How Pearl automates online eyewear ordering, prescription validation, and at-home try-on logistics
Pearl is a direct-to-consumer eyewear brand selling prescription glasses and sunglasses online, with virtual try-on, prescription parsing, and an at-home try-on programme that ships frames before the customer commits. Behind every pair sits a workflow stack that turns a face and a script into a packaged order with the right lenses.
Selling glasses online is harder than selling shoes. Every order has a prescription that has to be readable, a pupillary distance that has to be accurate, a frame fit the customer has not tried, and a lens build that takes a few days at the lab. Pearl automates the whole chain so a customer can pick frames at home, get the right lenses ground for their script, and have the package arrive looking like it was hand-fitted.
The four pain points Pearl's automation has to solve
Prescriptions arrive in every format imaginable. Paper photo, scanned PDF, emailed JPEG from the optometrist, doctor's handwriting. Pulling sphere, cylinder, axis, and PD reliably is the difference between a glasses order and a refund.
Customers will not buy frames they have not tried on. Glasses sit on the face. Without virtual try-on or physical try-on at home, the conversion curve collapses and the return rate climbs.
Lens manufacturing has its own SLA. Single-vision, progressive, blue-light, photochromic — each lens build has a lab lead time. Promising 5-day delivery without knowing the lens path is a refund waiting to happen.
Returns and remakes are expensive and slow. A glasses return is not just postage; it is a remake, a refit, and a customer who is not seeing properly in the meantime. Reducing remakes is the margin lever.
Four automation patterns that keep Pearl moving
Prescription capture + validation
Customers upload a photo or PDF of their script; the platform extracts sphere, cylinder, axis, PD, and add power, then validates against safe ranges before the order is built.
Virtual + at-home try-on
AR try-on previews frames on the customer's face; an at-home programme ships up to five frames to try before any lenses are cut. Conversion goes up; returns go down.
Lens-path-aware delivery promises
The promised ship date is computed from the actual lens build path — single-vision vs. progressive, coatings, lab location — so the date the customer sees is the date the package arrives.
Remake-rate analytics
Remakes are tracked back to root cause — script mis-parse, fit issue, lens defect, customer preference — and the loop is closed. Margin improves where the actual problems live.
The four-stage pipeline
Every order on Pearl runs through the same four-stage shape — capture the prescription, let the customer try the frames, build the lenses, ship the packaged order. The same pipeline serves a single pair of readers and a high-index progressive build.
Case study: Pearl
Pearl
Challenge
Sell prescription glasses online to customers who have never tried the frames, parse a prescription that may arrive as a phone photo of a paper script, build the right lenses on a promised date, and keep remake rates low enough to protect margin.
Solution
Pearl built an ordering pipeline that extracts and validates prescriptions automatically, lets customers try frames virtually or at home before committing, computes shipping promises from the real lens build path, and closes the loop on remakes by tracing every one to a root cause.
More from E-commerce & D2C
Huel — D2C subscription nutrition → Veo — sports camera D2C → Orchard — music distribution + royalties →Frequently asked questions
How does Pearl read a prescription from a phone photo?
The customer uploads a photo or PDF of their script and the platform extracts sphere, cylinder, axis, pupillary distance, and add power. Values are validated against safe optical ranges before the order is built — so a misread script does not become a wrong pair of glasses.
How does Pearl let customers try on frames before buying?
Frames can be previewed virtually with AR try-on, or up to five physical frames can be shipped to the customer's home through the at-home try-on programme. Lenses are only cut once a frame is chosen, which cuts both returns and remakes.
How does Pearl decide a delivery date?
The promised ship date is computed from the actual lens build path — single-vision vs. progressive, coatings, lab location — instead of a generic SLA. The date the customer sees at checkout is the date the package arrives.
Run your eyewear D2C the same way
Byteflow gives you the workflow shape — capture, try-on, build, ship — so your customers get the right lenses on the promised date.
Start automating →Easy automation. For everyone.