Boston Consulting Group research finds that the average fashion business operates on lead times of 37 to 45 weeks. This equates to estimating the styles, colours and sizes that shoppers will buy up to 10 months in advance.
Indeed, an accelerated, volatile trend cycle and logistical complexities continues to make the task of predicting consumer preferences a key challenge for brands and retailers. In 2023 in the United States alone, retailers were sitting on $740 billion in unsold goods, according to McKinsey research.
Further still, research from retail analyst IHL Group has found that retailers globally lost $1.77 trillion dollars in revenue in 2023 due to retail inventory distortion — taking into account both out-of-stock and overstock scenarios. How brands handle missing or excess inventory can negatively impact both brand affinity and the bottom line.
Since its inception in 2018, retail analytics platform Style Arcade has worked to optimise assortment planning for brands and retailers and power smarter product decisions. Finding solutions to fashion’s overstock issue — a $163 billion annual problem, according to Bloomberg Intelligence — is central to its mission. Style Arcade offers a complete suite of merchandising solutions including an AI-powered buy and size calculator to enable more intelligent forecasting.
These tools are designed to be leveraged across all functions within fashion businesses for a holistic understanding of product performance. Working with over 120 brands and retailers globally, it counts Amiri, Aje, Christopher Esber, Princess Polly, and White Fox among its clients. “Our annual benchmark study calculates that brands and retailers see on average an additional 4.5 points in gross profit within the first 12 months on the platform simply from better buying and sizing,” says CEO and co-founder, Michaela Wessels.
Now, BoF sits down with Wessels, a former VP of merchandising, to understand the opportunities that lie in leveraging product data, how to get ahead of the evolving demands of consumers and which AI-powered solutions are set to be a valuable differentiator.
What external factors have exacerbated the fashion industry’s inventory glut?
There are two key things that are critical to buyers’ success — they need to buy great products and then the right sizes in each product. If you are left with products that are not in high demand, or inaccurate sizes, then the best marketing in the world won’t save you from getting those two things wrong.
Historically, buyers would spend hours in either your generic business intelligence tools or your spreadsheets trying to quantify this. What was missing was a means of understanding what you could have sold without the inaccuracies of out-of-stocks skewing the data and, therefore, what your true demand is by store, by size and by product
Why have brands historically struggled to succeed in effective assortment planning?
Its a very multifaceted and complex problem. We have been relying on the human brain to compute the exact products to deliver to every single store in the right size, colour and silhouette, across every category as well as taking into account lead times, price points and the ever-changing trends.
On top of those asks, there is a level of granularity that is becoming increasingly hard to forecast. While weather variances and irregularities have thrown out traditional season patterns, best-selling products will differ hugely across store locations and on your online channels.
How can AI be applied to help match supply and demand?
AI needs to be user-friendly and fundamentally useful — both in terms of your teams’ day-to-day, and also the benefits for end consumers. In terms of matching supply and demand, Style Arcade AI helps by providing a recommended order quantity, by size, for specific items. Let’s say the brief is a white, short-sleeved mini-dress, delivered in April 2025 to a brand’s top five stores and priced at $199. AI can accurately determine the requirements, based on all facets of demand.
At Style Arcade, we believe that you can never take away the human influence and intelligence involved in trend forecasting. That buying intuition and deep knowing of your customer will remain key, but AI does help them get to the quantification faster and takes the grind out of the analysis.
What opportunities can smarter product data unlock over the medium-term?
I think many teams inherently know that sizing inaccuracy is a leaking bucket, but they haven’t measured and benchmarked their accuracy. Brands and retailers are losing, on average, up to 23 percent of profit on a monthly basis, due to inaccurate buying by size. Correcting this alone will close a significant profit gap for the industry.
Buying intuition and deep knowing of your customer will remain key, but AI will help get to the quantification faster and takes the grind out of the analysis.
But then if you think about the quantum of data that brands and retailers can get access to more broadly, it’s going to be able to solve quite sophisticated issues. They will be confident in predicting quite specific product information — they will know that in April 2025, the white, short-sleeved mini-dress, priced at $199, will be delivered — and it will be the right moment for their customer.
It’s a hard ask for buyers on their own, but with the data we can quantify this much more accurately. With smarter data, buyers can leverage all of their buying history. We see brands going quickly back to their archives, spotting bestsellers from three or four seasons ago that speak to today’s trends. They are then able to revive it quickly. Trends are cyclical — and that rich data is going to help brands to leverage that.
How is customer expectation set to evolve in this space?
Shoppers now absolutely expect the products that they want to be available, in their size, at their favourite retailers. I think consumers are still baffled as to why this problem of not being able to find their size still exists. The number one shopper complaint is still: “My size is sold-out” — which indicates how critical this is, not only for customer satisfaction and acquisition, but also to reduce the profit loss and wastage from inaccuracy.
The second key evolution, which will only become more of an expectation from consumers, is the time to market. They expect faster turnarounds on high demand products and in an almost counterintuitive way, they also expect that this speed to market will be provided in the least environmentally harmful way for our planet. To get ahead, the industry needs to get faster in our demand analysis and reaction times as well as far more accurate in our decision making.
Brands and retailers are losing, on average, up to 23 percent of profit on a monthly basis, due to inaccurate buying by size.
What trends do you foresee shaping the future of fashion retail?
The speed at which consumer expectation is evolving is going to drive our sector forwards. And they continue to raise their expectations — on hyper-personalisation and service, brand ethos and sustainability checkboxes. To meet those requirements, AI is going to become a trusted helper. The days of running retail businesses without AI is behind us, but it will take on more granular tasks which frees up teams to operate smarter and quickly.
The customer is saying, “Show me that you know me,” and that stands across all platforms — and across style, sizing and fit. All of this needs to come together, then we’ll have happier customers at the end of it.
