How we calculate some trend metrics is changing soon

This update will improve how we analyze key attributes within each product category.

This will make it easier to understand which colors, fabrics, prints, and product details are gaining or losing popularity within a specific product type, helping you identify opportunities and make better informed product decisions.

A few key concepts

To make the change easier to understand, let's start with some definitions.

Shapes are the main silhouettes within a category, such as T-Shirts and Jumpers for Tops, or Pumps and Mary Janes for Dress Shoes. A crossing is a combination of a shape with a color, print, fabric, or other attribute - for example Red Tops (a category crossed with a color), Lace T-Shirts (a shape crossed with a fabric), or Turtleneck Jumpers (a shape crossed with a neckline type). Crossings give you a more granular view of how specific attributes are performing within a given product universe.

What is changing?

Until now, we have calculated a trend signal for a shape crossing by taking into account the volume of its broad category. Going forward, it will be calculated relative to the volume of the shape itself.

In practice: Red Jumpers has been measured against the total volume of Tops, meaning that an overall shift in the Jumpers market could affect the metrics of its crossings. We will be measuring it against the total volume of Jumpers instead, so broader shifts in the Jumpers market will no longer influence crossing signals.

How it worked before

Previously, crossing trends were normalized against the total volume of their parent category - Tops, Bottoms, Boots, Sneakers, Bags, etc.

This meant that if the overall volume of Jumpers was declining, because fewer jumpers were being worn, that decline could pull down signals for many crossings under Jumpers. Red Jumpers or Turtleneck Jumpers could appear to be declining not necessarily because red or turtlenecks were losing ground within jumpers, but partly because Jumpers are declining within Tops.

The data was accurate. If fewer jumpers overall were being worn, then fewer red jumpers were being worn too - that's mathematically true. But it made the signals more dramatic, which could make it potentially harder to act on.

Why we wanted to change it

We understood that what would be more insightful when exploring jumper crossings isn't a reflection of whether jumpers as a whole are growing or declining. What you need to know is: within jumpers, what's happening with red? With turtlenecks? With chunky knits?

Whether jumpers are trending up or down as a category might influence how many pieces you put into that shape. But with jumpers, you will probably want a clean, undistorted view of how each attribute is performing within that universe - so you can make sharper decisions on color, print, fabric, and silhouette.

Under the old method, a broad decline in jumpers could make every attribute look like it was decreasing more than it should. A color that was holding its ground within jumpers could still appear to be declining because the overall jumpers market was shrinking.

How it will work

Crossing trends will be normalized against the volume of their parent shape - not the parent category.

Red Bare Shoulder Tops will be measured relative to the total volume of Bare Shoulder Tops. Cropped Bare Shoulder Tops will be measured relative to the total volume of Bare Shoulder Tops. This means that if Bare Shoulder Tops overall are declining, that decline will no longer affect the crossing signals. What you will see is purely how red, cropped, or any other attribute is performing within the Bare Shoulder Tops universe.

The overall direction of Bare Shoulder Tops as a shape will still be visible, but when you explore crossings, those signals will reflect what's happening inside the shape, independently of the shape's broader market trajectory.

For shapes that are relatively stable, crossing metrics won't be significantly affected. In the example of T-Shirts, which is growing by +0.3%, you'll notice little difference. The impact will be more noticeable for shapes experiencing stronger growth or decline.

What this will mean for your data

The underlying data won't change - we will still be using the same images, the same volume, the same panels. What will change is how we compute and present the signal.

As a result, some metrics you currently track may look different once the update goes live. Crossings that currently appear to be declining may show a more stable signal. This isn't the market changing overnight - it's the lens getting sharper.

If you have questions about this upcoming change or want to understand how it might affect specific signals you're tracking, don't hesitate to reach out to your account manager. Our product and fashion experts will be happy to give you more explanation and insights.