Nosto’s Jan Soerensen speaks to MartechView
Explore how Nosto leverages AI to fuel personalized shopping, tackle tech challenges, and empower brands to stay human-centered in a rapidly evolving market.
In this exclusive interview, Jan Soerensen, General Manager of North America at Nosto, delves into the evolving e-commerce landscape. We explore the rise of product-centric brands, the impact of celebrity endorsements, and the growing demand for authentic and sustainable shopping experiences. Jan also shares insights on the power of AI-driven personalization, the future of e-commerce technology, and how Nosto is helping brands navigate this dynamic and competitive market.
Excerpts from the interview;
As Nosto’s North America GM, what’s the biggest e-commerce shift you’ve seen recently, and how is Nosto adapting?
Product innovation (vs. brand innovation) has come to the fore
In the last few years, I’ve noticed that many of the most successful and trending brands have emerged through a heavy focus on true product innovation (vs. brand innovation).
Companies like On Running, Hoka, HexClad, and Kyte Baby (to name a few examples) have each introduced unique product innovations that have distinguished them in competitive markets:
- On Running revolutionized the shoe industry with its patented CloudTec sole, offering a distinctive cushioning technology that provides superior comfort and responsiveness.
- Hoka One One, founded in 2009 by former Salomon employees Nicolas Mermoud and Jean-Luc Diard, became famous for its oversized outsoles, which provided extra cushioning.
- Kyte Baby has also led with product innovation, using organic bamboo cotton to create soft, sustainable clothing and baby products.
- In a different category, HexClad transformed cookware with its unique hybrid design—a combination of stainless steel and nonstick coating that delivers durability and performance. Their patented design sets them apart from traditional cookware brands.
Regarding Nosto, our personalization engine is increasingly used to highlight brand and manufacturing characteristics, for example, for new visitors (vs reverting to a more transactional website for returning visitors).
Celebrities dominate modern product launches
Celebrity influence has become increasingly dominant across various industries in recent years, from alcohol and beauty to fashion and luxury goods. You’d be hard-pushed to find a major product launch that doesn’t rely, at least in part, on the support of celebrity influencers.
While the world of alcohol brands has seen celebrities like George Clooney (Casamigos), Ryan Reynolds (Aviation Gin), and Dwayne “The Rock” Johnson (Teremana Tequila) successfully launch their labels, the beauty industry’s welcomed players like Rihanna (Fenty Beauty), Kylie Jenner (Kylie Cosmetics), and Selena Gomez (Rare Beauty) in building entire empires. The presence of celebrities has become integral to brand success. These public figures attract their fan bases and lend credibility and excitement to a wider audience, creating a sense of aspiration and exclusivity.
Even in fashion, traditional brands have embraced celebrity collaborations to boost visibility and appeal. From Kanye West’s partnership with Adidas to collaborations between musicians, actors, and high-end designers, celebrities have become key tastemakers, shaping trends and influencing consumer purchasing decisions. What was once a niche strategy has become a core part of the marketing playbook.
Within Nosto’s portfolio alone, many brands are celebrity-owned or fuelled by celebrity ambassadors (e.g., Victoria Beckham, Kylie Cosmetics). And it’s not uncommon for such brands’ product ‘drops’ to attract shoppers in the hundreds of thousands. This brings technical challenges around serving, loading, and scaling millions of transactions within seconds. So, these merchants must work with platforms that can handle this.
Younger consumers are hungry for authenticity and more conscious consumption
The demand for authenticity is growing as the burden of proof increases, driven by the overwhelming amount of fake content online. Consumers, particularly Gen Z, prioritize transparency, security, and services that protect their data.
This same generation is also looking to consume more consciously—a countertrend to the rise of fast fashion. For instance, while brands like Shein have contributed to a toxic cycle of overconsumption that’s had both social and environmental impacts, as a New Yorker, I can tell you that the number of vintage shops has massively increased in recent years.
Similarly, companies like DVF and Ulla Johnson have addressed sustainability through initiatives such as pre-loved programs, where customers can return or buy gently worn, vintage items.
Nosto is fit to help merchants serve this segment, for example, by using its personalization engine to highlight vintage apparel to shoppers it’s detected to have an affinity for it.
Also Read: How AI Augments, Not Replaces, Human Marketing Skills, Says Amanda Cole
How is Nosto evolving personalized commerce to shape online retail’s future?
Automation is becoming more and more powerful, and as such, merchants are increasingly utilizing this to reduce time allocation for their teams while optimizing their catalog performance.
For example, we have an innovative analytics tool designed to help merchants overcome the challenges of accessing quick, actionable product insights, particularly when limited by resources like time, analysts, or varying levels of data literacy.
The tool provides simple yet powerful answers to complex merchandising questions, such as optimizing inventory, improving sell-through rates, increasing catalog conversions, and reducing return rates—often addressing multiple goals at once. In the future, it can directly apply the suggested strategies to the merchant’s store.
Additionally, when we look at ecommerce search, we see traditional keyword-based search engines being enhanced—or even replaced—by ‘vector search.’ Rather than relying on synonyms and various NLP techniques to match keywords, vectorization encodes multi-modal objects (such as images, long texts, and videos) into vectors. This approach excels in handling complex, long-tail keywords and the rapidly evolving nature of language, which is especially useful in today’s dynamic digital landscape.
For example, a keyword search for “notebook bag” might not return relevant results for a laptop bag. However, with vector search, the system can interpret the product description (e.g., “has a large main compartment with a laptop sleeve”) as relevant to the query for “notebook bag,” even though the terms don’t match exactly. This makes vector search more intuitive and effective in delivering accurate results.
With the rapid advancements in AI and machine learning, how do you foresee these technologies revolutionizing the e-commerce landscape over the next 3-5 years?
The rise of AI agents will profoundly impact ecommerce in the coming years. Specialized AI agents can collaboratively manage the entire strategy and execution of all e-commerce-related business functions.
Think about a tool that analyzes a store’s conversion rate and market trends, then optimizes onsite merchandising while suggesting buying specific items for the next season, all while creating a YouTube tutorial to generate more organic traffic! Currently, Nosto is experimenting internally with AI agents in business intelligence and exploring how to make them available inside other tools.
Similarly, co-pilots and proactive analytics will also be significant developments. Essentially, these deliver data-driven insights to a user, suggesting specific actions to take, saving a ‘data person’ from manually sifting through weighty reports in the hope of finding a single insight. Nosto’s Experience.AI or Klaviyo’s Guidance Layer are good examples of these.
Lastly, just like on the brand side, there’ll be significantly less headcount needed to actively manage accounts on the SaaS vendor side, with more automation within Customer Success/Account Management. This is likely to pave the way for more affordable ecommerce solutions.
Also Read: How Swiss Re is Navigating the Complex World of Reinsurance
What common mistakes do brands make in crafting seamless, personalized customer experiences in today’s competitive market?
The most common pitfalls we see among merchants are the following:
- Not relying on a unified personalization and merchandising approach
Despite the general shift towards a more consolidated tech stack, we still have brands coming to Nosto that are leveraging 4+ tools for merchandising alone (a search tool, a product recommendation tool, a post-purchase upsell tool, and a bundle tool). Nosto cuts this down to one, helping brands do more from a single platform.
- Not relying on verticalization of vendors
Nosto is making big bets on fashion, beauty, and lifestyle, and other vendors are under pressure to become more specialized tools. Upon vendor selection, merchants should seek to understand how a vendor’s feature set aligns with their vertical. For example, if you’re a beauty brand, can the personalization engine you’re looking at detect color affinity? If you’re an alcohol brand, can they facilitate replenishment recommendations?
- Not relying on A/B testing
Shopify brands are maturing, focusing more on structured CRO programs. Brands that don’t adopt these risk getting left behind. Shopify-centric tools like Nosto or Shoplift will help against agnostic platforms like VWO and Optimizely.
How does Nosto use data to create dynamic, personalized shopping experiences?
Nosto’s AI comprises four key competencies: predictive AI, Semantic AI, Visual AI, and Generative AI. Together, they power every feature of our platform so merchants can achieve more faster. Our AI engine is a merchant’s best ally, automating repetitive tasks and streamlining workflows while delivering intelligent commerce experiences that drive growth.
Predictive AI analyzes, tests, and automates operations from large data, while semantic AI helps understand and contextualize for human-like, conversational search responsiveness. Meanwhile, Visual AI finds, recognizes, and sorts images to streamline UGC and content campaigns, and Generative AI augments, enhances, and streamlines for improved efficiencies across the entire platform.
What are the key challenges in creating unified e-commerce experiences with third-party integrations, and how does Nosto tackle them?
The end of the ‘connector’ era integration approach: Even as little as a few years ago, SaaS solutions could integrate with ecommerce platforms and other vendors via a lightweight connector app. The app would usually facilitate a few steps of the integration process and, just as importantly, serve as a marketing vehicle for message compatibility. Expectations have completely changed with Shopify, particularly with the new generation of brands. A ‘compatible’ app is now expected to support all new feature releases harmoniously, work together across all touchpoints, and truly act as one integrated platform.
The increasing rate of innovation: All software has increased its ship speed. With that, API updates, deprecations, and framework changes are much more frequent.
AI agents: The future of integrations is having agents of one tool sit in another. Think about a Business Intelligence tool that tracks landing page performance and sees a drop in the KPIs. It could trigger a survey to get qualitative customer feedback, then perhaps action more personalization strategies through Nosto or another agent.
Also Read: Future of Growth Marketing: Insights from Franciska Dethlefsen
How does Nosto help North American brands leverage AI while keeping a human touch in customer interactions?
Firstly, by offering transparency around the AI. While Nosto heavily relies on multiple AI models, including LLM, transparency is always built into the tool. For example, suppose our AI was to find a valuable shopper segment for a merchant to target. In that case, there’d be a natural language summary of the characteristics of that segment for the merchant to view and finetune if they see fit.
Secondly, by offering a strong suite of merchandising tools. Black Box AI has been a turn-off for brands wanting to control the final shopping experience, and I bet it will stay this way. While every brand needs automation in its day-to-day operations, the secret sauce of merchandising and branding will largely stay in the hands of people involved in the brands’ intricate storytelling.
Lastly, we use LLM to generate synonyms for our Search engine, letting humans give the final thumbs-up. Instead of purely running this without input from a human, the merchandiser can give their thumbs up or down for these synonym pairs. A human has the final say.