Beauty Meets Fashion: Challenges in Extending E-Commerce beyond Fashion
Joyce Chen20:462023
A free lipstick sample can break a pricing model built for fashion
A lipstick shade called Velvet Teddy has no clean color mapping, and a free sample priced at zero breaks a data model that assumes every price is positive. Joyce Chen, VP of Engineering at Zalando, unpacks these gaps from bringing beauty onto a platform built for fashion, a trap any team faces when stretching one stack into a new category. Which parts should bend, and where should the fix live?
See how Zalando mapped beauty's quirks onto fashion tech, and the trade-offs behind each choice
Joyce Chen, VP of Engineering at Zalando, explains the engineering challenges of extending an e-commerce platform built for fashion to sell beauty. Using examples such as shade-based color pickers, zero-price samples, base price regulation and beauty advice questionnaires, she shows how to break a customer experience into technology building blocks and decide where to adapt the stack. She closes with experience level objectives and an engineering mindset focused on customer experience rather than systems.
“When you look at our color in beauty, it's descriptive words or phrase that are intended to invoke a certain mode of feeling.”
“So this is where you have a choice of adjusting this at the root or adjusting at the leaf, okay?”
What you will learn
Break a new product category's customer experience into technology building blocks to decide what to add, adjust or remove.
Spot where product attributes such as color, size and price behave differently across categories before reusing a data model.
Weigh fixing a requirement at the root (data model) against fixing it at the leaf (presentation), including for zero-price samples and regulation.
Define experience level objectives that measure complete customer journeys, not only services and APIs.
Design recommendation questionnaires around desired outcomes rather than sensitive personal attributes, with data models built for reuse.
Key takeaways from the speaker
Technology that was originally built for fashion e-commerce can be adapted and expanded to include beauty customer experience, and it is important for companies to consider this in order to offer a more complete and seamless customer experience
By breaking down the desired customer experience into technology building blocks, companies can identify areas where they can add, adjust, or remove technology to create and design a more effective and efficient system and architecture.
With the right engineering foundation, companies can build and expand their technology to meet the changing needs and demands of their customers, ultimately improving their business outcomes and staying ahead of the competition.
Details
Level
Intermediate
Format
Talk
Technical depth
Practitioner
For
Mid-career, Senior individual contributor, People manager
Length
20:46
Published
September 2023
Questions this talk answers
Why is color harder to handle in beauty e-commerce than in fashion?
Fashion colors map directly (blue is blue), while beauty shade names evoke a mood or feeling. The stack has to map each shade to a hex value so that filters and searches for colors like red or pink return the right products.
How can a pricing data model handle free beauty samples?
One option is to relax the always-positive price constraint and allow zero prices, which risks breaking things such as monitoring. The other is a rule-based discount at checkout, which keeps the model but means customers only see the adjusted price at checkout, possibly hurting conversion.
Where should regulatory display requirements like base price be implemented?
Either in the data model, by storing the data a certain way, or on the product detail page when it is displayed. Chen frames this as choosing between adjusting at the root or at the leaf.
What is an experience level objective?
Where service level objectives focus on services and APIs, an experience level objective measures a complete customer experience, such as the color picker or beauty advice from first interaction through to checkout.
What are the biggest challenges facing e-commerce companies today?
Chen points to rapidly changing customer behavior under global constraints such as the war in Europe and recession, acquiring and retaining customers, many new competitors, and generative AI such as Zalando's announced ChatGPT-based fashion assistant beta.
About this talk
This talk will delve into the engineering aspects of expanding technology originally built for fashion customer experience to include the beauty customer experience. Through specific examples, we will explore how to analyze the desired customer experience and determine the necessary technology building blocks needed to achieve it.
Why Velvet Teddy breaks a fashion color filter (75 s)
Free samples versus the positive-price data model (74 s)
Experience level objectives and owning the customer experience (88 s)
My name is Joyce Chen, and I am the VP of Engineering at Zalando. Today, I'm excited to talk about a topic that I'm incredibly passionate about, beauty meets fashion. Over the past years, I have been working on this topic extensively, and I'm excited to share with you some of the challenges that come with extending e-commerce. Beyond fashion. So, for the next 20 minutes, I will be diving into the topic and share my insight with all of you. So, before we start into diving into the topic of beauty meets fashion, let me take a moment to introduce myself and share a bit about my journey to becoming the VP of Engineering at Zalando. So, let's move on to the next slide.
Speaker 1: Right. So, I was born and raised in a small town in Taiwan called Douliu. And in this picture, you can see me participating in a math competition during high school. Math has been a passion of mine since junior high school. Speaker 1: And that love for problem solving ultimately led me to pursue a degree in computer science in the United States. And after completing my graduate degree studies, I made my way into Microsoft as an engineer, where I spent 15 years hunting my skill set and learning from some of the best in the industry.
Speaker 1: During my time at Microsoft, during my time at Microsoft, during my time at Microsoft, I was very proud to spearhead an initiative called Douliu. Speaker 1: Deep State. Speaker 1: Deep State, where leaders within the company come together to discuss diversity and inclusion. Speaker 1: From Microsoft, during Microsoft, I moved to Nostrum. Speaker 1: It's a luxury retail company where I was able to apply my skillset in the industry, where I spent 15 years hunting my skill set and learning from some of the best in the industry.
Speaker 1: During my time at Microsoft, I was very proud to spearhead an initiative called Deep State, where leaders within the company come together to discuss diversity and inclusion. Speaker 1: From Microsoft, I moved to Northstrom. Speaker 1: It's a luxury retail company where I was able to apply my skillset and lead a successful e-commerce transformation. Speaker 1: After three years at Northstrom, I joined Zalando in 2022 and relocated from United States to Berlin, Germany.
Speaker 1: These experiences have helped shape me into the leader I am today. Speaker 1: I'm excited to share my insight with all of you. Speaker 1: So Zalando also have a podcast program called Peace of Me, which I was also spearhead and was the first speaker on that program. Speaker 1: So now that's a take from a broad view, bigger picture and talk briefly go over what Zalando is about.
Speaker 1: And so Zalando, we are a leading online platform for fashion and lifestyle in Europe. Speaker 1: We bring head to toe fashion for more than 51 million active customers from clothing for wear, accessory and beauty. Speaker 1: And we have more than 7 billion visits per year. Speaker 1: So customer can enjoy a variety of more than 7,000 brands.
Speaker 1: The assortment of international brands range from world famous brand to local local labels. Speaker 1: And our gross merchandise volume is about 14.8 billion. Speaker 1: So we have more than 17,000 employees in Europe. Speaker 1: And as Europe, most fashionable tech company, we work hard to find digital solution for every aspect of the fashion journey for our customers, Speaker 1: Partners and every available player in the in Zalando today.
Speaker 1: And so we offer a wide and diverse assortment of fashion and beauty from niche to trend across all price point. Speaker 1: So for fashion with more than 1.8 million articles by more than 7,000 international brand, we offer Europe's largest fashion assortment. Speaker 1: We also have delegated content creation for the fashion store by Zalando context solution.
Speaker 1: So for beauty, we have created a bold assortment with more than 35,000 products, which are constantly expanding. Speaker 1: In 2021, we partnered up with Zalando to offer the most relevant assortment of prestige beauty to our customer. Speaker 1: And currently is live in Germany and Italy. Speaker 1: So for beauty, now we look at Zalando and now we zoom in a little bit into beauty.
Speaker 1: So for the beauty, we offer beloved international brand like name a few, MAC, SD Loader, Clinique, Origin. Speaker 1: And we also on the hunt for local brands to make the assortment even more relevant for each market. Speaker 1: So the Berlin based offline beauty store, beauty station puts the beauty savvy customer first. Speaker 1: We offer a tangible product experience through a curated range of beauty products from 75 brands.
Speaker 1: And so then that's the company level beauty. Speaker 1: And so now we zoom in a little bit even more on the commercial aspect on the commercial journey we travel. Speaker 1: So for 2018, we initiate that's the initial launch of beauty. Speaker 1: I remember two weeks ago, we just celebrate a five years anniversary on beauty with the team in the office in Berlin.
Speaker 1: And so we launched this and the 2021 we move on to do market expansion. Speaker 1: So from 2018 to 2021, Zalando has launched its beauty expansion in 10 markets. Speaker 1: Then in 2021 also take one more step in partnering. Speaker 1: So we started a partnership with Sephora to widen and deepen the assortment and especially for prestige beauty brands.
Speaker 1: This is where we really think the customer cares about all these prestige brand beauty brand. Speaker 1: And again, really want to offer all device selection and different ranges to the customer. Speaker 1: So 2021, we also have gorgeous release. Speaker 1: We started building the next level beauty customer experience resulting in several product launches.
Speaker 1: In 2022, we launched the Sephora experience. Speaker 1: So now we have the company level. Speaker 1: We have the commercial. Speaker 1: Next, we are going to the customer experience. Speaker 1: Okay, so customer experience. Speaker 1: We think about it from five area. Speaker 1: First is the shopping experience. Speaker 1: What's shopping experience?
Speaker 1: Shopping experience is we understand beauty specific customer need. Speaker 1: And we build a customer experience supporting the individual job that needs to be done. Speaker 1: Then beauty advice, inspiration. Speaker 1: So for inspiration, you think about example like beauty advice. Speaker 1: People need advice. Speaker 1: We translate this necessity to digital and becoming the starting point of beauty.
Speaker 1: And next is personal beauty. Speaker 1: Few choices are more personal than beauty brand, beauty and hair. Speaker 1: So we help customer find what is best for them. Speaker 1: And then expansion and growth. Speaker 1: We want the customer across the Europe are enable and incentivize to shop beauty with us. Speaker 1: And the last but not the least is strategic partnership and partner program.
Speaker 1: So then I don't want to offer the most for all this beauty assurance through strong partnership Speaker 1: We're leading beauty brands and retails and, you know, incentives to shop beauty with us. Speaker 1: Then now you have commercial and then the customer experience. Speaker 1: Then how does engineering support the success of all these?
Speaker 1: So for the beauty engineering, we look at it on three focuses. Speaker 1: We want to extend the technology foundation built for fashion. Speaker 1: And also, how do we think about operational excellence? Speaker 1: We long-term experience. Speaker 1: We want to make sure those in good quality and also the engineering mindset. Speaker 1: So now let's look at a couple of the, I pick a few challenges to share with you.
Speaker 1: And I hope it's a variable for you and also thought provoking on how you take it back to think about how you light up the e-commerce experience. Speaker 1: So for the challenge, first challenge, beauty, the color picker. Speaker 1: So you think about the color in fashion, the shirts, for example, I'm wearing a blue shirt, a blue carmagee, and then a white shirt.
The color is blue, white. It's a direct mapping, blue, white, green, and beige. Speaker 1: When you look at our color in beauty, it's descriptive words or phrase that are intended to invoke a certain mode of feeling. For example, violet teddy, and you wouldn't get it on the screen. So these are like, you know, pink color, red color, but these are emotional mood and feeling.
Speaker 1: So you think about how you serve this customer experience to the customer from the catalog experience on how you filter, right? Filter by color. And so therefore the tech stack will need to build a filter color and how to map color. Speaker 1: Well, I mentioned like for fashion, the color is a clear map mapping to existing color, green to green, blue to blue.
But beauty, you think about underneath the cover, the tech stack, right? Beauty is a hex number. So it's a hex number because it has a wide range to map to the actual blue and green color. Then for the search, when searching for red and pink, you want the color of velvet teddy or we wouldn't get it to show up.
So the corresponding color would need to be in the search result. Speaker 1: So next one is a size picker. Okay. So fashion size, you think about fashion needs. If you look at the fashion. So like I mentioned, I came from U.S. When I try to buy clothes, you know, here I need to map the size. So in U.S.
I might be size four or six. Mapping here is 32, 34, right? For the beauty, the size actually is more uniform. There is no mapping, right? Meanwhile, the price is different. So in fashion, the price is consistent. It's the same, right? And for beauty, the price varies. So for different size, the price is different. Also, there is the sample.
And shopping in beauty, you realize there are trial, there are sample. And the sample are zero price. So this in a way, break the price model, data model, right? Pricing data model in e-commerce, you think about it. Price has variance. Price, normally the data model is a positive number. You don't have a negative price. So there are two approaches to address this issue, right?
One is you do it, you actually break the price model. And you actually introduce a zero pricing into your data model. But when you think about this, this has a potential breaking, for example, your monitoring, right? And the cost of relaxed, the constraint of saying always positive price could be high. And the second approach is you can implement or think about a price policy.
Accept zero price at the checkout phase, okay? So at the checkout, you give discount, a rule-based discount. And the final price will be adjusted. The downside of this approach is the customer won't see the price adjusted until they check off. So this might affect the conversion rate. Next is a base price. So this is very interesting. Talking about regulation and compliance, this is the first compliance issue I work on when I mapped from U.S. to Germany.
So basically, on the display, visual-wise, this regulation is about you need to display the visual a certain way. So when you think about how you will implement this, how you light up this, how you extend fashion to support this is you have two approaches. You can do it again on the data model side, store the data in certain way.
Or you can do it at the product detail page at the visual side when you are visualizing it, right? So this is where you have a choice of adjusting this at the root or adjusting at the leaf, okay? Next one is inspiring. Inspiring has beauty advice, right? So customers receive high-quality product recommendation based on what answer they have provided.
Two key points. One is we don't ask the customer to identify their skin type or skin color. We ask what shape are they interested in, right? So this is really have compliance, have some of the privacy. And also think about in beauty, it's not about your skin type. It's really more about what outcome you want to get, what kind of shape.
And also you want to do your data model to both the data and also allow the data to flow to the system to allow extensibility and reusability for our questionnaire. So the example I have here on eyeshadow, that's not only advice. We will light it up. There will be more. And, you know, this is about traffic acquisition, the carousel.
So in fashion, we already have a carousel to display what's attractive to our customer. So we want to nudge the more fashion customer to beauty. We will have a dynamic carousel. And then we want to really introduce beauty to non-beauty customer. So this is about leveraging what's available in fashion and adding, you know, easy to attract more customers.
And also making them the one-stop shopping for both fashion and beauty. Okay. So operational excellence, see, I want to quickly touch on this. This is an important concept on experience level objective. So traditionally, you can think about service level objective, service and API. Also, there's critical business operational focusing on customer action. The experience level objective is focusing on experience.
So talking about beauty advice, talking about the color picker. Those are the complete experience from picking to going to, you know, check out how that experience is. Okay. Enjoying mindset. So enjoying mindset. Like I mentioned, there are variants when introducing beauty into fashion. An e-commerce stack, built for fashion. So you want to understand the invariance. You also want to extend the technical building blocks, right?
So you want to build it for more extensibility, for reusability, because there are many interesting customer experience to want to light up. And also the responsibility for the engineering team is for the customer experience, not the system. The system has teams managing those. So these are some of the three takeaways for changing and mapping your engineering mindset.
All right. So I think I'm a little bit past time. I hope I'm doing fine for the time. And thank you for all that for joining me today for exploring the topic beauty meets fashion and the challenge of extending e-commerce beyond fashion. I hope my insight have been valuable and thought provoking. And then you have learned something new that you can take with you in your own careers.
And remember, as women in tech, we have the power to shape the future of e-commerce and drive innovation into exciting industry. And so let's continue to work together to break down barriers and create a more inclusive, diverse and dynamic tech community. And thank you again. And have a wonderful day. The 20 minutes went by so fast for me.
20 minutes. Thank you so much, Joyce. It went by so fast for us, too, since we're very plunged into your very insightful and interesting presentation. I see how passionate you are about what you're doing. We have a couple of questions, a couple of comments. Maybe I will read some of the comments first. Thank you. Very inspiring career path.
Congrats. And we have a couple of questions. If you want, we can take them now. We can still have, you have a liberty to take some minutes now since we don't have the next speaker on the main stage after you. If you'd like, I'd like also to share with everyone who's listening to us that Zalando is also hiring.
So if you'd like to work on the Joyce team or to work with other of her colleagues on the change making work. So explore the opportunities. And here's a question. What are some of the biggest challenges facing e-commerce companies today's e-commerce companies are facing today's? That's a great question. Also, I want to remind the audience to, we are welcoming you at our booth.
And for any question that I did not get, won't get to today for sure, make sure drop in the booth and then ask the question there. We will for sure follow up. So the e-commerce, really, we see the, how do you call it, the micro e-commerce, right? The shop, the customer behavior are changing and they are, the global environment has all these constraints.
So, I mean, I would, the European war, right? And the, the market or are really going through some of the reception as we see. So the rapid customer behavior change are something we need to respond really quick to, right? And so also the, so these are the fundamental big challenges and how we want to acquire a new customer and retain customer.
And remember, we remain an interest where you will think about Zalando and you want to shop with Zalando. These are the biggest challenges. And also there are a lot of newcomers, right? There are a lot of competition coming in and they are all have many different angles. And also the chat GPT, Zalando has announced the chat GPT beta, you know, coming up, right?
So a fashion assistant, right? So these are a lot of big challenges and also fascinating. Yeah. Thank you. All right. Thank you, Joyce, for so much invaluable knowledge and experience from Salmira. Thanking us and many people are thanking you in the chat. Thank you so much for being with us. Thank you for this wonderful presentation. And as Joyce shared, join at the booth to ask your question.
I see some more questions are in the chat popping up. Thank you. Thank you so much. All right. So I will leave the studio. Bye-bye. Thank you.
Transcribed automatically; names checked against the speaker record. Select a time to jump there.
Article based on this talk
What breaks when a fashion e-commerce platform starts selling beauty?
Zalando's Joyce Chen shows how shade names, free samples and regulation test a stack built for fashion, and how to decide where to adapt it.
In short
Joyce Chen, VP of Engineering at Zalando, explains what happens when an e-commerce platform built for fashion has to sell beauty. Her answer: break the desired customer experience into technology building blocks, find where beauty behaves differently from fashion (color, size, price, compliance and advice), and decide case by case whether to change the data model at the root or adapt the experience at the leaf. She adds that engineering teams should measure and own whole customer experiences, not just systems.
Key takeaways
Technology originally built for fashion e-commerce can be adapted and expanded to include the beauty customer experience, and companies should consider this to offer a more complete and seamless customer experience.
By breaking down the desired customer experience into technology building blocks, companies can identify where to add, adjust or remove technology to design a more effective and efficient system and architecture.
With the right engineering foundation, companies can build and expand their technology to meet the changing needs and demands of their customers, improving business outcomes and staying ahead of the competition.
From a math competition in Taiwan to beauty at Zalando
Chen opens with her own path, starting with a photo of herself at a high school math competition in Douliu, Taiwan. "Math has been a passion of mine since junior high school." (0:59) That love of problem solving led her to a computer science degree in the United States and 15 years as an engineer at Microsoft, where she led an initiative bringing leaders together to discuss diversity and inclusion. She then spent three years at Nordstrom leading an e-commerce transformation before joining Zalando in 2022 and relocating to Berlin.
She then sets the business context. As the speaker cited, Zalando serves more than 51 million active customers, offers more than 7,000 brands and more than 1.8 million fashion articles, and has built a beauty assortment of more than 35,000 products. Beauty launched in 2018, expanded to 10 markets by 2021, and gained a partnership with Sephora to widen and deepen prestige brands, followed by the launch of the Sephora experience in 2022.
Five areas of the beauty customer experience
Before touching any code, Chen frames the customer experience in five areas: a shopping experience built around beauty-specific jobs to be done, beauty advice and inspiration, personal beauty, expansion and growth across Europe, and strategic partnerships. On personal beauty she notes: "Few choices are more personal than beauty brand, beauty and hair." (7:51)
Engineering supports these areas through three focuses: extending the technology foundation built for fashion, operational excellence, and an engineering mindset. The rest of the talk walks through specific challenges that show how those focuses play out.
When color, size and price stop behaving
The color picker is her first example. In fashion, a blue shirt is simply blue, so filters map directly. Beauty is different: "When you look at our color in beauty, it's descriptive words or phrase that are intended to invoke a certain mode of feeling." (9:39) A shade such as Velvet Teddy has to map underneath to a hex value, so that filtering or searching for red or pink surfaces the right products.
Size and price flip too. Fashion sizes need mapping between countries, while beauty sizes are more uniform but the price changes with each size. Then come samples: "And the sample are zero price. So this in a way, break the price model, data model, right?" (11:46) Chen sets out two options. Teams can relax the always-positive constraint and allow zero prices, at the risk of breaking things such as monitoring. Or they can keep the model and apply a rule-based discount at checkout, which means customers only see the adjusted price at the end, possibly hurting conversion.
Root or leaf: compliance, advice and cross-selling
Base price display, the first compliance issue Chen worked on after moving from the US to Germany, poses a similar choice: store the data differently, or adapt the product detail page when it is displayed. "So this is where you have a choice of adjusting this at the root or adjusting at the leaf, okay?" (13:39)
For beauty advice, a questionnaire gives customers product recommendations without asking them to identify their skin type or skin color, which helps with compliance and privacy. "And also think about in beauty, it's not about your skin type. It's really more about what outcome you want to get, what kind of shape." (14:05) The data model is built for extensibility and reuse, so questionnaires beyond the eyeshadow example can follow. A dynamic carousel, reused from fashion, nudges fashion customers toward beauty and supports one-stop shopping for both.
Measuring experiences, not just systems
On operational excellence, Chen introduces experience level objectives. Service level objectives cover services and APIs, while "The experience level objective is focusing on experience." (15:22) An experience level objective tracks a full journey, such as using the color picker or beauty advice through to checkout.
Her engineering mindset has three parts: understand what varies and what stays the same when introducing beauty into a fashion stack, extend technical building blocks for extensibility and reuse, and take ownership of outcomes for customers. "And also the responsibility for the engineering team is for the customer experience, not the system." (16:24)
In the Q&A, she names the biggest e-commerce challenges as rapidly changing customer behavior under global pressures, acquiring and retaining customers, a wave of new competitors, and generative AI, pointing to Zalando's announced ChatGPT-based fashion assistant beta.
Watch the talk
Beauty Meets Fashion: Challenges in Extending E-Commerce beyond Fashion, by Joyce Chen (21 minutes): https://www.womentech.net/video/beauty-meets-fashion-challenges-in-extending-e-commerce-beyond-fashion
About the speaker
Joyce Chen is VP of Engineering at Zalando, with a 25-year career in tech that took her from entry-level programmer to engineering lead, manager, director and VP. She is a strong advocate for diversity, inclusion and belonging in tech, and in recent years has combined her passion for engineering with her love for art and fashion.
Questions this talk answers
Why is color harder to handle in beauty e-commerce than in fashion?
How can a pricing data model handle free beauty samples?
Where should regulatory display requirements like base price be implemented?
What is an experience level objective?
What are the biggest challenges facing e-commerce companies today?
Drafted with AI from the talk transcript. Editor review pending.