Three tools, built for three different problems, with one thing in common. Each answers a question about the customer that no amount of product copy could answer. This case study covers what each one does, why it exists, and what running everything inside the browser buys you.
At a glance
| Tool | Problem it answers | Where it runs |
|---|---|---|
| Lash Mirror | What will these lashes look like on my eyes | In the browser, camera-based |
| Nail Mirror | What will this shape and length look like on my hand | In the browser, camera-based |
| VetAge | How old is my dog biologically, by breed and size | In the browser, no camera |
Lash Mirror
Choosing false lashes online is mostly guesswork. The product photography shows the lashes on a model, in studio lighting, on somebody else’s eye shape, and the customer is left to imagine the rest. It’s one of the main reasons people buy the wrong style, and one of the reasons they hesitate before buying at all.
Lash Mirror asks for camera access, then uses a face landmark model running inside the browser to find the upper lash line on each eye. The chosen style is drawn along that line in real time, following the actual shape and angle of the customer’s own eye. Switching between styles updates the preview instantly, so the difference between a wispy and a cat eye is something the customer can see on their own face rather than infer from two product shots.
The styles aren’t invented for the tool. They match the collections the site already sells against, and selecting one surfaces a genuine product from that collection with its live price and a link to the product page.
Alongside the try-on there’s a guide to the eleven eye shapes the site merchandises for. Choosing a shape explains what defines it, recommends a style and applies that style to the preview, so a recommendation becomes something the customer can look at rather than something they have to trust.
Nail Mirror
The same idea applied to press-on nails: try shapes, lengths and finishes on your own hand using the camera.
It’s a narrower problem than lashes and a more visual one. Length in particular is almost impossible to judge from a product photograph, because the reference is somebody else’s hand.
VetAge
Not everything needs a camera. VetAge works out a dog’s biological age by breed and size, using veterinary data rather than the seven-year rule everybody has heard and nobody can source.
The seven-year multiplier is the kind of widely repeated figure that’s easy to publish and wrong in a way that matters, because ageing varies enormously by breed and size. Giving an owner a number grounded in real data is more useful, and it opens onto something a product page can’t do: showing the veterinary care milestones and the products that suit that particular life stage.
That’s the commercial logic of a tool like this. It segments the customer by something real, then recommends against that segment, rather than asking them to work out which life stage their dog is in.
Why everything runs in the browser
For the camera tools this is the answer to the first question customers ask. The face detection runs inside the customer’s own browser, so no video is uploaded and nothing is stored.
That’s worth stating plainly on the page rather than burying in a privacy policy, because the alternative is a customer who declines camera access and never sees the tool work. A privacy answer given up front converts better than a reassuring one given afterwards.
It has a practical benefit too. A self-contained tool with no server behind it is simple to host, simple to move between platforms and has no per-use running cost.
What I learned
The recommendation matters more than the rendering. The eye shape guide in Lash Mirror, which explains a shape and then applies the style that suits it, does more commercial work than the try-on itself, because it turns a preview into advice.
The second learning is that a tool is a segmentation opportunity. VetAge asks two questions and can then recommend against a genuine life stage. Most sites ask for an email address and learn nothing.
Limitations
Camera tools depend on the customer granting access and having reasonable light, so there always has to be a route for someone who won’t or can’t use one. A try-on is a preview rather than a photograph of the product on the customer, and shouldn’t be presented as more than that. VetAge is a guide based on published data, not veterinary advice for an individual animal.
Services used
Interactive tools and calculators and app and tool development. The lash work sits alongside the FalseEyelashes.co.uk case study, and the ring size tools cover the measurement side of the same idea.
FAQ
Do virtual try-on tools upload customer photos?
The ones here don’t. The face detection runs inside the customer’s own browser, so the video never leaves their device and nothing is stored anywhere. Saying that clearly on the page matters, because it’s the question customers ask before deciding whether to grant camera access.
Does a try-on tool actually reduce returns?
It can, where returns come from customers misjudging how something looks on them rather than from quality or delivery problems. Working out which of those is driving your returns is the first job, because a try-on won’t help with the other two.
Do these tools work on a phone?
Yes, and phones are where most customers use them, since that’s where the camera is and where the shopping happens. All three are built to run on a phone first, with the desktop version treated as the secondary case rather than the other way round.
Why not use the seven-year rule for dog age?
Because ageing varies a lot by breed and size, so a single multiplier gives a misleading answer for most dogs. Using published veterinary data gives an owner something more accurate, and it lets you recommend care and products against a genuine life stage.
Got a question your product pages can’t answer?
Get in touch and tell me what customers keep getting wrong. That’s usually where a tool earns its place.