Virtual try-on has become a standard feature of coloured contact lens retail, and it is genuinely useful — but not for the reason most listings imply. It answers a colour question. It does not answer a fit question, and the two are routinely conflated in product pages.
What the technology actually does
An AR try-on renders a coloured iris overlay onto a detected pupil position in a live camera feed. The overlay is a texture, not a simulation: it applies the lens artwork’s colour, graphic diameter and limbal ring to a circle in screen space. The result is a good approximation of hue and pattern placement, and a poor approximation of everything that depends on the physical lens.
Three variables are outside what the render can show:
- Base curve and fit. Centration, movement and edge lift depend on the corneal curvature and the lens geometry. A rendered overlay is always perfectly centred.
- Colour under different light. Camera white balance and screen calibration shift hue noticeably. A grey that reads neutral in the render can read blue in daylight.
- Opacity and layering. The render composites at full opacity. A real lens sits over the natural iris, and the visible result depends on the pigment density and the wearer’s own eye colour.
Where it genuinely helps
Used honestly, try-on shortens the shortlist. A buyer arriving at a listing with forty patterns cannot evaluate forty; the tool reduces the set to three or four candidates that are worth ordering as samples. That is a real reduction in sample cost and in return rate, and it is the argument that holds up.
It also helps the retailer, not just the buyer. Listing pages with try-on hold attention longer, and attention is the metric marketplaces rank on. The commercial case is engagement, not accuracy.
What to publish instead of a fit claim
If the goal is to reduce returns rather than to increase time on page, the specification is more useful than the render. Four figures do most of the work:
| Parameter | What it tells the buyer |
|---|---|
| Overall diameter (DIA) | Total lens size, and whether it will cover the visible iris comfortably |
| Graphic diameter | The coloured zone; this is what changes the perceived size of the eye |
| Base curve | Fit range on the cornea; the figure that decides comfort, not appearance |
| Water content and material | Handling, dehydration behaviour and how the lens behaves over a wearing day |
A product page that states graphic diameter and base curve next to the try-on button answers the question the render cannot. Buyers who return lenses usually return them for fit, and fit is a specification problem.
How to read a try-on demo before you trust it
- Check whether the overlay tracks head movement. A static image with a pasted circle is not try-on; it is a mockup.
- Compare two patterns in the same lighting. If the demo only shows one look under one light, it is a marketing asset, not a decision tool.
- Look for the stated graphic diameter. A try-on without a published graphic diameter is showing an effect with no reference point.
None of this makes try-on useless. It makes it a discovery tool with a clear boundary: it decides which four samples to order, and the samples decide the rest.
FAQ
Does AR try-on show how the lens will fit? No. The render places a colour overlay on a detected pupil position; it cannot reproduce base curve, centration, edge lift or how the lens behaves on the cornea.
Why does a colour look different in the render and on the eye? Camera white balance, screen calibration and full-opacity compositing all shift the result. A real lens sits over the natural iris, so pigment density and the wearer’s own eye colour change the final appearance.
What should a product page publish instead of a fit claim? Overall diameter, graphic diameter, base curve, and water content with material. Those four figures answer the fit question the render cannot.
