Closing the social gap in Warby Parker's Virtual Try-On

Hypothesis: Adding a second-opinion prompt inside the Virtual Try-On view reduces the gap between try-on and Add to Cart — and turns the share into an acquisition channel.

Brand of the Week

Warby Parker changed the way people buy glasses online.

Before Warby, buying prescription eyewear on the internet meant committing to a frame you had never seen on your face. Warby's Home Try-On programme — five frames sent to your door, free — built trust by solving that problem with a physical object. When the box arrived, you tried the frames on at home. You asked someone nearby. You took a photo. The decision became social.

As Virtual Try-On became the primary digital fitting experience, the convenience improved. The social element did not carry over.

Today, Warby Parker's Virtual Try-On lets a visitor see any frame on their face using their phone or laptop camera. The technology is good. The rendering is accurate. But the experience is designed for one person.

Word-of-mouth is a genuine part of how Warby Parker grows. Reviews reference friends, campus conversations, recommendations passed between people who already own a pair. That consideration happens — it just happens off-site, after purchase, in places Warby can't reach.

The purchase funnel ends before that social loop can start.

The challenge

The Virtual Try-On solves "I need to see it on my face." It doesn't solve "I want to ask someone." The moment a shopper most wants a second opinion is the exact moment the current experience gives them nothing to do with that feeling.

Warby's word-of-mouth is real, but it arrives too late. Reviews like "I bought these — apparently everyone on my campus does too" show that social validation is already part of how people talk about these frames. It just lands after the sale, not before it.

So what would we A/B test?

Control: Warby Parker's Virtual Try-On with no sharing mechanism. A visitor tries on frames, views the result alone, and either navigates to Add to Cart or exits the modal.

Variant (built in MidaGX): A "Second Opinion" button added directly inside the Virtual Try-On modal. The flow is short: a prompt, a 3-2-1 countdown, a capture from the live canvas — not a product shot, but what a friend would actually see sent to them. From there, native share with the photo attached.

The share copy invites the recipient to try the frame on too.

That last detail matters. The recipient lands on the Warby Parker try-on experience, prompted to try the same frame. A second-opinion request becomes a warm introduction for someone who wasn't in the funnel at all.

Our hypothesis

Surfacing a second-opinion prompt inside the Virtual Try-On, at the moment fit doubt and commitment doubt are highest, reduces the friction between try-on and Add to Cart. The mechanism doesn't change the product or the page — it changes what a visitor can do with the feeling they already have.

And by inviting the recipient to try the frame, each share becomes a low-friction entry point for a new visitor. Warby is already generating word-of-mouth consideration in reviews, threads, and campus conversations. This test gives that consideration a path back into the purchase funnel, where it can influence an outcome.

The CRO principle underneath

This test is built on messenger-driven decisions. For products where appearance matters — particularly eyewear, where other people see you in the frame every day — a friend's reaction carries more weight than any brand copy. Shoppers already know this. They already want that input. The question is whether the product experience can provide a mechanism for getting it.

When Home Try-On existed, the mechanism was the physical box. Bringing frames home meant other people were in the room. That natural moment of seeking a second opinion was baked into the format. The digital experience removed the friction of waiting for a box — and in doing so, also removed the social layer that came with it.

This test is a proposal to rebuild that layer digitally, inside the experience where the decision is already happening.

Want to test this kind of experiment on your own site? Try Mida free — no account needed to get started with MidaGX.