The End of the PC: Leadership in the Age of Connectivity
According to the GSMA, 4.6 billion people (57% of the global population) now use mobile internet on their own devices. Connecting those with access to existing mobile internet services could deliver an additional $3.5 trillion in total GDP during 2023-2030, with 90% of this impact benefiting low- and middle-income countries (LMICs). Statista estimates that mobile e-commerce sales made up 60% of all e-commerce sales around the world in 2023. These massive shifts have also transformed leadership and organizational design.
Amrop shares this mobile-positive vision and spoke with two global leaders in the e-commerce space:
- Boozt is a Nordic technology company selling fashion and lifestyle products online. They offer customers a curated and contemporary selection of brands through their multi-brand webstore Boozt.com and their Nordic fashion outlet Booztlet.com.
- Stylitics is an Inspirational Commerce platform which integrates outfitting and product bundling solutions across the entire shopper journey.
Bo Ekelund (Amrop Sweden) spoke with Andreas Juul Hirszhorn, Director at Boozt Innovation Lab, while Bill Hogenauer (JM Search) sat down with Rohan Deuskar, Founder and CEO of Stylitics.
Mobile technology plays a pivotal role in shaping the strategies of both Stylistics and Boozt. Mobile usage continues to increase, driving companies to innovate and adapt strategies to meet the demands of a mobile-first world. Boozt and Stylistics emphasize the importance of AI in enhancing personalization and efficiency, and stress the need to focus on "inspirational experiences" to engage customers and improve key metrics.
A conversation with Andreas Juul Hirszhorn, Director at Boozt Innovation Lab
Boozt is a Nordic technology company selling fashion and lifestyle online. They have transformed their e-commerce operations over the past five years by focusing on native app development and mobile web improvements, like transferring web views to native views, enhancing user experience, and implementing new features. In the interview they discuss these transformations, their leadership culture, where they’ve created a balance between innovation, play and experiment, and operational efficiency, AI and machine learning, which are valuable in process optimization and customer service, as well as challenges, which include adapting to regulatory changes and maintaining a collaborative, innovative work environment.
Bo Ekelund (Amrop): Boozt have recently put a lot more focus on mobile device based engagement and commerce. How has the rise of mobile shopping transformed e-commerce operations and customer engagement?
Andreas Juul Hirszhorn: Since I joined Boozt we have undergone a transformation both due to the situation in the market and to the general adoption of native apps, as well as mobile web usage. Early on the app clients were very much like mirrors of the mobile web, leading to the question: why should people use them? We were not fully leveraging the advantages of a native app and its unique features. Over these five years, we have transformed, but this process takes longer than we all initially anticipated. I wanted things to move quickly, but sometimes you have to take the time to do things properly; focus on the end product and patience is key. It was a strategic choice to ensure a native feel in the app and provide a very responsive user interface. Customers can easily identify if an app feels sluggish, which is often caused by using WebView technologies in a native app. We have transferred everything that makes sense to native views and implemented various improvements. We can now truly start capturing the value and enhance our offerings based on this solid foundation.
BE: What specific strategies has Boozt implemented to adapt to these changes?
AJH: We have created a solid foundation, and now we can build extensively on top of it. One of the recent features we have released is the wallet feature. In the past, the flow on the web was shaped by historical reasons, often leaving users locked out in the browser. You cannot always anticipate that a user is logged in, which creates a lot of issues and flows around that. We want to avoid annoying the user with unnecessary login prompts, so we push the login to the latest possible moment in the checkout process. In a native app users are typically logged in for extended periods. Regular customers who return to the app, say, once a month, will almost always have their login refreshed. This change allows us to expect a much higher percentage of users to be logged in, enabling us to adapt all our features accordingly. For example, the wallet feature is most effective when the user is logged in because that’s where they can access available vouchers and gift certificates. If only 5% of users are logged in, it wouldn't make sense to invest heavily in this feature. However, if 80% of users are logged in, then it becomes worthwhile. To illustrate, the traditional flow for applying a voucher code involved receiving an email or notification with a code, which users would often screenshot or copy and paste back and forth between apps to use during checkout. With the wallet feature, customers can simply click a button, open their wallet, apply the voucher, and avoid the hassle of copying and pasting or searching for notifications. We now consolidate all communications from Boozt to customers in one place. Customers can easily access all communications from Boozt in one convenient location.
BE: In your view, what role does the mobile device play in shaping the future of online shopping, and how does it influence consumer behavior? Can you provide insights on customer preferences regarding mobile versus desktop shopping?
AJH: A growing and significant percentage of customers use their phones for almost everything, especially among the younger demographic. For instance, I wonder if today's teenagers would ever use a desktop device for anything. Phones will increasingly become the primary tool for many tasks, especially with advancements in speech-to-text and text-to-speech capabilities. Consequently, desktop computers may become less necessary. People tend to use computers more for work-related tasks. In their free time—when they would typically be shopping—they are likely to rely on their phones. This trend supports the idea that the purchase moment is shifting toward mobile.
Regarding browsing, it can be done through mobile web or native apps. If a company lacks a native app, customers will turn to the mobile web. If an app exists, there will be a division between mobile web use and app use, depending on what the app offers that the mobile web does not, or vice versa. As for customer behavior, it depends on how companies differentiate their mobile app and web experiences from desktop. If businesses simply scale down their desktop sites for mobile, users might have a worse experience. Therefore, it’s crucial for companies to prioritize a mobile-first design and organisation, even if it feels inconvenient – companies should start with mobile web features rather than desktop, given where the growing population of customers spend their time.
Moreover, it’s also essential to rethink features. For example, now that we can expect users to be logged in, we can personalize content more effectively. Instead of presenting a one-size-fits-all page to groups (like men, women, children), we can segment our audience into many more user types based on the richer data we have. For instance, if a user is logged in, we can display the order status of their current order on the homepage. If they are not logged in, we can't provide that information because we lack visibility into their orders. This capability fundamentally changes how companies must think about the customer experience.
BE: As a leader, how are you navigating the technological advancements and rapid changes brought on by digitization in the e-commerce landscape? What key leadership qualities do you believe are essential to effectively guide a technology company through this evolution?
AJH: The ideal position lies in the sweet spot between being a first mover and a late adopter. So much is happening all the time - just look at the current AI developments that everyone is discussing. Some aspects may be overhyped, and the pace of change is so rapid that if you're an early adopter, you may gain valuable knowledge but often at a significant cost. You can end up spending considerable hours on technologies that either get overshadowed by others or ultimately fail. It's important and a learning experience to experiment with new technologies, but this should only go so far. You need to determine when something is here to stay or when experiments yield positive results. We have found a way to do that. It’s important to keep projects small in scale. For instance, when experimenting with AR, VR, or AI, start small and implement these technologies in areas where they can deliver immediate value if successful, rather than trying to completely overhaul core strategies. Otherwise, you could end up spending six months to a year on something that ultimately doesn’t integrate well with your existing plans.
In my role as the director of Boozt's innovation lab, we have experimented with a variety of projects. In the initial years, the goal was to innovate across many fronts. One of the main challenges is ensuring that ownership of projects remains anchored in the right teams. If a random team presents a new tool that they’ve created without engagement from the core team, it can be challenging to secure buy-in. Whereas if you empower team members to explore and tell them to invest time in creating a prototype using new technology, you can generate excitement and foster a collaborative environment. Often, many experiments lead to valuable learning experiences, even if they don’t result in launch-ready products. Say, one out of ten experiments might yield something promising that you want to build on. This approach allows for rapid failure, enabling you to pivot quickly and avoid investing time in doomed projects.
Initially, we conducted innovation projects within the innovation lab only, and then presented the results to various teams. However, we found it challenging to foster true ownership of those projects. Now, we encourage all teams to experiment and innovate independently, which they do consistently.
BE: When you seek individuals with these suitable skills, do you have any inspiration or guiding principles that you aim to emulate? For example, do you look to specific companies or methods as a benchmark? Since you are testing and iterating as you go, do you draw inspiration from other industries that have successfully implemented similar practices?
AJH: I'm not sure we can argue that Boozt has directly copied their approach, but I personally as a leader find inspiration in the game development world, which inspire most of us in this area. While larger studios operate differently, mobile game development, in particular is quite dynamic. In that environment, many small teams are often hired by larger companies to create a variety of small prototypes. They use a lot of data to determine what works and then focus on the ideas that show promise. It's a continuous cycle of striving for success and proving that you're delivering results, which allows you to gain more resources and move to the next level. Eventually, the company may decide to concentrate its marketing efforts on certain successful projects to give them a significant Boozt.
However, it’s getting much more challenging compared to when I first started app development. Earlier it was possible to hit the jackpot and achieve great success on the App Store. Nowadays, those chances are slim due to the dominance of large companies that have extensive user data. They often have hundreds of teams developing applications and can select the best options based on the gathered data for further investment. This process is similar, just on a smaller scale. You want to ensure that you focus your energy on the projects that show promise from the beginning, avoiding the waste of time on initiatives that are likely to fail.
BE: As a leader, are there other skills or qualities you think you’ve needed to develop in order to become a more effective leader?
AJH: As a leader, it’s essential to encourage a playful environment so to question is essential. While there are many operational tasks that must be addressed—such as ensuring the website and apps run smoothly—this foundation should come before experimentation. Innovation won’t happen if you simply ask someone to generate ten AI-related ideas. Many of those ideas may not be good - perhaps none of them will be initially - but over time, as individuals experiment and become familiar with the technology, someone unexpected may come up with a valuable idea, and that’s what you want to focus on. To foster this culture, leaders must not stifle creativity. Unfortunately, I've observed that in the software industry, there is often a strong emphasis on delivering features, which can crowd out the space for innovation. It’s important to give teams some breathing room and promote initiatives like what we call chapters, modeled after Spotify. These chapters consist of groups of developers who share expertise in specific technologies, such as iOS or Android development, despite being in different teams. In these chapters, the focus should be on technology-specific advancements, encouraging members to stay informed about the latest developments in their fields.
If we fail to promote knowledge-sharing naturally within our teams, improvements won't happen on their own. It’s vital that the drive for improvement comes from the developers themselves. As a leader, you shouldn't push from above; rather, you want developers to feel ownership of their codebase. Instead of pointing your finger at a problematic codebase, encourage them to acknowledge it and collaborate with others to devise solutions. My role is to ensure they have the time and resources to address technical challenges without being overwhelmed by feature demands. It's their responsibility to create an efficient and robust codebase and to resolve any issues they identify.
BE: There are many leadership challenges. How do you nurture curiosity within these chapters? Does the leader of a chapter have a budget or resources to spur innovation within the team?
AJH: We have a rule of thumb where we say we split the time in 50:25:25. This means that 50% of a team's time should be focused on feature development - big features and new initiatives - 25% on maintenance, smaller features, and bug fixes, and 25% on more technical development areas. When looking at a team over the course of a year, the accumulated time for all members should roughly align with this distribution. It’s important to note that not all developers need to adhere to this split, but as a team you should strive to include all aspects. This is primarily the team lead’s responsibility, rather than the chapter lead’s.
In our case, on the App branch, all teams run the Scrum model. The team lead is involved in sprint planning, long-term roadmaps, and ensures that the team does not commit to too many features, collaborating closely with the project manager and the leads of the web and backend teams to maintain alignment. The chapter lead’s role is to facilitate discussions within the chapter, highlight new and interesting areas, encourage team members to challenge themselves, and promote education in different technology domains. For instance, seniors take on tougher challenges, such as major migrations. We’ve seen this in the past, one example is when we shifted from Java to Kotlin.
In Boozt, we have both Boozt and Booztlet apps, sharing a codebase. The architecture is designed very thoughtfully, which is not solely a chapter leader's task; it involves all chapter members, particularly the seniors, who tackle the more complex changes. The chapter leader facilitates these discussions, guiding the team to reach conclusions about how to proceed, whether that means adopting a mono-repo or a multi-repo structure. Currently, for example, we have two different build systems for Android that emerged from the chapter's ideas and direction.
BE: How has the above way of working changed impact on customer experience and operational efficiency?
AJH: We already have a department focusing on insights, which includes significant work in machine learning and data analytics. They initiated many projects before the recent anticipated and now realised hype, so in that sense, we are in a very good place. However, our new generation of AI is different; it is more accessible to everyone, not just specialized departments. That said, we need to be careful about where and how we use it. One area where AI truly shines is in repetitive tasks like generating and translating text. For instance, with the upcoming Accessibility Act, instead of having someone manually add text to all images, an algorithm may handle the initial step, with someone editing the output later. We see and work on many opportunities for process optimization through AI. We continuously develop helper tools to streamline product reviews or improve image handling processes. However, I prefer to have ten small AI tools than one large, all-encompassing system. We identify specific problems that AI can help solve, whether that's organizing financial factors, enhancing product descriptions, or sorting images that require attention and we address them one by one. This strategy enables us to experiment in multiple areas simultaneously, retaining only the successful solutions and discarding the ones that don't work.
BE: Will AI tools for you be more important to enhance the customer experience or for internal operational efficiency?
AJH: I believe it will be a combination of both, difficult to say exactly how it will be split. There are definitely aspects that significantly improve customer experience, such as chatbots, enhanced product descriptions, or personalized product finders that help customers discover items tailored specifically for them. At the same time, we also focus on making our own lives easier. This includes developing faster and improving efficiencies within departments, like finance, which can save significant time.
BE: Do you think the hype around it is overrated, or do you believe it's genuinely here to stay? Will you be using it over the next 10 years?
AJH: All companies will use it in some way over the next 10 years. I do think it is overhyped, but it's definitely here to stay and will improve our efficiency in various areas. However, it won't solve all our problems, and not all developers will be out of a job in five years.
BE: What do you consider to be the most significant challenges faced by online retailers today, and how is Boozt addressing these issues?
AJH: Yes, main street stores are probably more challenged by us. One of the significant challenges is the changing legal and structural requirements, such as GDPR and similar regulations. Big companies like Boozt can survive these changes because we adapt, implement new regulations, and comply with the rules. However, smaller companies may struggle more, as they need to adapt to these extensive regulations, including the upcoming Accessibility Act. Can smaller companies survive these changes? That’s a big question. Furthermore, small retail shops are slowly dying unless they cater to a very specific niche. For example, a K-pop store is unlikely to be replaced by an online store anytime soon, having built an unique community around it. However, random clothing stores may face tough times due to the higher cost of living compared to operating a large warehouse. The efficiency, both economically and in terms of CO2 footprint, benefits larger operations, as we may run more leanly and efficiently without the overhead of a physical storefront.
BE: Is there any question related to leadership and the challenges ahead that we haven't touched upon yet that seems important?
AJH: I want to add that it's very important to have a collaborative organization and a cooperative way of working together, where everyone, including developers, are involved in making the product. Co-ownership is crucial for developer satisfaction and, in fact, for overall employee satisfaction as well. For me, this is a key aspect of our work culture. I came from a consulting background. In a product company, your voice can be heard, which is not always the case in consulting firms. There, you often do as you are told because you’re being paid to deliver specific results. In contrast, in a product company, it is much easier to engage in discussions and collaboratively develop interesting and effective products.
A conversation with Rohan Deuskar, Founder and CEO of Stylitics
In this interview Rohan Deuskar, talks about the company's evolution from a digital closet platform to an inspirational outfitting and styling technology for retailers. He highlighted that over 60% of Stylitics' traffic comes from mobile, leading to a shift towards mobile-first user experiences. Rohan emphasized the importance of AI in enhancing personalization and efficiency, and stressed the need for retailers to focus on inspirational experiences to engage customers and improve key metrics like average order value and repeat traffic.
Bill Hogenauer (JM Search): First of all, thank you for doing this with us, Rohan, I appreciate you taking the time to discuss today's focus on the increasing role of mobile in e-commerce. Could you first tell us a little bit more about yourself and what Stylitics is all about to give us the context for our conversation?
Rohan Deuskar: I'll start with a brief bio about myself. I’m Rohan, the founder and CEO of Stylitics. I started the company about 13 years ago, originally as a digital closet platform and technology for consumers. About eight years ago, we transitioned to using inspirational outfitting, styling, and visual recommendations, along with enriched data, to help retailers enhance the experiences of shoppers. In the 13 years prior to founding Stylitics, I was in business school, spent some time at Amazon, and before that, I helped grow a mobile technology startup. We were one of the first companies to pioneer the use of SMS as a responsible marketing and communication channel with a company called Vibes Media. Additionally, most of our customers and retailers now see more than 60% of their traffic coming from mobile, if not higher, so this is definitely a topic that is near and dear to my heart.
BH: Great, and that’s a perfect segue into the heart of the discussion. There are several areas we’re really curious about, particularly the rise of mobile shopping and how it has transformed e-commerce - not just in terms of purchasing, but also in e-commerce operations, customer engagement, and the data that comes with it. What strategies has Stylitics implemented to adapt to this rise in mobile usage compared to using a standard web browser?
RD: It's been a massive shift. As a company, there are some simple things we've done, as well as more sophisticated changes we’ve implemented. To give some context, we power outfitting and styling technology and recommendations for about 200 major global retailers in around 50 countries. Last I checked, over 60% of that traffic was mobile, which marks a significant shift that occurred about three or four years ago, and it’s not going to reverse anytime soon. When we noticed this change occurring at such scale - across 30 billion sessions we assist with every year and 100 million orders - we had to make a clear pivot to becoming a mobile-first company from a user experience perspective. This means that, given the libraries of user experiences we provide that sit on the retailers' sites, we are an integral part of their mobile web and mobile app experiences. This shift is not unique to us; many good companies have been doing this for several years. Initially, some thought this was a constraint, but I believe it can actually offer a significant advantage due to the simplicity it fosters. It's important to recognize that user behaviors differ across platforms.
BH: Can you give some examples of the way you’ve taken advantage of this shift?
RD: More recently, we’ve implemented two major initiatives. The first is our Composer technology, which allows us to deploy very nuanced user experiences, user features, CSS changes, and different form factors across mobile, desktop, and tablets. In running numerous user experience tests live with retailers, we found, for example, that on mobile, a certain version of our hotspots experience - where users tap into an outfit to see the products light up - produced a mere 1% increase in revenue per visitor, translating to millions of dollars annually, compared to desktop. This illustrates that a one-size-fits-all approach doesn’t work, so being able to conduct different tests for each form factor and channel has been a considerable asset. Additionally, we’ve invested heavily in our Mobile SDK, as that’s where mobile truly shines. Historically, our approach was to simply port features over, but more recently—over the last year—we've aimed to build mobile app-first experiences. The advantages of an app can be quite different from mobile web. Users tend to be more loyal and engaged when they have the app, resulting in a different context and user experience. Therefore, our strategy has been to lean into the dynamics of shopping as they happen in that moment.
BH: That's fascinating! In a way, you've answered part of the follow-up question I have regarding the role of the mobile device in shaping behaviors. You mentioned that app users might be more loyal and discussed the importance of the SDK piece when working with your customers. Considering that mobile experiences already exist, do you have insights on the preferences of today’s e-commerce shopper, particularly regarding their general preferences regardless of whether they are using a desktop or mobile device?
RD: Definitely. The first observation is that most people are not just one thing or the other. While individuals have preferences, we focus on understanding the intent of the shopper in that moment. This involves considering the form factor they’re using—whether it’s mobile, desktop, app, or in-store. By intent, I mean determining whether they are looking to replenish an item like a white shirt, trying to understand a trend, attending an event, or making a high-stakes purchase like a prom dress, and so on. All of these factors, combined with the form factor, reveal significant behavioral differences. We also need to incorporate personalization and relevance, treating shoppers differently on mobile and desktop. Generally, people move through these contexts over the course of their multi-touch journeys. That said, there are specific moments that call for distinct experiences based on context and intent. For instance, every retailer has large grid pages. On desktop, this layout is acceptable because users can quickly browse through many products, such as jeans or button-up shirts. However, on mobile, users typically scroll vertically, which changes the site experience significantly. Our job is to break up that grid and inspire users by highlighting categories and surfacing the most relevant items for them. We have a product called Dynamic Galleries that targets mid-funnel engagement. We observe completely different behaviors; for example, while this product is very effective on desktop, it’s less effective on mobile—especially during sales. During sales, outfit recommendations must showcase best-selling products differently instead of interrupting an already extended scroll. Contextual awareness is crucial, and while personalization and relevance are important, factors like context, form factor, and intent are equally significant. In general, we’re seeing trends that require bolder, clearer calls to action on mobile. For instance, something as simple as “Shop This Look” needs to be much more prominent than on desktop. This is because mobile users have developed a tendency to act quickly, almost like “thumb instinct”. If you want to grab their attention, your messaging needs to be compelling and distinct—something that stands out amidst the fast-paced content consumption typical on platforms like TikTok and Instagram. In contrast, desktop experiences are more akin to a sit-down meal, where users may take a bit more time and thought in their browsing.
BH: You certainly nailed it with regards to my shopping behavior. When I suddenly realize I need something, I can easily make that purchase on my mobile device because it's a targeted decision - I know exactly what I want. However, if I need to think more thoughtfully about a larger purchase or want to explore options, I typically sit down at my desktop browser. The larger screen provides better visual real estate for comparison. That said, there are many interesting factors at play with both mobile and desktop shopping, including technological advancements and the importance of latency—nobody wants to wait for a page to load. Additionally, there are significant global partnerships in e-commerce, which enhance access through catalogs and other means. Given these rapid technological changes, what has Stylitics done in terms of leadership, hiring, and bringing in the right talent to navigate this dynamic environment? You have a purpose and a unique solution, but how are you managing the surrounding factors to stay focused on your mission?
RD: That's a great question. The biggest shift we've tried to instill, with considerable success centers on our approach to technology and the mindset of technologists. Traditionally, there were significant, baked-in decisions regarding the tech stack and toolkit, often influenced by strong, almost religious adherence to specific languages and technologies. These were heavy, long-term decisions that shaped hiring practices, deployment options, and more. However, the switching costs have changed. For example, we recently transitioned from using some Amazon AWS tools to Google Cloud Platform tools for various reasons. This was a straightforward project without service interruptions. My point is that we want people focused on solving customer problems, eager to use any high-quality tools available - not frivolously, but with an understanding of their responsibilities. The passion should shift towards problem-solving rather than sticking rigidly to familiar methods, akin to having a hammer and seeing everything as a nail. Some individuals may seem enthusiastic about new technologies, but when asked to implement them, they often revert to what they know. This is acceptable for certain cases, but we are a company committed to solving retailer and shopper problems by providing inspiration and style advice at scale and with quality. We will use whatever the best, enterprise-grade, secure tools are to achieve our goals. This fundamental shift requires more than just the CEO's influence; the CEO often gravitates toward the next shiny object. We need leaders, like our CTO and other product heads, to reinforce this mindset, as well as engineering leaders who can support and uphold it. That’s how we achieve organizational change.
BH: That is a very difficult thing to manage, truly. Being in the talent business, I empathize with you; the type of person who is curious is not necessarily the type who will stray far from their comfort zone, so I appreciate that insight. Now, on to the topic that everyone is thinking about these days: AI. I'm sure that Stylitics is using AI in your tech operations in some way, whether internally or throughout the product and customer experience. Without going into specific technologies or toolsets that you are deploying, how has AI impacted your product and the overall experience?
RD: Our company firmly believes that AI is a transformational tool, and it might be even more than that. There are several ways we’re using it across the spectrum. For instance, there's the way people are using generative AI, particularly large language models, in everyday work. This is useful; it's a time-saver and provides other benefits, and I won’t delve too deeply into those aspects, as they are fairly understood. What’s much more impactful is that over the last year, we’ve built out our AI data pipeline. This is a significant development for our business and for our customers because we aspire to deliver an experience akin to a personal shopper - not a chatbot - but rather an assistant that flows through the entire journey, is personalized, and can respond quickly. We can generate much more content that can then be further personalized. However, there are limitations to machine learning models and human-in-the-loop systems. We’ve developed a system where we can take any part of our operations and replace it with an AI-driven or hybrid solution throughout the workflow - a major advancement. For example, we can choose to create 15 outfits for a product using our more traditional technology and automation, or we can opt to create 200 outfits and filter them differently, providing each shopper with relevant outfits generated quickly. Because it hasn’t taken us years to develop, we can offer these features for free or price them reasonably. Otherwise, with a dedicated engineering team of 30 people on that one problem, we would have to charge differently. It's exciting because our retailers will receive supercharged, innovative experiences that were previously unfeasible at the same cost or even included with their purchase.
BH: Are retailers promptly recognizing the benefits of this type of product?
RD: I’ll mention a significant caveat: many retailers are currently being inundated with offers from both small and large technology companies claiming they have AI or generative AI capabilities. However, most enterprise use cases require pairing them with other data sets and systems to truly reap the benefits. In our context, one of the things we aim to enable is that as users browse the site, we want them to see the right outfits based on their browsing behavior - without any personal information - complete with captions relevant to their journey. For example, if someone is price-sensitive, we want to highlight that a product is a great value for specific reasons, especially if they tend to prefer bright, bold colors and are very trend-focused based on their browsing behavior. This level of personalization requires two key elements: the ability to write those captions and maintain brand consistency, as well as a robust personalization engine. We observe the journey across 26 different dimensions, including price sensitivity and color preferences, which inform the language model needed to generate the right output. If you only have the language model, you can’t achieve the desired outcome. Many companies either arrive without the necessary supporting systems or underestimate the challenge of realizing their vision without the right data sets or systems in tandem with the language models.
BH: Valid point—data is king, right? Continuing to delve into the technology, particularly concerning the AI question, there are certain limitations associated with deploying AI. Some of these are related to computing, while others pertain to latency and cost issues. When considering the deployment of AI in a solution, have any of these aspects posed challenges for you? I mean, this is the nascent stage of AI, right? We're just at the beginning of a massive revolution with these tools, so I'm sure the economic factors and usage patterns will change. But have there been any challenges in actually deploying them due to some of these concerns?
RD: It's a great question. We engage in extensive data enrichment on product attributes for our retailers, and we're actively expanding this by adding trend, style, occasion, and theme attributes that will enhance SEO, SEM, and other areas. This is a significant initiative for us, but it entails pulling in millions of products and using our hybrid system alongside LLMs to further enrich them with various attributes. There is, of course, a cost associated with this. Initially, our team approached this like we always do, thinking, "Let's assume it's 25 cents per product." We did the math and saw that we would either have to price it at a certain rate, or it would become cost-prohibitive. I wanted to offer a baseline version of this for all our retailers for free as a value add, with the option for them to purchase premium features if they wished. However, the calculation made it seem prohibitive—a situation we faced about a year ago. Fortunately, during this time, the price for implementing these solutions has dropped by 75%. We proceeded with building the system, anticipating this decrease, and it indeed materialized. Now, the economics make sense for everyone involved: it works for the retailers, it works for us, and it has become a non-issue. We're certainly riding that wave beneficially. Something I've observed is that throughout history, if you wanted better quality, if you wanted something faster or in greater volume, you would have to pay more. What's both confusing and novel about AI, particularly generative AI models, is that you can actually achieve higher quality, faster results, and greater volume for less than you would expect. This has fundamentally changed the economics of these processes, and I believe that's one of the greatest strengths of this technology. We're only beginning to see how this will percolate through society.
BH: Taking into account everything we've discussed so far regarding the role of mobile, your underlying technology, and your approach, when you look at your customer base of online retailers, what are you hearing from them? What are you interpreting based on the use of your product that would be considered their most significant challenges? What kind of pressures are they facing that make your solution important to them? Additionally, how do you think those challenges or pressures will evolve and impact Stylitics’ evolution over time?
RD: Retailers are under tremendous pressure across many different fronts, even the very successful ones, like Boozt. They face margin pressures, and while this may be a slightly US-centric view, I believe it is more global - we have many global customers, and I hear similar concerns from them. It's also becoming more expensive to acquire customers, and there are increasing distractions in channels. Retailers are no longer just competing with one another; they're also competing with TikTok Shop and other platforms. When retailers are set up in a certain way—transactionally, meaning they focus on getting a customer to find a product, buy it, and leave—this entire funnel is becoming more expensive and more challenging. There are two key areas right now that I believe will be very impactful for them and will change the equation in a positive way. One is AI; it hasn't been widely adopted yet, but retailers will likely adopt it first through partners. This includes call center operations and other in-house processes. The biggest benefit comes when partners become smarter and more capable through AI without passing on the costs to retailers, which is a significant advantage. The other key area - around which we've built our company - is shifting from purely transactional to also inspirational. When you spend a considerable amount to bring customers to your site, you need them to convert better and increase the average order value. Shoppers, especially on mobile devices, tend to have more transactional behaviors as they flip through items, but if you can catch their attention and provide more value, they may think differently. For instance, instead of simply looking for a scoop, they might explore more and discover items they didn’t realize you offered, like specific bags. This shift in mindset is crucial because it still leverages your own products and brand while showcasing different offerings. By doing this systematically, you can build positive associations with your brand. This can lead to increased organic traffic, repeat visits, and more time spent on the site. I believe there are many challenges, but there is also good news. Retailers should absolutely examine these two areas. They should seek out trusted partners and established players. We are very fortunate to now be working with Boozt, and we're live on their site after a couple of weeks. This is the right approach because the danger in this moment is that there is so much noise in the market - digital and marketing leaders can spend a lot of time and effort trying to navigate through various offerings, only to find that after three years, they’ve cut most of them because they didn’t scale to enterprise needs or weren’t the right fit. So, I would advise retailers to trust the established partners in the space, hold them to high standards in using these technologies, and ensure they’re passing on the benefits effectively.
BH: That's a great point. We have this discussion a lot in our business about whether we are acting transactionally or building relationships with our customers. As you mentioned, these e-commerce retailers spend a significant amount of money getting customers to their site, so you don’t want them to just pop in, make a quick purchase, and leave; you want to build a connection with them. Although I’m not heavily involved in social media, I do hear about influencers and how they’re persuading people to buy. In a way, a lot of what your technology does is invisibly create a better bond between the e-commerce retailer and the shopper, almost akin to relationship development. It’s great!
Thank you so much for your time; this has been an enjoyable conversation. I’ve certainly learned a lot more about you and Stylitics, and thanks again for all the time and investment you’ve put into this.
About the Authors
Bo Ekelund is a Partner at Amrop Sweden and a member of Amrop's Global Digital Practice, who leads high-profile searches for transformative CEOs, board seats and a range of C-suite positions. Through extensive experience in transformational leadership, Bo helps client companies build sustainable results in a digitalized world.As an executive in Ericsson, Sony Ericsson, GN Store Nord he has led multinational line organizations across timezones, cross-border post-merger integration, and led successful corporate-wide transformation programs with groundbreaking bottom line impact.
Bill Hogenauer is a Partner at JM Search (Amrop’s strategic partner in the US) and a member of the Firm’s Technology Practice. Bill leads the company’s center of excellence for technology and product leaders (CTO & CPO). He specializes in executive search engagements for high growth early-stage, scale-up, and middle-market technology companies, as well as private equity sponsors and their portfolio businesses. Bill joined JM Search in 2013 after four years of leading corporate and business development as a Vice President for a global manufacturing technology business.
The authors wish to thank both interviewees and their respective organizations for their collaboration on this report.