Eyewear suggestions based on facial analysis

Online eyewear retail is particularly challenging as the optician doesn’t have the opportunity to get to know their customer casually during the (in-person) eye examination, before moving to the frame selection process. Hence, selecting relevant suggestions to a first time visitor has been much of a hit-and-miss game.

With this challenge in mind, Styleriser has developed a professional instant facial analysis tool, allowing users to generate highly relevant recommendations simply by sharing their selfie.

The software detects the user’s predominant facial shape and evaluates facial lines and complexion, understanding the particular user’s style-mix in order to find matching glasses for them.

In a previous step, the entire eyewear stock has been tagged with a comprehensive set of style attributes so that the matching (or recommending) happens instantly:

  1. User shares a selfie
  2. Face is analysed on the cloud in a few seconds
  3. Product suggestions are displayed immediately along with explanations
  4. User makes informed buying decision from a small and highly relevant range of products
Shutterstock 2266610809 optician styleriser eyewear suggestions

Two possible settings

A shopper enters the web shop and uses the facial analysis and eyewear recommendation tool to see best matches

  • User does not have a particular model in mind and wants the retailer to help them with suggestions
  • They see the service button/teaser on the page and start the recommendation process by taking a selfie
  • A few seconds after the selfie upload, a small range of suitable glasses is presented to them along with a few sentences where their own face shape and features are summarized
  • The presented selection is matching their age range, gender, style type and color tone
  • The user has the opportunity to use additional filters representing their choices or preferences. Here they can choose between ‘progressive’ or ‘harmonic’ to see either more extravagant or more classic products in line with their individual face type
  • With the reassurance that the presented selection is customized to their own appearance, the user is more confident to go ahead and select frames
  • This process increases sales and reduces the rate of bad buys

An online shopper found frames they like but is insecure to follow through with the purchase, opts for virtual style check

  • A customer has placed a couple of frames they like into the basket but is insecure whether those products really look good on them
  • They see the style check button/teaser and start the recommendation process by taking a selfie
  • A few seconds after the upload, they get the results for their facial analysis in a few sentences along with match scores for the glasses they previously selected
  • In addition to the glasses they selected, a small number of further products with higher match scores are presented
  • This professional consulting process helps them with their purchasing decision and increases sales, significantly decreases cart abandonment and returns

While this is particularly relevant for online eyewear retail, the AI assistant is an equally powerful tool supporting the in-store consultant helping their customer find best matching frames.