Updated: Nov 30, 2021
What is Beauty Tech, and how does it relate to health? Learn about virtual skin care, virtual make up, and explore the psychology of transforming personal appearance as Lygeia Ricciardi of AdaRose interviews Wayne Liu, Senior Vice President of Perfect Corp, a beauty tech company leading the world in Artificial Intelligence (AI) & Augmented Reality (AR) for skin care and virtual makeup.
New technology including AR and AI makes it possible to analyze skin problems and find the right products to treat them from afar. You can also use it to find the perfect shade of lipstick, or see how you'd look with rainbow hair.
While VR makeup has predominantly been used by teens and young adults to enhance their social media images, it's finding a new market among busy adults who want to improve their skin and project a professional image over Zoom and in real life. Learn how the latest tech can make your skin healthier and give you a whole new look—all from the comfort of home. Watch the video (click on the arrow) or read the full transcript, below.
Welcome to the AdaRose video podcast, I'm your host, Lygeia Ricciardi. AdaRose is a community of curious people who are exploring and redefining health and wellness by leveraging the latest technology. Our guest today is Wayne Liu, who's the Senior VP of Perfect Corp, and our topic is virtual technology for skin care and makeup.
Thank you. Thank you for having me. It's my great pleasure to be here, to talk to you and your audience.
So Perfect Corp is the leading company in beauty tech, they're using augmented reality and artificial intelligence to transform the beauty industry, and they have their own suite of apps. I'm sure Wayne will fill us in on this a little bit more, but they have over 800 million downloads annually. Let me just say that again, 800 million downloads annually!
So this is a pretty big deal. And some of the companies they work with, you may not be familiar with their names, but they work with a lot of the companies that are real household names like Estée Lauder, L’Oréal, Johnson & Johnson, and others. So, um, I guess Wayne, with that, could you tell us a little bit more about your background and also about Perfect Corp, for those who may not be familiar with it?
What is Perfect Corp? (1:21)
So, um, Perfect Corp, we've been in the business for about almost six years. So we started the company in 2015 and then our focus is really to utilize technology like augmented reality and then AI, artificial intelligence, and machine learning, all kinds of things to help the beauty industry.
Okay, so currently we work with 300 million customers, "customer" means a business customer or the brand, and then we have 10 billion trials per year through a combination of our app and our customer’s activation. Yeah, since then—actually last year—we started working with a platform company like tech giants like Google Snap. So Google, actually currently if you go to Google and do a search or go to YouTube, if you see the virtual try-out, you can be sure it's powered by us.
So basically that's what we've been doing, and then we have about 250 people, currently 250 people, and we are in 11 countries to serve our customers. Our main actual headquarters for the development of technologies is in Taiwan, and our headquarters are in Taipei, and then I'm running the America business, Perfect America, which is our largest operation.
So that's me, and I've been with Perfect since day one, and then before my job with Perfect, I worked with Alice at her previous company. So I've been working with our founder and CEO for more than 11 years, and right now I'm kind of like a manager of the organization.
Very cool. Would you mind taking a second and telling us a little bit about that prior business that Alice was involved with, and you also? I'm just always curious about how these things evolve.
Yes. So that company was called CyberLink. So CyberLink is a company that has been in the business for about 23 years. So the company is a publicly traded company in Taiwan actually. So basically it was, you know, the CyberLink was, uh, I don't know if some of your audience probably know that the product called "Power DVD" in the old days, people still play DVD if they want to watch a DVD from their PC. So almost every PC or almost every notebook has a Power DVD, or some kind of player, which also was white labeled by Power and CyberLinks.
So that was CyberLink’s business. And then CyberLinks started doing all these video editing and photo editing tools through this year. So right now Cyberlinks is also, um, heavily invested in, you know, like facial recognition and all kinds of, uh, technology. So this was, uh, it's our parent company, but right now we are separate. So they are one of Perfect Corp’s investors, and also founded by Alice, which I worked with her there for more than five years.
How does beauty tech relate to overall health and wellness? (4:17)
Very interesting. So, um, I look at things from the health and wellness perspective, and I tend to think about beauty as part of that universe, but I wanted to sort of get your take on when you think about beauty, do you think of it, or how do you think of it, relative to health and wellness?
Yeah, so I guess beauty, it’s not like you can categorize beauty as a single thing. Like it's pretty much representative of your status, your health, this, if you are happy. If you are healthy, you look beautiful, you know, and of course if you put some makeup on top of it, you look even better. So I guess beauty is a part of your life, your entire wellness, and also probably your lifestyle. So it is just representative of that, one of the presentations of your health status. So that's what I look at for beauty.
I think that makes sense. I mean, I would even go so far as to say that it can go the other way too, in that I remember, um, hearing stories about, for example, like, uh, elderly ladies who would go to the hairdresser and they'd get their hair done and they'd feel fabulous. So obviously, not just for that demographic, but the point is, if you feel good and/or are able to kind of express who you are and feel good about yourself, that that can add to an overall kind of wellness, mental health, and a positive, um, you know, vibe that can contribute to your overall well being. So I think it is closely related.
How is the Perfect Corp software being used? (5:51)
I want to better understand some of what your software can do. Like what are some examples of how it's being used?
So, um, yeah, I’d say in the past five, six years, our offerings tend to expand quite a lot, but let's start with something basic. So that's where we start with virtual try-out, if you see "VT," okay, virtual try-out. So basically that idea is very simple. So you, uh, we try to digitize all kinds of makeup products, including lips, eyes, uh, you know, and consider all kinds of things, and then you don't need to wear the real product, but you can try it.
So our technology is really our contribution to make it true, to real, true to life. So we can see the lighting, we consider your native skin tone because we need to, uh, do a blending, right? So otherwise it's more like a sticker instead of makeup. And so that's our basic technology.
Then we expanded to skincare because most of our beauty customers, they also offer skincare. And then, you know, skincare actually is a pretty big, uh, attractive category, it’s growing pretty fast. So then our customers challenged us to, can you get some skincare products using technology, which I guess, um, skincare is not really about trying. It's about understanding both your skin condition and the product, so you can do a match. So we use a different approach. We use artificial intelligence, which is machine learning, this is possible, right? So it's everywhere, AI and machine learning.
What we’ve been doing with the skincare is, uh, we have an engine, of course. And then we train the engine with data. Some of the data is from our customers, some of it’s really, uh, from more like a clinical study, which are provided by our brand partners or some other partner. So we use this real product, real data, to train the machine. So it's fantastic, so much is done knowing what the skin looks like.
We teach the [the product], okay, first what’s skin? Then we teach the product, here’s a wrinkle. So this is a different kind of wrinkle, we call one wrinkle. And then we call another acne. This acne, we have a black head, we have a whitehead. And then we talk about different moisture levels. So we start to show different examples to teach the machine. So gradually, gradually throughout all this year, we've been feeding it with the right data and then we verify.
So, actually our skincare engine becomes a good dermatologist, or can be a skin consultant if you put it this way. So they, uh, right now you just show your face basically in front of a camera, it can be just your iPhone camera. So scan it, analyze it, within a second. We tell you your skin age. We tell them, you know, right now we can identify like 10 different skincare concerns. And then we tell you what's the best product for you, what's the best routine. So that's pretty much what we’ve been working on recently.
New opportunities with skincare and skin health through artificial intelligence (9:04)
Of course, I tried that and I was curious, I mean, I don't know if you would admit this to me, but whether you gamed that system at all, like, does [the product] give you a younger skin age than you might actually have, so that you feel good? Or is it really, really like honest to goodness, it's trained to try and as accurately as possible assess your skin age?
Yeah. So that's a good question. So from our app, we pretty much, uh, you know, we are not really using the zero to one [scale], so we're kind of shifting a little bit. So the basis is not really zero. So the basis is probably like a 60 or something. So, uh, we can definitely, um, adjust, the number, the way it looks. But it's relative, it depends on how the brand wants to represent it, how we want to represent it. But for the skin age, it's really the number. So whatever number shows, is your number.
So some of you may see that if you try that, you see that younger [age], because, think about this, if you're wearing makeup, your skin looks better and then, or you put some makeup on top of it. Of course, when you scan it, then the age you [get] tends to be younger. So, um, yeah, it really depends. So if you remove the makeup you can also try and see how it goes, but, uh, it's been verified with sound clinical data. So it tends to be pretty accurate.
Are dermatologists involved when assessing skin conditions using this technology? (10:23)
Yeah. That's pretty cool. Have you, um, again, just kind of in exploring the, the lines between sort of wellness and health and healthcare, have you guys worked with dermatologists at all or talked about running into the issue of diagnosing conditions that you think, you know, that could potentially be cancerous or dangerous in some way [such that] you should go see a doctor? How do you navigate those kinds of questions or do you not, not sort of get involved?
We work with dermatologists because pretty much it's like, uh, we need to have access to clinical data, so it can be, the clinical data can be from the brand. Okay. Really or from some institute. Okay. So that's the situation here. Yes. We are working with a professional on the skincare side. Uh, however, when we are going to look at the skin, um, cancer, I guess the potential yes, we can, because with machine learning, the [programs don't] have a preference, so it really depends on how you teach them, how you train them.
So right now, currently we're focused on skincare for beauty, but if we want to go to the professional and medical part, uh, definitely that's doable, but we probably just need to reset all the, uh, the criteria verification. And then of course you probably need to go through FDA approval. So yeah. So do some clinical trials. So I guess that's the direction high-level, technically that's totally possible.
But that's not where you're focused from a business case perspective, which totally makes sense.
However, one thing here is, uh, because our skin direction is that we have a function called the so-called "diary." Okay. So we can actually record it. So by recording your condition, and then you see the numbers, we may not be able to identify your skin [diagnosis], say like, a cancer, but if it's something wrong, like sometimes the color changes and you get some kind of, uh, you know, different texture. We can actually tell that, so we can see that somehow the score just tends to be getting lower, lower, lower—then you probably need to pay some attention to it.
Virtual reality and personalized beauty (12:47)
Yeah. That's very interesting. So you were talking about, again, we were talking about sort of categories of ways that we can use the software. You're saying obviously it's AI, you could train it in a whole variety of ways, but right now you guys, you mentioned virtual try-on and skincare. Are those the main ones? I think I was reading, too, about things like matching makeup. Let's say you wanted to match for whatever reason, maybe a company is doing an advertising pitch and they want like a very precise shade of some color. Um, is that the kind of thing that's available to general customers?
Yes, actually, um, that is the direction actually we are heading toward. We call, um, the AI, attribute, face attribute. So basically we, um, we look at your whole face, like the shape of your eyes, the formation of your eyebrow, and then the shape of your lips and your skin tone, and then possibly your hair color. So we look at everything and we identify say, like 27 different attributes. And based on the recommended combination, of course, the brand, you know, they can tell us which product or which look they want to recommend. So this one is really getting to the so-called personalized ability that we now only look for you to try. We actually, based on your, uh, face attributes, we create a look and then recommend it to you.
So basically that's what we’re doing here. Yeah. We are heading into that direction. And then some of the brands actually are about to launch. We are very close to some official launch commercials available in the next couple of months.
Very cool. So, um, I was checking out your app and I noticed, and I realize you have multiple apps, but the one that I think it was the main Perfect Corp one that I was checking out, there were a lot of people who were live streaming, like beauty influencers, mostly younger women who were really using makeup. Um, a lot of it, I would say was for, it was kind of like for exploration and for fun, like it was like trying to, uh, replicate a look that in some cases was very maybe exotic, you know, like put like stars all over your face or something that's very different from just taking a regular face and trying to maybe smooth over a skin issue.
Um, so I was interested in that and is that, I guess how much are people doing that versus using the tools in more traditional ways to kind of look more like themselves? How much is it like fantasy versus sort of, um, more traditional kind of conservative makeup looks?
Yeah, so actually that's a good observation. So from our app, we have this live streaming offering, um, long since we've been doing this for more than three years, almost four years. So we are always a part of the future. So we have an influencer to talk about different looks or makeup or hair. And then the good thing is, uh, our live stream is different from other, you know, live streams. So we can do live streams in many places, but right now for us, we can do AR try-on.
So when this, the beauty advisor, whoever is the influencer, they talk about the look, or they talk about some color, their hair, or they are pretty much creating something, you can make a push, uh, you know, over the air. And then from your phone, you can actually see that virtual look, and you could try by yourself. So that's more like a more interactive, uh, live streaming. So we've been doing this for years and actually the Estée Lauder company, they are very interested in this technology.
So they adopt, we actually create a so-called art platform. Art means artificial, augmented reality training. So we created this, uh, training platform for them. So they use, uh, their internal training globally, so they can run, instead of flying to the places where they do the real training, uh, they can do this virtually. So they've been using it for more than two years, and now things have changed since the pandemic. Uh, you know, there's lots of places which are closed. You cannot go to the beauty advisor, you cannot talk to them.
So you go to e-commerce, however, e-commerce is more passive, right? So you search, uh, you cannot find too many people that could give you advice just like in real shopping. So, uh, some of the brands are adding this, the live streaming part. So instead of just shopping, they can do, uh, live streaming commerce. So they start to promote the product.
One of the early adopters is Ulta, Ulta Beauty. So if you go to their website, they call it their "Ulta Beauty School." So they have their brand representatives talk about different, uh, products, live streams, and then a customer can react with a real time question, emoji , all kinds of fun things.
So I guess, uh, this kind of thing is really, I will say, compared to traditional AI trials, this one's still percentage wise, is still small, relatively less, because it just started. But the way it grows is extremely fast because, uh, you know, we see from our app, since the pandemic, we see like a 32% increase in people, like watching the show and then the number is still going up. So I guess, uh, that one is more like, I would say that's a trend, especially in the beauty industry.
Yeah. Do you think after the pandemic, when people have the opportunity to go back and step into stores like Sephora or others, do you expect that the trend is going to continue? Do you think it'll level off, or you think this was going on anyway, and that the pandemic just sped it up?
I think there will be co-existing. So it's a complement to each other. So people get used to watching this live streaming, it becomes a habit. So it's part of the entertainment. So right now there is a very famous word, say shopping as an entertainment [shoppertainment]. So you watch all your shopping, just like a QVC, right? So in the old days, so QVC, um, I guess, you know, going into the store and then watching this live streaming, you'll get a totally different experience. Uh, they will come side by side, but I guess the people will continue to watch the show because they just love it if it's entertainment.
What are the demographic trends for beauty tech? (19:23)
Yeah. So, um, it sounds like there's been tremendous change in a very short period. Not only with COVID, but even just in a few years, this is like a whole new industry that didn't really exist before. So I'm curious, it seems like the people who are at the sort of cutting edge, at least looking at the app that I was playing with, it seems like they're mostly, um, like younger women in their, I would say teens, maybe early twenties at the most.
Are you seeing signs that this is something that you think is going to spread to a broader population, either women in an older demographic and/or, you know, demographically diverse in other ways, like men and other, you know, I guess diversity in terms of culturally, ethnically, you know? What are you seeing in terms of, or what do you predict in terms of patterns about where this is starting and where it's going?