel of you wearing the latest Gucci or Valentino or whatever you want to buy. AI Headshot, this is another feature that we put on the app. It's very, very popular. I don't remember the last time I went to a studio anymore, right? So when I need my photo for my speaker bio or my LinkedIn, I just put my selfie image in the AI machine. And it generates me different styles with different prompts, right? So I can choose to wear the tie or not the tie. And even hairstyles change.
If I want to go even a little bit more artistic, this is another feature we launched recently in our app called AI Studio. This is my colleague. She's working in Taiwan. Look at these images. These are almost like professionally done. But these are all created just by AI. I wouldn't dare to be taking a photo with a leopard or a wolf. But she can do it with AI, right? It looks like it's just very fashionable. I love it. You're going to be amazed how many consumers are willing to pay a subscription product to have access to that just because they can share socially. It's just amazing. There are more different AI for AI replacement, AI removal. It's just a lot of creativity going on in that field. OK, lastly, just a quick note on financial highlights.
Again, we just filed our annual report, 20-F, a couple of weeks ago. So welcome to follow that with us. Last year, we did $54 million in top line. Last quarter four was a very good quarter, 28% growth for $14 million in that quarter. In terms of gross profit, the whole year gross profit margin was about 81%. It slightly declined versus the previous year, mostly because our B2C business is growing at a faster pace. B2C business, we have to pay the Apple tax and the Google tax, right? So it's 13% there. But the B2B is a very high-margin business. We don't pay royalty to anyone. We develop everything in-house. In terms of the bottom line, adjusting the income last year was $7 million, about 13% in margin. Net margin, I think, was $4 or $5 million. So good business.
Another thing I would like to note is our cash flow generating machine, right? So last year, our operating cash flow was $13.6 million in the full year, the audit report, right? Compared to the previous year, it was -$3.3 million, right? And previously, it was the year of the listing. There were a lot of fees. So I think it is a nice business model where we are contractual mostly, our recurring business. So we get paid in advance. And we deliver the service for the next 12 months. So it's a good cash flow generating machine. Last year, the growth line I mean, last year, the first half was challenging. I think many enterprises were unsure about the future of the economy. So enterprise spending was lower. But the second half actually recovered quite well. So I don't see people talking about any hard landing anymore these days.
So for this year, again, we have been beating our consensus for four or five consecutive quarters. So I think that's good. So we always try to be conservative on our guidance. So we are estimating about 12%-16% in top line. Actually, the pipeline looks a lot more robust. I think the pipeline looks a little bit like 20%. But on IFRS recognition base, I think it would be somewhere in our 15%-16%. So my last slide here, where do we grow from here? There's a lot of opportunities. Beauty SKUs are a very big opportunity for us. So we want to deepen our penetration in those top 20 beauty groups, get more brands, more cross-sales across more SKUs, more countries. But in the last two years, we're spending more into skincare AI because, again, it's not just a skincare product brand.
It's about the masstige, the different channels. It's more the regional business. So we're spending that in more countries. Penetration in the fashion jewelries, this is very high demand, high-quality luxury business that we are now having Hermès, Cartier, Louis Vuitton, Chanel. They are all already on board on our platform. This is great. Keep innovating on Gen AI features. And lastly, our balance sheet, which I didn't touch, we are now sitting on $154 million in cash, right? So because it's a money-generating machine, and we have, we raised $110 million to $115 million in the IPO. We haven't really used that any more yet. But the idea was to look into synergistic M&A opportunities, right? Something to help us accelerate our growth. As a SaaS company, at least in the US, I think it's important to get as quick as we can to $100 million ARR, right?
Doing that organically or through other opportunities. Everything is on the strategy. So with that, thanks for listening. And if any minutes for questions. Partnership and also the investors, did that come with exclusivity in any way?
No. Yes, at the beginning, right? So it was exclusive only in China, right? Tmall is the biggest beauty merchant. So it was obvious to have that as part of the deal. They invested. But that exclusivity ended at the time of the IPO, right? So the preferred share were all converted. And so all the exclusivity went to us. So in China, we're working now with TikTok, Douyin. And we work with WeChat as well. Yes.
Tom, the B2C side, can you speak to customer retention and LTV?
We haven't released that in the metric, in our regular earnings. I think when the time is right, I think we're going to announce it. The B2C used to be a very small part of the business, like 10%, right? So it wasn't really a key focus. But in the last 2 to 3 years, it has been growing so fast. So we're seriously looking into the year. I think from the data we track compared to different comps and different peers, I think it's about the average of the same category.
We have time for one more question.
Yes.
As you continue to grow on the B2B side, what do you see as the main hurdles that you have to overcome to manage that quality?
I think what we learned from the enterprise, and especially beauty and fashion, they are not a very fast company to work with, right? Some of them are even over 100 years old companies, right? So I think sometimes I wish they would really move faster because the change is coming. And they do need digital technology. And if you look at their earnings, they spend 0% to 1% of their revenue in R&D, right? So they do not have in-house capability. That's why the opportunity comes for us. We're becoming really the engineering arm for them. So I wish that everything changing in this world, this enterprise business, will move slightly faster. And then we will grow with them. Thank you very much.