Demand sensing in fashion: real-time signals and forecast adjustments

What is demand sensing and how does it differ from traditional forecasting?

Demand sensing leverages real-time data and improved analytics to understand and anticipate customers, unlike traditional forecasting which relies on weekly or monthly cycles. Blue Yonder uses short-horizon machine learning to incorporate near-real-time signals such as point of sale data, weather, promotional activity and market indicators into daily or weekly replenishment cycles. These systems continuously detect weak demand shifts in real time.

Real-time sales data, social media monitoring and predictive analytics are vital tools for capturing and analysing current customer behaviour and identifying emerging trends promptly. The goal is to respond to actual market signals before a trend peaks or a product sells out, rather than relying on a static projection fixed weeks or months earlier.

Which data inputs do demand sensing systems use?

Demand sensing platforms pull together internal operational data and external market signals. Blue Yonder Demand Edge ingests external data such as weather, local events including concerts and sports, holidays and social trends, and correlates them to sales history. The system can ingest point of sale data throughout the day, and if a store sells 50 per cent of its inventory by noon, the system flags a risk of stockout immediately.

Beyond the shop floor, fashion brands also track upstream indicators. Real-time data mining supports efforts by highlighting evolving customer preferences about how, when and where they want to engage with a brand. AI forecasting systems update predictions continuously as new data comes in, adjusting for real-time sales velocity, competitor promotions, trend signals and regional inventory levels.

The table below summarises common data sources:

How often do merchandisers adjust forecasts when using demand sensing?

Traditional seasonal planning operates on monthly or quarterly cycles. In pre-season planning, decision speed follows monthly or quarterly cycles, whereas in-season planning operates on a daily to weekly basis. Modern platforms provide real-time reporting for in-season course correction.

Intraday sensing does not wait for the end of week report; it can ingest point of sale data throughout the day, and if a store sells 50 per cent of its inventory by noon, the system flags a risk of stockout immediately. For fast fashion and trend-driven brands, this continuous updating is the capability that matters most.

AI forecasting systems that incorporate sell-through velocity and remaining season length can flag slow-moving SKUs early enough to take action, including a targeted promotion, a price adjustment or a channel redirect, before the item becomes dead stock.

What technology platforms support demand sensing in fashion retail?

Based on a survey conducted for the 2025 State of Fashion report, 75 per cent of fashion executives are prioritising AI for demand forecasting, inventory optimisation and cost control. Several enterprise platforms now offer demand sensing capabilities tailored to fashion and apparel.

Blue Yonder Demand Edge for Retail uses advanced machine learning to predict consumer demand at the SKU and store level by analysing hundreds of local causal factors such as weather, events and price elasticity in real time. In August 2026, Blue Yonder announced new retail innovations as part of its Cognitive Solutions portfolio, designed to help hardlines and softlines retailers move from insight to action with agentic AI and connected retail decisioning.

In Q1 2025, organisations in automotive, consumer goods, fashion and apparel, and specialty retail selected o9 to improve planning, forecasting and decision-making. On 25 March 2025, o9 announced that Mango, one of the leading international brands in the fashion industry, has chosen o9 to enhance its planning capabilities and foster more data-driven decision-making across its operations. In Q2 2025, multiple companies across a wide range of industries, including food and beverage, electronics, fashion and medical equipment, went live with o9 for assortment planning, supply and demand planning, inventory optimisation and order management.

In 2025, more than 72 per cent of retailers use cloud-based POS systems, allowing for real-time updates, remote access and seamless integrations with e-commerce platforms. This shift to cloud infrastructure enables the rapid data exchange that demand sensing requires.

Which fashion companies have deployed demand sensing tools?

Fashion brands have begun modernising their planning architecture to support faster forecast cycles. With approximately 2,850 stores in 120 markets, Mango is digitally transforming its Merchandise Financial Planning, Assortment Planning and Demand Planning capabilities, replacing fragmented, outdated systems with an integrated, AI-powered and data-driven planning framework.

JD Sports Fashion stated that as the company continues growing its business, it looks forward to partnering with o9 as they understand the business model and unique requirements to enhance assortment planning capabilities. In Q2 2025, o9 partnered with JD Sports Fashion to optimise assortment planning for scalable growth and went live at Prada Group with advanced production and procurement planning capabilities.

Blue Yonder Demand Planning earns 4.2 out of 5 in a 2026 evaluation, standing as the most widely deployed demand planning solution in North American retail, with Walmart, Tesco and Albertsons among its flagship customers. Over the 2025 Thanksgiving weekend, Blue Yonder surfaced inventory availability to its customers in as quickly as 10 to 12 milliseconds and estimated delivery dates for over 1.2 billion SKUs, driving improved conversion for retailers.

What are the operational challenges and implementation requirements?

Demand sensing places new requirements on data quality and system integration. Investing in a unified order management system and warehouse management system infrastructure before deploying AI forecasting tools matters, because the systems need to share a single source of inventory truth to generate predictions that are actionable.

With inventory accuracy falling to an industry average of 65 per cent, forecasting variables present an ongoing variety of challenges for estimating stock cover and perceived demand; according to the Business of Fashion McKinsey State of Fashion 2025 report, 33 per cent of brands continue to struggle with inventory positions. Merchandising teams are often ordering manually, using spreadsheets to plan product, store and size breakdowns, and that process is not only complex but reinforces existing inaccuracies in the data.

Organisational change also matters. Modern platforms use AI-driven demand sensing that reacts to market changes faster than static forecasting models, scenario planning that helps teams compare inventory, margin and sales trade-offs, and cross-functional workflows that connect merchandising, finance and supply chain decisions. Teams must be ready to act on daily signals rather than wait for monthly review cycles.