Trend forecasting platforms compared: what do WGSN, Heuritech and others measure?
Trend platforms differ less in their AI than in the data they watch. What WGSN, Heuritech, Trendalytics and EDITED measure, and how to choose.
KEY TAKEAWAYS Summary by the editors
- Fashion trend platforms differ mainly in the data they observe: expert-curated forecasts, social media images, search behaviour or retailers' live product listings.
- WGSN combines more than 2,800 third-party data sources, social posts, retail SKU tracking and a team of more than 250 experts, and describes an 'analyst in the loop' approach for its TrendCurve forecasts.
- Heuritech analyses social media images with computer vision, recognising more than 2,000 fashion attributes, and says it forecasts up to 24 months ahead; its published accuracy figures are company claims.
- Trendalytics combines search, social and product data across more than 10 million products, while EDITED focuses on retailers' assortments, pricing and discounting.
- For design, 3D and sourcing teams, the best platform is the one whose data matches the decision being made, and whose output can be translated into the company's own attributes and materials.
Trend forecasting platforms such as WGSN, Heuritech, Trendalytics and EDITED all use data and machine learning, but they watch different things: expert analysis of runways and culture, images posted on social media, consumer search behaviour or retailers' live product listings. Which one fits depends on whether a team needs creative direction, an early read of rising attributes, or evidence of what is actually being stocked and discounted. This guide compares what each type measures, based on what the providers publish.
What do trend forecasting platforms actually measure?
Broadly, there are four kinds of input. Expert forecasting uses trained analysts who interpret catwalks, trade fairs, street style and wider cultural shifts. Image analytics applies computer vision to large volumes of social media photos to count how often attributes such as a neckline, a print or a fabric appear. Search and social analytics measures what consumers look for and engage with. Retail analytics tracks what retailers list online, at which prices and with what discounts. Most providers now combine several of these, but each has a centre of gravity.
How do WGSN, Heuritech, Trendalytics and EDITED compare?
| Platform | Centre of gravity | Published data scope | Typical output |
|---|---|---|---|
| WGSN | Expert forecasting supported by data | 2,800+ third-party sources; 100k+ social posts a month; 300m+ SKUs tracked daily; 5m+ catwalk, trade show and street style images; 250+ experts | Seasonal direction, colour and material forecasts, TrendCurve forecasts |
| Heuritech | Image analytics on social media | Millions of social images a day; 2,000+ fashion attributes; edgy, trendy and mainstream panels | Attribute-level trend curves, forecasts up to 24 months ahead |
| Trendalytics | Search, social and product data | Millions of search data points; 10m+ products and 20k brands | Trend momentum, product attribute and pricing views |
| EDITED | Retail analytics | Competitor assortments, pricing, discounting and site merchandising (volumes not published on the page reviewed) | Assortment and pricing benchmarks |
WGSN describes its method as multi-layered, combining proprietary and third-party data with internal frameworks and expert analysis. Its Fashion TrendCurve model, it says, applies deep machine-learning algorithms enhanced by analyst expertise, and it claims high accuracy for these forecasts without publishing a single figure on the page reviewed. Heuritech states on its website that it forecasts with 92 percent accuracy 24 months ahead; such claims depend on how accuracy is defined and should be read as the company's own assessment.
Which data source suits which decision?
- Creative direction for a new season: expert-led forecasts give themes, colours and moods that design teams can interpret.
- Choosing which attributes to scale: image analytics shows whether a silhouette or print is spreading from early adopters to the mainstream.
- Checking consumer intent: search data shows what people actively look for, which is closer to purchase than visibility.
- Pricing and assortment gaps: retail analytics shows how crowded a trend is, at which price points, and how fast competitors discount it.
How do trend platforms connect with 3D design and sourcing?
Trend outputs become operational only when they reach the tools where products are created and sourced. For 3D teams, the useful format is concrete: colour references in a standard colour system, material descriptions that match the digital fabric library, and silhouette descriptions that can be turned into a base block. A forecast that says 'softer tailoring' is inspiration; one that specifies dropped shoulders, wider trouser legs and fluid suiting fabrics can be modelled in 3D and reviewed before any physical sample is made.
For sourcing, the critical link is materials and timing. If a platform signals rising demand for a fabric or finish, sourcing teams need to know early enough to reserve capacity or develop alternatives. Retail analytics can add a reality check: if competitors already list the trend widely at low prices, a brand may decide to source a differentiated version rather than compete on price.
How should you evaluate a trend platform before buying?
- Start from your decisions. List the three or four decisions the data should improve, such as colour selection, option count per trend or price positioning.
- Ask how the panel is built. Which platforms, regions and audiences are covered, and how does that compare with your customers?
- Ask for the accuracy definition. What exactly was predicted, over what horizon and against which outcome?
- Test on your past seasons. Ask the provider to show what the platform would have said before seasons you already know the outcome of.
- Check the attribute taxonomy. Can the platform's attributes be mapped to your PLM and product data, so signals can be compared with your sales?
- Clarify integration and export. Can data be exported or accessed programmatically, or is it locked in dashboards and PDFs?
What are the limits of trend forecasting platforms?
All platforms share structural limits. Social image panels reflect people who post publicly, which may not match a brand's customers. Retail analytics shows what is listed, not what sells, unless the provider infers sell-through from stock changes. Expert forecasts are influential partly because many companies read them, which can make them self-fulfilling and push competitors towards similar ranges. In an NPR report from October 2025, both a Heuritech representative and WGSN's chief forecasting officer described AI as a tool that supports human forecasters rather than one that predicts fashion on its own.
For most brands, the sensible approach is to combine one creative source with one quantitative source, link both to internal sales data, and review after each season which inputs actually improved decisions. A platform that cannot be connected to the company's own attributes, materials and sales will remain a source of inspiration rather than a planning tool.
Frequently asked questions
What is the difference between WGSN and Heuritech?
WGSN combines expert analysts with a wide range of proprietary and third-party data, including catwalk images, social posts and retail SKU tracking. Heuritech focuses on computer vision analysis of social media images to measure how fashion attributes spread among different audience panels. Both publish accuracy claims for their own forecasts.
How accurate are AI trend forecasting tools?
Providers publish accuracy claims, such as Heuritech's stated 92 percent accuracy 24 months ahead, but definitions vary and figures are rarely independently audited. Buyers should ask exactly what was predicted and test the platform against seasons whose outcome they already know.
Which trend forecasting tool is best for a fashion brand?
It depends on the decision. Expert-led services suit creative direction, image analytics suits deciding which attributes to scale, search data shows consumer intent, and retail analytics shows competitor assortments and prices. Many brands combine two sources.
Can trend forecasting data be used in 3D design?
Yes, if the output is specific enough. Colour references, material descriptions and silhouette details can be translated into digital fabrics and base blocks in 3D software, so trend-led concepts can be reviewed before physical samples are made.
One edition every weekday morning. Read in five minutes. Free for industry professionals.
