Understand the “Why” Behind the “Buy”

Fashion is a dynamic industry, and as a result, there are several complex variables. Deconstructing customer intent is the key to reversing the downward trend.

Retailers are spending hefty amounts on customer analytics but one thing i still missing—customer intent. Defining it as the crucial “why behind the buy,” Sarah McVittie, Co-founder at Dressipi, says the key to successful intent-driven personalisation lies in understanding how and when to deploy it for optimal customer experience.

Recently acquired by mapp, a marketing cloud provider, Dressipi transforms how retailers engage with their customers through deeper, entirely personalised experiences. Using technology like computer vision, natural language processing (NLP), and deep tagging, the company helps apparel retailers show each visitor the items they’re most likely to buy – and keep.

“Fashion retail presents unique challenges and if a retailer is using a generic recommender powered by a standard collaborative filtering model, these challenges aren’t going to be addressed, “ says McVittie, explaining how solutions engineered specifically to handle fashion-specific challenges are a must.

CXM Today spoke with Sarah about tackling retail challenges for the fashion industry and the use of technology in achieving fruitful analytics.

Excerpts from the interview:

Tell us about Dressipi and what challenges it solves for the industry.

Fashion retail faces a critical challenge today: despite significant technology investments, retailers struggle to understand customer intent – the crucial “why behind the buy.” With conversion rates stagnating at 2-3%, return rates exceeding pre-pandemic levels, and full-price sell-through stuck at 50-60%, the industry faces a profitability crisis, with operating margins down 20% since 2018. Dressipi’s AI platform tackles this head-on by deconstructing products into customer-centric attributes, enabling truly personalised experiences and smarter operational decisions that can reverse this downward trend.

How does Dressipi’s AI help optimise product assortment and inventory management?

Our AI platform leverages the fashion industry’s largest dataset to decode product performance at the attribute level. By combining this with sophisticated customer propensity modeling, we deliver accurate predictions and forecasts for all products – including new items. This empowers buying teams with demand-driven insights about which style combinations will resonate most with customers, leading to more strategic assortment decisions that align perfectly with customer preferences

How can the technology help reduce stockouts or overstock situations for clients?

We address this challenge through a dual approach. First, our personalisation algorithms create precise customer propensity models at both size and feature levels, enabling retailers to sell fragmented stock at full price without resorting to discounts. Second, our demand forecasting solution integrates seamlessly with existing processes, delivering accurate size and volume predictions across products, categories, and stores. This attribute-level understanding of products and customer intent helps retailers optimise inventory allocation across channels, resulting in an 8% improvement in full-price sell-through rates.

What are elements specific to fashion and apparel that make Dressipi’s offerings unique?

Fashion retail presents unique challenges and if a retailer is using a generic recommender powered by a standard collaborative filtering model, these challenges aren’t going to be addressed. Fashion is a dynamic industry and as a result there are several complex variables: diverse customer preferences that change frequently, a constant influx of new products (33% monthly) creating cold-start problems, sparse customer data (70% make just one annual purchase), high return rates, and size availability complexities. Our solution is specifically engineered to handle these fashion-specific challenges and deliver 8% incremental revenue growth.

What is the role of AI in demand forecasting?

Demand forecasting represents an ideal application of big data and machine learning, providing the right quality, breadth, and depth of data. Our approach builds prediction models based on customer propensities and garment features, then combines these with historical sales data, customer behaviour patterns, and store profiles. This creates more accurate forecasts by understanding demand at a feature level, meaning retailers are able to identify money that would have otherwise been left on the table with a simple historical data analysis.

How should marketers approach unstructured data challenges with AI?

Our AttributeAI product addresses this directly by automating data enrichment and standardising attribute tagging through physical, contextual, and dynamic data points – all in customer-centric language. This comprehensive approach to product attributes streamlines operations from design to SEO, reducing manual effort while improving accuracy. The result is faster time-to-market, better product discovery across channels, and a 10% improvement in ROAS.

What can you tell marketers about utilising intent-driven product recommendations to enhance the purchase experience?

The key to successful intent-driven personalisation lies in understanding how and when to deploy it for optimal customer experience. Our personalisation engine creates individual “best aisles,” “best outfits,” and “best edits” for every visitor while maintaining brand DNA in storytelling. Beyond product recommendations, we emphasise understanding the most relevant context for each customer. Our unified approach – using consistent AI models and data across marketing, merchandising, and buying teams – creates a truly customer-centric retail operation that delivers results across all channels.