Cutting Apparel Waste With ZyseMe (Bobby Östberg) | Fashion Made Podcast

Hi everyone, welcome to another episode of Fashion Made. On this week's show, I speak with Bobby Westberg, who is the CEO and co-founder of a company called Sizeme, and we talk about sort of what their technology does. They're taking a slightly novel approach to taking customer measurements and empowering things like on-demand manufacturing with those measurements and no scanning, all that kind of stuff. So should be a really, really interesting conversation. Um, obviously we don't just talk about the technology, uh, we're going to touch on some of the wider issues around customer data and the sort of differences between sizing, fit, and style, which I think is really interesting. And, um, obviously we mentioned sustainability as it's such, um, a permanently boiling hot topic within the industry, rightly so. So look, here's the conversation I had with Bobby. I hope you enjoy and I look forward to seeing you All the best. Hey Bobby, welcome to the show. Nice to speak to you.

Very nice to talk to you too.

Yeah, excellent, excellent. It's really, really cool to hear a bit more about Sizeme. You're obviously the CEO and co-founder there.

Yep, that's right. So exactly, been doing Sizeme, actually grew out from a B2C business that we had before. So now we are focused very much on the B2B model where we connect consumers or brands' consumers with their factories.

So how have you digitized that?

No, we, we have— we started off with essentially a questionnaire. So, uh, it started with algorithms where we ask a number of questions and we could, at the high accuracy, predict individual body measurements. Um, obviously there goes a lot into not only having someone's body measurements, but also what would actually that mean for the garment measurements. So that resulted in us building artificial intelligence to better, more accurately predict someone's— what they would be happy to wear in terms of the fit and size of the actual garment.

Okay.

That was the starting point and quite frankly, something that has worked. I know I'm a founder, so obviously I should believe in that, but I think we all were surprised about how well that worked. And we see that as well in the data we have from our customers.

Can I ask, were you concerned that a questionnaire might not quite sort of fit the bill, if that's the right kind of phrase? Do you know what I mean?

Yeah.

Was there enough information there?

100%. I think in the beginning we were, quite frankly, we explored. We just said it's I mean, the problem we had with our B2C model back then was the customer loved the product, but it was just really hard to buy from us. So that sort of triggered, is there any way where we can make it easier? And obviously we start looking at the data we had, we start to see what if we asked the right questions, would we be able to use that data to replicate the measurements we got from hand measurements? We then start building a dataset with body scanners to see what that would mean for us to extend our datasets. What we did see though was with the body scanners, when we collected that data, we also produced for the consumers. And interesting enough, we had higher satisfaction rates when we only worked with questionnaires versus people getting a full-on body scan. And I don't think that's because our AI was necessarily smarter or better than, than a body scanner. I think the trick It's that if you— I think there's a very fine balance between the effort and expectation and the product you get. And if you come and you spend time going and getting to a body scanner, you know that someone has perfect measurements of you, you expect a perfect product. And effort you put in is pretty high. So while from our side, while we got really great measurements of body, it didn't tell us much about the preference. Of the consumer. So the mismatch of having perfect measurements, the high expectations, and the output of the actual garment just wasn't as satisfactory to people as when they could answer a questionnaire in a minute and a half and they would get a product better than what they've had before. So I think that is sort of the balance between the effort and the results is what drives satisfaction.

For sure. You know, fit and style do not run in a parallel line. Um, so sure not.

I think that's the biggest learning we had. Even if you are 100% accurate on the body size, it just doesn't result in perfect garments.

You know, I tried Zozo myself. Yeah. I mean, I tried Zozo myself, you know, I got the suit and, um, You know, the, the fit was tight, I will say, but if I was, if that was my style, then it would have been great, but that's not really my style. I'm a bit more laid back and I like things a bit looser. So actually I found everything was a little bit small for my, for my style. So, you know, that's, I guess, a perfect example of why maybe a little bit more information, um, would be useful in those scenarios. You know, body scanning isn't everything. And you could argue whether they had the right algorithms translating the datasets they got. But yeah, regardless, I think you're right. You do need a little bit of nuance there that's preferable, I guess, to the consumer. For me, another thing I'm thinking is maybe sort of how someone— their gait or their stance, you know, how you're standing in one position and how you move. Are two very different things. Do you take any of that into allowance in your questionnaire or into your AI?

Um, today we don't look at, uh, necessarily how you move. What we do take into account is the activity levels of people, so that tells us quite, quite a lot about how the volume is distributed. So the, the questions are typically the basic ones are height, weight, age, um Shoe size interestingly matters, so we see that pretty strong correlation, especially on the length variables. And then of course activity, if you go a lot to the gym, of course that's gonna, it's gonna change your proportions. And then a couple of, a couple of questions related to self-perception, so the shoulders to, to essentially the drop between shoulders and waist. And so on.

Fantastic.

But relatively few questions. The result has been, I would have to say, surprisingly good. That said, of course, there are many other methods, and I think there are many ways to get to science. I think questionnaires, one method that worked well for us, but obviously there are other ways to do this beyond sort of questionnaire and AI.

Yeah, and look, I don't think You know, Sizeme isn't just a questionnaire, right?

No, I think that's one element of getting input, essentially getting the consumer input. So what we really do today, since we pivoted towards a B2B business, and the result of that was essentially us getting a lot better technology than being sort of a brand ourselves, plus a broader market need that we start Experience made us decide that if we can provide a technology that helps innovative brands connect their consumer, getting their wish, essentially getting the wishes and their input in order for them to produce and sell customer-centric products. And of course, there's a whole flow of information and technology goes from consumer input, which the questionnaire is one element of that. And then the whole way where we essentially feeding the right information into the factory and into the cutting machine that orders can essentially go from consumer input to being produced within minutes rather than the current model where it takes weeks and months.

Yeah, so that's really the challenge that we're solving here is the time and efficiency. Challenge.

Exactly, exactly. So we know that the model that the fashion industry is deploying today is going through large volumes, essentially pushing out products. So you have a designer who decides how the garment should look and how it should fit, and then that's pushed into the market. This obviously is moving now with the consumers having much more power that they can influence and co-create the product, and then having that produced essentially on demand as the orders or as the demand comes along. And that's really what we're aiming to do is our bigger vision is providing technology that helps brands provide higher value products to the consumer. And produced on demand. So essentially getting rid of all the wasteful overproduction and doing that by producing based on actual demand. Okay.

And you look, you're trying it out with, I believe, H&M Lab amongst maybe some other people. Talk to me a little bit about the H&M Lab collaboration that you're working on.

Yeah. No, so we work with H&M now for H&M Lab for I think we started about a year and a half, 2 years ago. That was initially a sort of a number of steps towards going live towards the consumer last year. The way that was set up or is still set up is consumer can go on, essentially set up, it's a microsite where the consumer can buy, in this case, men's shirts. They can customize the fit or the size using our questionnaire, and then the product then is sent to H&M's sort of factory where they're being produced 100% individual for the H&M consumer, and then shipped to the consumer so they can get their individual product.

Okay, great. So yeah, I mean, it seems to be working quite well then.

Yeah, it's been great fun. It's been great fun. Obviously working with innovation arm of H&M has been obviously great for us. Obviously been really good to see that it works in an environment like that. The result has also been really good, really good results and feedback from from the customers. So in that sense, super happy and proud of what the team has achieved on both sides.

Yeah, 100%. I mean, you should be. Is there any way that I would, you know, if I wanted to use Sizeme right now, you know, what would I, where would I go and experience this, you know, at least the consumer end of this journey?

Yeah, so one would then be h&m-lab.de. So that's essentially mainly set up for the German market. Other place would be, we're working with a retailer called Hildemar Eckerle, and also in Germany. Depending on when this will air, there will be a few more places where you will find the product. So I think the next coming months and quarters will be quite a lot happening.

Interesting. I'm looking forward to it. Um, have you nailed down, is there any particular, um, you know, is it tops or jeans or, you know, any particular garment that has excelled more than the others or has been bought more than the others? I know that might be a design thing rather than a technology thing, but yeah, I'm just interested to know.

No. So, so we started on men's upper body, uh, and then shifting sort of still, still men's, but now we are Within this year, we will also launch the first women's categories. But I would say so far, obviously shirts is our sort of initial core product, has been I think our staple product so far, but also quite proud and excited what is coming next.

I'm sure women's wear, some people would say depending on which category you go into, but it can be slightly more challenging, I guess.

Yeah, it for sure can. It's a whole different, it's a different purchase behavior. It's also different, I would say different also when you look at the body constructions is obviously different from men. So there are a few challenges that are there for women that may not be as hard for men when it comes to sizing. Of course, customization is essentially the same for both gender. And there's a couple of different methods you can apply also on this, on the sizing side that we are, that we're quite, quite excited about.

Will you change the questionnaire at all?

Uh, for sure. For sure. It's, it's a different process.

Yeah, for sure. Um, I guess no matter if it's men's or women's though, this is a great thing for sustainability as well, which obviously there's, um, such a keen focus on within the industry, has been for a couple of years now. And, you know, everyone it's starting to become more ingrained, I think, in our thinking as an industry. And obviously this helps that.

Yes, no, I for sure, I think the— what I think most people have gotten aware, and quite frankly, I don't— I also think the industry itself hasn't truly been aware about the impact of, of, I would say, the, the, the environmental impact clothing production has. I would say though, one of the biggest drivers is the overproduction. So you see today that a large, a large share of what is being produced are never worn, oftentimes not even sold, but when it's sold on high discounts, people may pick them up, but they don't really need a product, they might never wear it. So we see that reducing, and if we can reduce overproduction, obviously has a massive impact on on the sustainability aspect. So shifting model towards on-demand driven model, I think is one of the biggest improvements that can be done from a sustainability aspect. And just to play a little bit with that, there is by UN, I think estimated between 8 and 10% of the CO2 emissions come from the fashion industry. Around 30% today is overproduction. So if you could reduce that, of course, the overproduction itself contributes to a couple of percent of CO2 emission, which is quite upsetting, to be honest. Things that have been produced and not really wanted. So if we can help reducing that, that's obviously something that we all strive to do, and I would love to be part of that shift.

Well said. It's almost crazy to think that a couple of percent of our CO2 is not only unsustainable, it's completely unnecessary or unwanted.

Yeah, I think that's— yeah, I think unwanted is a great word because one thing is we know other things we do maybe is not great for the environment, but at least we get something out of it. When it comes to to essentially producing something that is not wanted is just pure waste. Rather than, I mean, at least if you fly, you may get to a destination where you can relax or you go to meetings and so on, but production for the sake of producing and it's not—

Yeah.

Doesn't feel right in this time and age.

No, 100%. Obviously Seismic is just one part of a large chain of tools that everyone uses. You know, the, the industry is digital from start to finish in one way or another. Obviously there's a lot of manual processes in there as well, but how does Seismic connect with all of the other tools that people need to use? You know, is it fairly integratable? Is it easy for companies to implement it?

Yeah, no, so obviously there is a couple of different levels. One would be sort of front end. The questionnaire is a Essentially super simple, essentially script that goes in, so that's fast. Then it depends a little bit on the production side. There is a sort of a setup with the factory that is a one-time thing. Typically it's quite a simple process, but obviously here it's a little more discussion with the brands and where they produce, but relatively simple. Obviously we're working to make that essentially one click on all ends. So Yeah, that's the, the idea is that we provide a technology where there is essentially one-click setup into integration to all the required systems. We're not there for all of them just yet, but obviously a great focus for us.

Yeah, absolutely. Well, it sounds like exciting times are ahead. I don't know, Bobby, if there's anything else you wanted to mention on the call, if there's anything we haven't covered quite yet.

Maybe the one thing that I often get asked about is a little bit of why now. So essentially, we know that customization has been a topic for quite a number of years. We know that on-demand production has been discussed for quite a while as well. And we haven't really seen the takeoff just yet. I think what we are experiencing right now is at least At this point, I can look back almost 10 years of being working on customization on-demand production. I think we hit an inflection point where it's a couple of drivers all come together at the same time, which is, to be honest, quite exciting. That will be the technology hasn't really been there until now. So that means on the front-end side, now is to having good visualization tools, just having an easier, more accessible way to get individual measurements. At the same time, you start having more digital flows towards sort of the production side. I would also say that the sustainability start to make the shift where brands start thinking more about different methods. But when it came to on-demand production, it was still sort of discussed, I would say, before the pandemic. I think the pandemic itself have unleashed a whole bag of issues for the industry where a lot of brands and retailers sit on unprecedented levels of inventory, tying up the capital that they need for other things, not knowing what to purchase and produce for next season and so on. So all of a sudden, what used to be, let's say, priority, strategic priority number, you know, 7, has now risen to the very top of the agenda where where if we can produce in a different way where we don't have to pre-purchase and sit on all that inventory, but connect the consumer and actual demand to production, that's something that is a discussion. Well, to be honest, we have it with almost all our clients and factory clients, and we see that the priority have really risen to the top of the agenda. Over the last few months, which is quite exciting for, I would say, the industry. Of course, for us it's very exciting, but I would say also for what is coming next for the industry and various players in this field.

Yeah, 100%. I really like that idea, you know, the inflection point. You're absolutely right. You know, we have discussions weekly, you know, we have events going on and we've done a lot of online events over the last year as well. And It's been fascinating. It's been a fast-moving industry before the pandemic, technology-wise. There was a real shift from 2D to 3D reasonably quickly compared to maybe some other industries which took a slower route that way, but maybe started it a long time ago.

Yeah.

But it's absolutely skyrocketed off in the last year for people to change the way— it's like we've always been having these discussions and now we know we have to. To change. So you're absolutely right about that inflection point. So yeah, it's exciting for you guys as well.

Yeah. And I also think it's quite, quite interesting to have the digitalization started on the consumer side. So essentially the front end, how we sell things, essentially making more targeted offers and so on. And now I think it's moved really all the way back into how we actually produce product. Which I think is where a lot of the value can be unlocked. It's just if you reduce the waste, it's not only for the environment, good for the environment, it's also for the brands because obviously you have less forced discounts and so on. So essentially, at the end of the day, you end up with a— it's a win for the consumer because you get a higher value product, It's a win for the brands and retailers because you have a more efficient production and sales models and is also good for the environment because we don't overproduce as much as you would do in the traditional model.

I guess that's something we haven't quite talked about a lot is the money side of things, you know, can this make a brand and a retailer more profitable, let alone just more sustainable?

Oh, for sure, for sure. There is no doubt. I mean, there is— we know that even at— we know that done right when it comes to production, and if you look at— I think the one thing that needs to be looked at a little different is the traditionally you look at the in price, essentially what you buy the product for, produce a product for, and then what you expect a product to sell for. But if you take a different angle and look what the actual average sale price is after discounts and write-offs, then the marginal increase in cost that comes from customized items and producing that on demand is still a fraction from what you save in terms of getting a full price, or in some cases also premium price on the product. So there's no doubt that the financial aspect is massive. We see that in every single calculation with our customers as well. It's just a lot better. It's a just more efficient model for sure.

Yeah.

Which I think is needed because that's how you drive change. So the sustainability without having the financial incentives is always hard. And I think in this case, they are very much aligned, which is why I'm quite optimistic about the future.

Yeah, you know, they have to be, they have to be. You can't be sustainable if you're out of business, you know. So you still have to make that profit.

Yes, yes, for sure. Yeah, and that's— yeah, exactly, exactly. So at the end of the day, those two needs to work out both, and in this case they really do. If you have the right process in place, you can produce at the marginal cost increase, but you save so much on what you can sell the product for, that it makes on the bottom line, it's a massive increase.

Brilliant. Hey Bobby, thanks so much for talking to us today, it's been a pleasure.

Oh, thanks David, this was great fun.

Yeah, look, thanks to Bobby, that was a really great conversation, and thanks to you for listening as well. That's all we've got time for this week, I hope you'll join us again for the next show. Until then, wish you all the best. Take care of yourselves. Bye-bye.

Everybody get up!

Woo!