Released in The Interline’s AI Report 2025, this executive interview with Alvanon is one of an eight-part series that sees The Interline quiz executives from companies who have either introduced new AI solutions or added meaningful new AI capabilities into their existing platforms.
For more on artificial intelligence in fashion, download the full AI Report 2025 completely free of charge and ungated.
Key Takeaways:
- AI is rapidly progressing from broad promises to delivering quantifiable ROI by transforming data into operational improvements, particularly in areas like software development, where it significantly boosts efficiency and agility. This shift targets where fashion currently experiences the most value leakage.
- Stubbornly high e-commerce return rates, sometimes reaching 50%, are primarily due to inconsistent sizing standards and consumer uncertainty, leading to multi-size purchases. AI tools like virtual fitting rooms and sizing recommendations will only effectively reduce returns if companies first resolve these fundamental sizing inconsistencies and integrate data from across teams, as demonstrated by platforms like MyAlva.
- AI is predominantly being deployed as a backend enabler for specific tasks, empowering human decision-makers rather than serving as a primary frontend interface. Its future lies in supporting areas like design ideation, product development automation, and AI-generated content to reduce costs, increase efficiency, and accelerate time to market across the entire supply chain.
Where do you believe we currently are on the progression curve from AI as an extremely broad set of capabilities and promises, to AI as the foundation for applications and services that can deliver a measurable return on investment in well-defined areas?
I think we are well on our way. AI already surpasses humans in various tasks that are based on reiteration and reinforcement. Software development, for example, is a space where leveraging AI can produce a greater ROI by increasing efficiency and agility. Beyond generative imaging, AI is now being used to turn data into operational improvements, and that’s where real business value starts to emerge.
To be more specific about where that kind of quantifiable return might be found, it seems logical to look at where fashion is currently letting the most value escape, and to ask what AI can do about it. With returns from eCommerce now as high as 50% in some market segments, that seems like a strong target – but it’s also one that technology providers have tried to hit from multiple different angles in the past. From a cultural and a technological perspective, why have return rates remained so stubbornly high? And what do you believe AI tools, based on a huge amount of body data, can do to reduce them?
The main issue is that most companies still don’t use consistent sizing standards. If your products aren’t consistent, consumers can’t trust what size will fit them, and that’s the biggest reason for returns related to sizing and fit.
Returns unfortunately cannot be solved by simply applying some form of AI. If that was the case, then we would see much more success from the current solutions in the market, like sizing and fit recommendations and virtual fitting rooms.
The other reason why returns have remained high is simply due to the inherent lack of confidence that the consumer has in terms of the product sizing and fit and the ease of free return policies. Customers with low confidence in a brand’s sizing are buying multiple sizes in a single garment, and then returning the ones that don’t fit. If companies solved their sizing inconsistency issues then, overtime, these types of hedged purchases will go away and consumers will exercise brand loyalty and increased conversion rates.
If companies adopt and execute sizing standards properly, then AI tools like virtual fitting rooms, and sizing and fit recommendation engines will work much more effectively than before and the industry will see much lower return rates. If those AI-powered tools are based on real human measurements, then even better.
Another area that a significant amount of value leaks out of the typical product lifecycle is waste. Whether it’s iterative sampling and all the rework that creates, low materials utilisation, or a lack of standardisation in sizing, a lot happens upstream to determine how much waste is created in a single style’s route to market. These are not the most outwardly-compelling use cases for AI when compared to generative product photography, for example, but sourcing teams and their strategic partners are perhaps some of the best examples of end users who could benefit from AI. But what would that be like in practice to make a meaningful difference?
Absolutely. While generative AI gets more attention, traditional machine learning is already making a big impact across the supply chain to great results.
Let’s look at 3D product development, for example. It is proven to drastically reduce the number of physical samples, which cuts both costs and environmental waste created during the product development cycle. AI is also being integrated into these digital workflows through developments such as automated pattern generation, material costing, and usable digital twins that can create more efficiency and increase ROIs. These AI developments are underway at this moment and will have sizable contributions for increasing speed to market and reducing waste.
The most effective way to avoid creating waste or incentivising returns is, of course, to bring the right products to market in the first place through intelligent merchandising, and then to pair them to the right consumer through smart inventory planning and product recommendations. This kind of forecasting and trend analysis has historically been primarily backwards-looking, based on retrospective market signals and prior performance, combined with intuition – and product recommendation engines have been notoriously rigid and deterministic. How do you see this changing thanks to platforms like MyAlva and the kind of real-time analytics you aim to provide on shopper preference, body data trends, and size-related frictions in the shopping experience?
MyAlva is an AI tool that allows us to capture the consumers body shape and size based on several parameters – height, weight, age, through consumer input or scanning. While this tool is similar to other existing sizing finders, the difference is in the utilization of gathered data.
Traditionally, e-commerce, merchandising, and product teams don’t share data effectively. Useful information that is gathered by ecommerce is not being transferred in meaningful ways across all teams.
That’s where MyAlva changes the game, providing transparency of the gathered consumer body data to all parties.
Product teams can look at measurement data and body shape data to improve their product sizing and fit. Merchandizing teams can look at sizing distribution charts to better determine sizing allocation in different regions. They can also easily determine if there are opportunities for new size ranges and product categories based on what consumers are looking for. And ecommerce teams get stronger insights into consumer preferences and reduce size-related friction in the shopping experience.
MyAlva and its AI-powered insights provide another important data source for decision makers, when combined with existing data sources, like historical sales and returns data, competitor analysis, or trend forecasting, adds up to better product offerings and more accuracy in terms of sizing.
We’ve talked about AI for very discrete use cases, but there’s also a growing argument being made for layering it across as much of the value chain as possible, as a way to bridge different disciplines and to bring structure and smart decision-making to the full spectrum of product design, development, and retail. How are you approaching that angle? And do you see that kind of enterprise-level intelligence as being off-limits to smaller companies, or is it theoretically accessible to everyone?
I touched upon this on the previous question, but to elaborate, AI is an enabler that should be used throughout the supply chain, bridging creative, technical, and manufacturing. It enables consistent decision-making and faster product development.
The important thing is that it is made useful for the different roles, so all teams can fully utilize the insights tools like MyAlva provide. Like the example above, a technical designer is going to want to dig into detailed body measurements to build patterns, while a buyer will want the AI simply to guide them on what sizes to buy while planning inventory. An intelligent system will be able to cater to both these demands, and this is the main differentiation point for MyAlva. We realize that AI doesn’t make decisions in the company; it is the people. The AI is simply there to empower them. MyAlva is not an enterprise level system, it is a simple, powerful tool that can be used by startups or corporations.
What do you believe are the next steps for how AI in general is deployed and used? Is it more likely that AI will solidify its place as a new human interface paradigm the frontend of tools and workflows? Or is its future closer to what cloud infrastructure has become today – a quieter commodity that is still the foundation for the next generation of applications, but in a less obvious way than what we’ve seen over the last couple of years? Or is it both?
I think that AI, in the near future, will be used primarily as backend enablers for specific tasks. For example, we are not going to see a Hollywood film created entirely by AI, but the film industry is certainly using generative AI to help them create frames and scenes for their movies during production.
While on the frontend in apparel we are going to see more virtual try-on AIs, I think that a lot of the AI integration will be happening in the backend. Design Ideation, product development automation, AI product image and video generation, and other tools will all be used to reduce costs, increase efficiency and speed to market.
