The Shopify merchant's buyer's guide to virtual try-on

What virtual try-on actually solves, and what it does not

Virtual try-on solves one narrow, real problem: a shopper looking at a flat product photo cannot see the item on their own body. Closing that gap reduces the imaginative work a shopper has to do before committing, which is why the technology exists at all. That is the whole promise, and it is worth being precise about it before you evaluate any vendor. What try-on does not do is also worth listing. It does not rescue weak product photography, because generated results depend on decent source images. It does not answer whether the medium will fit, which is a sizing question, not a visualization question. This is why some vendors pair try-on with separate size guidance — Looksy, for example, ships a Size Help feature built on the merchant's own size charts rather than folding fit into the render. And no vendor can honestly guarantee a conversion lift for your store. Your catalog, traffic mix, and price points are not anyone else's. Treat every case study as a description of somebody else's shop, and plan to measure your own results — a later section covers how.

Photo, video, and AR are three different products

Photo-based try-on asks the shopper for a single uploaded image and composites the product onto it. It is the lowest-friction approach because it needs no live camera and no special hardware. As a concrete reference point, Looksy's published photo flow works from one uploaded photo, returns the result on the product page in roughly 20 seconds, and requires no app install, no account, and no camera permission. Whatever vendor you evaluate, ask the same questions: how many photos, how long, what permissions. Live video try-on renders the product continuously on a camera feed. It is more impressive in a demo, but it requires camera permission from the shopper and more compute from the vendor, which is why it usually sits on higher tiers — in Looksy's published plans, realtime video appears only on the $79 Scale tier. AR try-on overlays a 3D model, and suits categories like eyewear or furniture where geometry matters more than drape. It requires a 3D asset for every SKU, which is a production cost most apparel merchants underestimate. Match the approach to your catalog before you compare any two vendors on price.

Questions to ask every vendor before you install

Start with the meter. What exactly consumes a credit — every render, only successful ones, or every attempt including retries? Looksy's shopper flow includes a retry option, and its listing does not spell out how retries are billed, so we would put that question to any vendor, ourselves included. A pricing page that defines a credit loosely is a pricing page you cannot forecast. Then ask about shopper photos. Where are uploaded images stored, for how long, and are they ever used to train models? None of this is visible in a demo, and the answer belongs in writing. Ask the same about language support: a listing may state a number of supported languages — Looksy's enumerates 47 — but what that covers can vary. Confirm whether it means the shopper-facing widget, the merchant dashboard, follow-up emails, or all three. Finally, ask about exit. How is billing handled — through Shopify's own invoice or a separate charge? Is there a contract? What is left in your theme after uninstall? A vendor that answers these quickly is telling you something; one that stalls is telling you more.

How to run a free trial that actually tells you something

Decide the shape of the trial before you install. Pick a window — two to four weeks is realistic for a small store — and a subset of products that represent your catalog: a bestseller, a hard-to-photograph item, something in an unusual cut or fabric. A trial run only on your three easiest products tells you nothing. Then test like a shopper, not a merchant. Use your own photos in different lighting, at different angles, on different body types, and mostly on a phone, because that is where your shoppers are. Time the render yourself. Deliberately feed it awkward inputs — busy backgrounds, partial figures — and note what a bad result looks like and whether there is a retry or feedback path when it happens. Know the free tier's constraints going in so they do not surprise you mid-trial. Looksy's free plan, as published, offers unlimited try-ons but caps additional revenue at $100 and applies Looksy branding to results. Every free tier has an equivalent boundary somewhere; find it first. And before the trial ends, uninstall on a duplicated theme and inspect what remains.

Reading pricing models: credits, caps, and overages, with a worked example

Most try-on pricing runs on two meters at once: a credit allowance that limits renders, and a revenue cap that limits how much attributed sales value a plan may generate. The worked example below uses Looksy's published Shopify App Store pricing — our own numbers — because they are public and current, not because the structure is unusual. The published plans: Free at $0, with unlimited try-ons, up to $100 in additional revenue, and Looksy branding on results. Starter at $14.99 per month, with 100 photo credits, a $500 revenue cap, branding removal, and extra credits at $0.14 each. Growth at $29, with 300 credits, a $1,000 cap, and $0.12 overage. Scale at $79, with 600 photo or video credits, no revenue cap, realtime video try-on, and $0.10 overage. To read any such table: divide the monthly fee by the included credits, compare that with the overage rate, then ask which ceiling your store hits first — credits or the revenue cap — and what happens when it does. Ask how additional revenue is attributed, because a cap is only as fair as the attribution behind it. Prefer usage-based billing through Shopify with no contract; that is how Looksy's listing describes its own billing.

The integration surface checklist

First, how the app touches your theme. The clean pattern is a theme app extension: the widget is enabled from the theme editor, nothing edits your theme code, and removal is complete on uninstall. Looksy's listing describes exactly that — theme app extension, no theme code edits, clean uninstall — and whichever app you choose, verify the claim yourself on a duplicated theme rather than taking a listing's word for it. Second, what the app connects to. Check the works-with line on the listing — Looksy's names Checkout, Shopify Admin, Flow, and Klaviyo — then trace each connection to a workflow you actually run. If the app captures shopper emails, ask where the addresses land and which tool sends the follow-up; in Looksy's case, follow-up runs through Klaviyo or Shopify Flow rather than a separate sender. Third, the data features. If there is size guidance, does it build on your own size charts or on a generic model? If there is analytics, what does it record — Looksy's covers what shoppers try on, add to cart, and buy — and can you reconcile those numbers against your own Shopify reports?

Write the measurement plan before you install anything

Take a baseline first. Before installing, record two to four weeks of add-to-cart rate and conversion rate on the specific products where try-on will run. Without that, any post-install number is a guess dressed up as a result, and you will end up comparing against memory. Decide the comparison up front. The cleanest option for most stores is enabling try-on on some products and not others, then comparing like against like. A simple before-and-after works too, but seasonality and promotions muddy it. Be honest about traffic: a store with a few hundred sessions a week will need a long window before any difference means anything at all. Scrutinize attribution. An app's own analytics will report revenue it credits to itself; ask how that attribution is defined and over what window, and cross-check it against Shopify's reports before trusting it. If reducing returns is part of your case for try-on, track return rates as well — and remember that returns lag sales by weeks, so judge them on a longer clock.

Red flags that should end the conversation

Some answers disqualify a vendor outright. No plain, written answer on where shopper photos go, how long they are kept, and whether they train models. Pricing that requires a sales call to learn basic numbers, or credit definitions vague enough that you cannot forecast a monthly bill. A guaranteed conversion lift — nobody who has measured this honestly guarantees it. Watch the mechanics too. Manual theme code edits with no documented uninstall path age badly. Contracts and lock-in on what should be month-to-month usage billing deserve a hard question about why they exist. And treat unverifiable case studies — no store name, no method, suspiciously round numbers — as marketing, not evidence. Finally, read reviews with proportion. A small review count is weak signal in both directions: at the time of writing, Looksy's own listing shows 4.6 stars from six reviews, and we would tell you to weigh that exactly as far as six reviews deserve. Ratings, ours included, are not a substitute for the trial and the measurement plan above. Run those, and the decision mostly makes itself.

Facts in this post come from the public Shopify App Store listing for Looksy. Last checked 29 July 2026.