The short answer
The Visualizer adds one script to your storefront, about 65 KB compressed, and the browser only requests it once your page has finished loading. We ran Lighthouse on two Shopify stores, a busy retail store and Imersian's own demo store, five times with Imersian and five times with it blocked, on mobile and desktop. Performance score and layout shift did not change beyond normal run-to-run variation on either store. The cost we could measure is blocking time: on the lean demo store, on Lighthouse's slow-phone profile, the script added about 60 to 70 ms after the main content had painted. On desktop it was too small to report. The 3D view and room imagery load only when a shopper taps the button.
How it loads
Imersian connects to your store through a native Shopify app, which keeps your products in sync with Imersian and makes product onboarding and implementation easier. The Visualizer button on your product page is a normal part of your theme, styled in Shopify like any other button. Behind it, Imersian loads in three stages.
- Page load: nothing from Imersian. Your theme, images and checkout scripts load as they always do. The button is already in the page, so it does not pop in later or push content around.
- After load: one small script, then a few tiny lookups. Once the browser fires its load event, the storefront adds a script with the async attribute, served from a CDN with Brotli compression. It then checks whether the product has a 3D model, in a few requests of under 1 KB each. On both stores we measured, the script request started within 3 ms of the load event.
Repeat visits: cached in the browser. After a shopper's first visit, their browser keeps a copy of the script, so repeat page views don't download it again.
- On tap: the Visualizer opens in its own frame. Only when a shopper taps the button does the Visualizer load, inside an iframe on a separate Imersian domain. The 3D model, room imagery and Spatial AI processing all happen there. Rendering takes about 10 seconds and reconstruction about 3 to 4 seconds. Shoppers who never tap download none of it.
What we measured
We ran Lighthouse 12 in October 2026 on ten product pages. Each page was tested five times with Imersian and five times with it blocked, on mobile and desktop, and we compared the medians. Here is what changed when Imersian was switched on.
- Script size
- 65 KB
- Requested after the page has finished loading
- Extra blocking time
- 60 to 70 ms
- Throttled mobile profile, after the main content has painted
- Layout shift and score
- No change
- Beyond normal run-to-run variation
- 3D view and room imagery
- On tap
- Nothing loads until a shopper taps the button
Lighthouse 12, median of 5 runs with and without Imersian, October 2026.
| Measure | Mobile: on / blocked | Desktop: on / blocked |
|---|---|---|
| Performance score | 60 / 61 | 95 / 92 |
| Largest Contentful Paint | 11.7 s / 9.8 s | 1.4 s / 1.7 s |
| Layout shift (CLS) | 0 / 0 | 0.008 / 0.007 |
| Total Blocking Time | 230 ms / 159 ms | 0 ms / 0 ms |
How to read this
- Blocking time is the one real cost. On a slow phone, the script runs two short tasks that add about 60 to 70 ms of blocking time. They run after the page has loaded, once the main content is on screen. On desktop the cost was too small for Lighthouse to report.
- On a busy store, it is a small share. The busy store spent about 2.3 seconds blocked on mobile. Imersian's part of that was 27 ms. Score and LCP moved a little in both directions between runs, by less than the normal spread from one run to the next.
- Lighthouse is a lab tool. It does not measure Interaction to Next Paint, which comes from real-user data. Check your own store's figure in the Shopify Web performance report or PageSpeed Insights before and after enabling the Visualizer.
What you control
- The button. Text, icon and placement are set in Shopify, the same way as the rest of your product page. It renders with your theme, so it adds no layout shift.
- Which products are enabled. You choose products with a Shopify tag, and Imersian syncs tagged products every 24 hours.
- What shoppers download. Shoppers who never tap the button get only the after-load script. The Visualizer itself loads on tap.
Run the same check on your store
You can repeat our test in Chrome in about ten minutes. The same method works for any third-party script.
- Open a product page that shows the Visualizer button, in a Chrome Incognito window.
- Open DevTools (F12), go to the Lighthouse tab, choose Navigation, tick Performance, pick Mobile, and run it three times. Note the median score, LCP, CLS and Total Blocking Time.
- In the Network tab, open the request blocking panel (Ctrl+Shift+P, then type Show Request blocking) and add two patterns: *static.imersian.com* and *api.imersian.com*. Turn blocking on.
- Run Lighthouse three more times and compare the medians. Repeat both sets with Desktop selected.
- For real-user data, including Interaction to Next Paint, check the Shopify Web performance report or the field panel in PageSpeed Insights.
Run each set at least three times. A single Lighthouse run on a busy store can swing by 10 points on its own.