At the start of the year, Gwan Yip, CEO & Co-founder of Code & Craft, had a chat with our CEO. Gwan is a seasoned technologist with experience in VR, AR & MR design and development, online marketing, e-commerce, user experience design and web development. He discussed VR and AR at great length and with such passion, that even our CEO was often at a loss for words.
Our Conversations series launched in 2017 and continues this year with just a select number of interviews, held with true innovators and entrepreneurs in our industry.
Name: Gwan Yip
Occupation: CEO & Co-founder, Code & Craft
Likes: Family, friends, my toy Pomeranian dog, 1am bowling at the Gutter and PUBG
Words to live by: ‘If you’re always prepared you never have to get ready’ – Will Smith
Mark Harrop: Gwan, to start could you give me the ‘elevator pitch’ on you and your business? How did you get where you are, and where are you going in 2018?
Gwan Yip: [Laughing] I’ll do my best to keep this short. My background is in e-commerce. I’ve worked my entire career in technology – the majority of which was in e-commerce for the fashion and retail industries. It all resulted in myself being tasked with starting an entire e-commerce division for a Maggy London International (a women’s dress company) that, historically, didn’t deal directly to the consumer, but knew they needed to get into the game. That’s how my path crossed with theirs.
A few years ago, I left that job as I felt I was getting pigeonholed into a corporate role, whereas my background is a lot more ‘hands on’. I also saw, from a technology perspective, that things were moving very, very quickly. I hadn’t heard of VR or AR at the time, but I wanted to get more involved from a practical standpoint. Specifically that meant getting very involved in user experience design and web development – which were two things I was really interested in, but didn’t have the time to dedicate to them.
Mark Harrop: As well as the user experience and design side of things, when you talk about ‘practical’, where does that go? Are you talking about the ability to code, or at the business process level?
Gwan Yip: At a point, it was code. Because I felt that I was fortunate enough, given the opportunities I had quite early on in my career, to have had a lot of exposure and some experience in the practical applications when it came to business processes. Because I was tasked with setting up an entire division for e-commerce, it was really down to me to put into place all of those processes and practices. I really do think that experience has given me the kind of insight and empathy I’m using on a daily basis when talking to brands and retailers about the adoption of something like augmented reality and 3D – because I was there doing the same thing with e-commerce not that long ago.
[Laughing] It was quite a huge job to give to somebody who was so young at the time. It started off with the COO and I, and we grew it to a ten-person team on-site, split across five different divisions. It was not a small project, and we managed to get it into a very good place during the three years I was there.
To answer your question, that’s where I got a lot of hands on experience of the business processes, behind the scenes. When I quit, it was really to get more hands on in the development and technical design side. From a technical perspective, in order to “innovate” you really need to have an understanding of how it works. If you’re fortunate enough to continuously understand how things work, you can relate that back to how you’re going to implement that within the context of something like the fashion and retail industries. And that’s really how we started Code & Craft.
Mark Harrop: I’d like to support that point. One of my benefits is having skills on both sides of the fence, and that’s the kind of thing that really matters.
So, let’s move on to starting the new business. What was, or is, the goal and where did the inspiration come from?
Gwan Yip: When we started the company, we originally tried positioning ourselves as a third party R&D team to help implement technology into e-commerce companies – so it really was an extension of what we just discussed. Closing that gap, that cycle time, to develop ideas, build prototypes and test that within the environment of our clients was really where we saw a lot of the value that we would bring to the company. That was the genesis behind starting the company, and what brought me and my business partners (Michael Giuliana, CTO, and David Lee, COO) together.
We did that fairly successfully for a few years, but about three years ago I bought the Samsung Gear VR headset, put it on, and was completely blown away. At that point, I knew that was something I wanted to dedicate the rest of my career to.
Mark Harrop: That nirvana moment.
Gwan Yip: Exactly. For me, it completely surpassed anything I could imagine, for where that particular technology was at that stage. I took the Gear VR back home with me to England for Christmas, and did this kind of ‘guerilla testing’ and I put this headset onto everybody I saw: my dad, my sister, my grandmother, my niece – essentially anyone from a 13-year-old kid up to a 72-year-old grandmother – and they all had the same reaction, which was an emotional response to what they were seeing. For me, having worked in technology for a while, I’d never seen something that was so cutting edge, so easily experienced and adapted by everyone. That was the wow moment, and what really kickstarted my personal journey into VR and AR development and design.
That was nearly three years ago. The first year I really spent trying to familiarize myself with the technology, with the theory behind working in 3D and virtual reality, and being able to design and code in VR. It consisted of a lot of hackathons [laughing], which I swore never to do again because I’m definitely too old to stay up two days in a row! It was a lot of fun, a great experience, and I met a lot of incredible people. Actually, one of the VR developers who’s working for us, I met at the MIT hackathon; we got along so well we ended up working together, and he’s now employed by our company, which is fantastic.
Mark Harrop: How far have you taken that moment, and thought about what you’ve done in the past? And how do you see some real, practical examples with the use of VR and of augmented reality? On a personal level, I’ve always thought there’s room for AR in Fashion.
Gwan Yip: That’s a great question …that I hopefully have a good answer to! The reason I always start this story off by going into the VR side of things is because that really was the catalyst that got me into this industry, but after about three to six months of building VR experiences for the fashion retail industry, pitching them to people, and presenting and talking to contacts, the feedback I was getting was consistent. People thought it was really great, but they wondered how they would get it into the hands of their customers – whether that’s internal customers, business customers, or end users – and I didn’t have a good answer for that. A large part of me knew that was going to be one of the challenges with adoption of virtual reality in general.
Mark Harrop: It’s the same challenge with any new technology. People can’t see the real use case for it, so if we can show them the use case, with the return on investment alongside, they will come.
Gwan Yip: The additional challenge of VR specifically as a medium, as you know, is the dependency on equipment. So, I was at a VR conference just last year, sitting next to a gentleman who worked in the education space, and we were having the same conversation: he loved VR, but said there was no way he would ever be able to bring it into the classroom because of the challenges we just discussed. He asked if I’d ever done any work in augmented reality because that was something, to him and the industry, which was of definite interest. At the time I had no idea how to build any augmented reality apps, but I of course told him I could [laughing]. So, naturally, I ran back to the office after the conference, and proceeded to Google, “how to build augmented reality apps”. Fortunately, you use the exact same technology to build AR apps as you do VR apps.
We really started to build AR apps at that point, and we’ve never looked back. All of the benefits we discussed – in terms of accessibility, collaboration, ease of use – coupled with Apple’s release of the ARKit and Google’s announcement of ARCore, are going to make mobile AR so much more accessible. And, this is a point we can come back to later, but I was reading your 5th Edition on 3D (which was fantastic by the way), about this idea of “seeing is believing”, especially when it comes to 3D. To have the power to pull out your phone mid-conversation, and say “look at this”, has changed everything in terms of where we feel the opportunity is – especially in the practical applications of this technology to the fashion and retail industries. And that has really been the driving force behind why we’re really focusing on, specifically, augmented reality experiences – at least for now – in the format of native mobile applications, powered by our photogrammetry pipeline process that we’ve developed in house, which can produce photo-realistic content at scale.
That’s the other big challenge we acknowledged quite a while ago, especially with my experience working in the fashion industry. More from a 3D perspective, access to content or the ability to produce content is going to be one of the biggest challenges for adoption. Typically the fashion and retail industries haven’t really worked in 3D like the hard surface industries. You can lie out this fantastic technology, but it’s up to the brands and retailers to power it with content they have to provide. It’s a really big ask, and is kind of impossible, at least for now.
Mark Harrop: Well, footwear 3D has been around since the ‘60s, and really took off in the ‘80s. The vast majority of footwear in this world is designed in 3D. The photo-realistic true-to-life products in footwear are common – from Nike, adidas, Under Armour, Jimmy Choo and so on. There isn’t an issue in footwear.
Ever since we produced the 5th Edition that you mentioned, we’ve seen a surge. Numerous companies in our space are developing 3D further. We’re still in the early days on the apparel side, but the content is coming.
What we want to try and do is to make sure people are getting in early and using the best possible technologies to scan; unless we scan in, at the highest possible resolution, we’re unlikely to get true-to-life product. You can’t make a silk purse from a sow’s ear, as they say.
I wouldn’t argue with your point, but I just wanted to share that we have come a long way, and we’re now entering this space. The technology is there, and the infrastructure is in place, on the cloud, that is capable of looking after the storage (when it comes to these hi-res, large scans). It’s nothing to be frightened of.
Gwan Yip: It’s about finding a solution that’s right for them today, as well as in the future.
Mark Harrop: Absolutely. It’s about picking what’s available now, so it’s future safe.
Gwan Yip: I like that term: future safe.
The direction we originally had, and still have to some degree, of the technology we built – which is augmented reality native applications for your mobile to produce augmented reality experiences powered by the scans we produce – was all targeted in a B2C environment, targeting the end consumer. It wasn’t until fairly recently, when we started working with the Fashion Tech Consortium – which is an organisation in New York helping to develop relationships between start-ups and established fashion brands and retailers – that they started to open our eyes to the opportunity to bring what we’ve developed further forward into the product lifecycle management of a garment. And this is when I started to learn more about the history behind the fashion industry trying to incorporate 3D into their internal processes. What I discovered is …not to quote you back to you, but in the 5th Edition (I’m coming back to it, again) you explained that the ultimate goal is to completely remove the need to produce physical samples – that’s the Holy Grail. The challenges with that are associated with bridging the psychological gap, and also the skills gap, to product something in pure 3D that is good enough to use as a sample for the design team, the manufacturing team, the buyers and so on. I’d love to get your feedback on this. I’ve spoken to a few people within various departments who, unfortunately, cannot get over the ‘fakeness’ of looking at these garments in 3D. That’s not to say that isn’t the future, as I think it will be, but there are certainly challenges we need to overcome – from a technical, graphical, processing standpoint, but also from a cultural and societal understanding of what 3D is. That’s the long tail objective.
Mark Harrop: I’m in full agreement with you. I’m very pleased that the industry has started, and I’ve been very clear that footwear is a solid product, and works well. A flowing silk garment, with all the creases, shadowing, stretches and drape, is a challenge. But I have spoken with a whole host of companies who absolutely love working in 3D for apparel.
It was the same with PDM and PLM in the beginning. There was a lot of initial skepticism, but the market leaders started with it, and went through the pain, and now everybody has PLM. It’s a journey. Will we get there? Sure. Will there be hurdles? Certainly, but it’s a start.
Gwan Yip: Sure. And I think the approach we’re taking in terms of producing photorealistic assets, is very different from something that other 3D vendors are doing. For us, we need some sort of physical garment to act as the base scan. The benefits of that are obviously that we can scan in the detail like draping, how it’s structured, and how it fits on a mannequin. The two main implications of this process are that we don’t capture the physics of the garment (like seeing how a garment adapts to a model’s body walking down the runway), and the dependency on actually having that initial garment to scan in. Once we do have that garment, we can do pretty much everything else in post in terms of producing photo-realistic, high quality materials that we can then layer on top of the actual silhouette itself.
All of this is done in a highly optimized way, because we’re leveraging a lot of the best practices from the video game industry. In order for all of these assets to be rendered in augmented reality, they have to be rendered in real time. The file sizes are small enough (we’re actually talking megabytes) to be sent and shared across the web. All of the backgrounds in my team are from the web, so we constantly think in terms of app storage, open APIs etc. So when you layer in the opportunities and accessibilities of mobile AR, and then the ability to collaborate in real time from anywhere in the world, that’s when it gets really interesting.
Mark Harrop: Now, you’re talking about augmenting to the scanned product. So, let’s say we have this scanned white shirt. You’re then going to augment onto it. Are you, in your mind, talking about augmenting a new print on there, and/or specific styling details? Or, are you talking about augmenting almost off the shirt with a table (or something similar) displaying size details, colour details and so on?
Gwan Yip: Both.
Mark Harrop: So, in some respects, you’re talking about a creative design solution with different pieces of information augmented over the top when you hover over certain areas?
Gwan Yip: Exactly. Towards the end of our promotional video you can see a partial demonstration of what augmenting the model with additional content and information would look like. So the use case that we’re working on is, say, if the design team in New York and the manufacturing team in China are trying to add notes and information to the actual garment itself, you would be able to do that through AR.
So you’d have the ability to add these hot spots on a model, and then assign information to those hot spots.
Mark Harrop: Yes …the scale looks very good as well.
Gwan Yip: Well that’s another benefit of working in AR: scale is something you can manipulate very easily. Part of what we wanted to show was that if you’re sitting around a table with a group of people, you don’t necessarily want a full-scale model, so you could shrink that down to 50 or 40 percent, so you could see that within the context of where you’re talking. But you could also easily flip that up to true-to-scale, and get a sense of what it would look like in real life.
Mark Harrop: You could even augment into a different setting – like a beach for a model wearing swimwear.
Gwan Yip: Exactly. The reason we’re taking this approach is partly because we know that we can scale the production of this type of content fairly easily. And we know that, with the demands of fast fashion and in terms of the size of the market, for this service to be successful, it needs to be able to scale. The difference between using something like what we develop versus a 3D software program, is that there’s still a human expertise dependency on building something from the ground up in 3D, and making sure that the end result (the 3D asset) looks great.
Mark Harrop: And with your approach you’re working for ‘the many’, whereas other 3D software solutions work ‘one to few’. You’ve got to be a fantastic driver of that technology to get the best out of it and, sadly, not all of us possess those skills. You have this ‘one to many’ approach where it’s affordable, it’s easy, and can be used by everybody.
Gwan Yip: Exactly. And to reference your 5th Edition just once more, the requirements, from a skills perspective, of being able to make something truly beautiful and good enough to replace a physical sample, there are so many dependencies – from a business standpoint, in the value of seeing a new type of technology. And that type of technology has a dependency on re-training, hiring, and re-organizing. These are the challenges of adopting tomorrow’s technology today.
To add to that, the success of a technical design team adopting a new technology is always going to hinge on the anxiety of “will it replace me?” The inherent ‘touch and feel’ aspect of being able to drape something, and cut it, and do all of the things that they need to do to build the foundation of a great product is still going to be there for the foreseeable future. With the approach that we’re taking, it’s really a win-win scenario. Their design team, technical design team, their patternmakers can actually build the base of the garment which still needs their hands on skills; but you can then scan that in and provide them the kind of flexibility that 3D can.
Mark Harrop: So, where are you today in terms of your development? Do you have an offering yet? Any early customers?
Gwan Yip: Mmhmm. Where we are today is that we have the internal photogrammetry pipeline process down, so we can scan in (and have been scanning in) products easily. In terms of customers, we have an ongoing pilot with Maggy London at the moment. We’re very lucky in that respect. I have an ongoing relationship with the company (having worked there previously), which has helped in understanding some of the pain points they’re going through, and in adapting our service to providing a solution for them. We built the application they’re currently piloting at the moment.
To be fully transparent, we spent the second half of last year really going after the B2C AR applications, but the challenge that we had was the fact that, when you’re talking about a B2C mobile AR application, you run into a whole suite of issues that have nothing to do with technology or your offering. For example, where does it fit in from a budget standpoint – is it the technology department or the marketing department? Do they have an existing app? Is this going to be integrated into this existing app or will it be a standalone app? If it’s a standalone app, how do we support from a marketing effort? And so on.
The response that we got was incredible. We’ve had some fantastic conversation with some very well known, large, brands. Everybody loves what we present, but it then becomes more of a business issue around how we put a test case together. When you’re talking about a B2C environment, it tends to fall into this promotional category.
Mark Harrop: I can see that. On the other side, going upstream to the supply chain, there’s a value.
Gwan Yip: Precisely. And that’s the thing. We’re such die-hard believers in the potential of this technology; we’re adamant this isn’t a gimmick, or a marketing or promotional campaign. But when you’re talking about a B2C context, that’s where brand’s heads naturally go.
Mark Harrop: I really understand that. I can understand the B2C issues. And I can also understand the value to be had here. In that respect, have you yet defined the platform, and the pricing and that offer?
Gwan Yip: We have. We have a base pricing structure associated to what we’re doing, but it’s highly dependent on what the potential customer or client actually wants or needs. If you look at it from a high-level perspective, there’s the actual content generation component to what we provide – which is like a per-product pricing structure dependent on the complexity of the product – and there’s the functionality in the app that the customer would want us to develop. We have a very baseline app that we can offer fairly cost effectively (which is the app in the promotional video for Maggy London), which is basically uploading all of the content information, assigning a 3D asset, and then spitting out an iOS or Android app. But if, say, the customer wanted specific functionality, we would need to build that for them.
Mark Harrop: How does your customer get content onto the image quickly? Or is that part of what you call the scanning piece? Fast fashion is so quick today, that retailers and brands can’t be waiting for a third party to be adding data in.
Gwan Yip: In terms of annotating the model, that’s something the user would be able to do on the mobile app itself – 100%.
I do have more to say on the development we’re doing but, unfortunately, at this time it’s still under wraps. Hopefully I can share it with you very soon.
Mark Harrop: Sure. I think we could go on forever (which those readers who know me will know that I’m very good at!) but you’ve given us a very well rounded look at Code & Craft.
*Keep up with Code & Craft here.