[wizzwizz4]

Thought of a solution. Rather than treating models as interchangeable, consider grouping by model (and then perhaps grouping models by category). If people know how a particular item of clothing fits on them, and how it fits on the model, they can generalise to how other clothes might fit on the model vs on them.

Like old-school media reviews: not everyone agrees with Roger Ebert's film criticism, but after reading his reviews of films they've watched, most people can read a review of his and have a reasonable idea what they will think of the film.

[Deekshana_Reddy]

I would like to dig deeper to fully understand the challenge. When you mention grouping models by fit, is the main issue that the current presentation doesn’t help you imagine how clothes would look on your body? For example, do you feel the models lack diversity in proportions (like height, width, or shape), or is it about not knowing which body type matches yours? If so, what would make this clearer or more relatable for you?

[wizzwizz4]

I don't know your labelling scheme. I do know what my measurements are, or what clothes I have already (and to what extent they fit / don't fit). In general, hierarchical classification schemes are unsuitable for an inherently multidimensional / unstructured search task. What if, for example, I'm half-way between two of your "body types"? Do I have to keep switching pages?

If I could, say, look up a particular item of clothing, see the pictures, pick a model and see other clothes that fit that model, that would be superb. Or have that pre-populate the model measurements sliders in a search, which I can then edit. (Really, I'm just trying to describe the features, not the implementation, here.)

It might be my lack of experience, but I have no idea what the "male" body types are supposed to mean. The diagrams do not appear to match the models, and there is overlap between the categories. I don't know how I'd shop for "male" clothes using this system.

Btw, I noticed the "chat with AI" button that doesn't do anything. I think an "AI" could work quite well, if you go old-school with it (behaviour trees implementing an expert system version of the script you work out for your manual consultations) rather than trying to get an expensive, unreliable LLM thing going (as is the modern fad). See "How to get people to say and type what computers can understand" by Elizabeth Zoltan-Ford (https://www.speech.kth.se/~edlund/bielefeld/references/zolta...) and related literature.

[Deekshana_Reddy]

The idea behind our system is that every body shape can broadly fall into one of these categories: Triangle for a broad shoulder frame Rectangle for a leaner, straighter build Rhomboid for a muscular frame Apple for either plus-size or someone with a belly These categories simplify the process while remaining relatable to most shoppers. For users who are unsure about their body shape, we’re currently building a chat feature to help identify their specific body shape more accurately and enhance their shopping experience.

Alternatively, you can book a call with me, where I personally create a 3D model of you and explain why you fall under a particular category.

I’d love for you to test the "Chat with AI" feature when it’s ready! Let me know if you’d like to try it out.