Guide · Shopify
How much main-thread time does an accessibility widget's script use on a Shopify store?
How this guide was made: Written with AI assistance and checked against the sources linked on this page.
Updated
Short answer: Across 10 public Shopify homepages that load an accessibility widget, the widget's own script was attributed a median of 756 ms of blocking time and 1,049 ms of main-thread time in our lab runs. Blocking time is the part of a task that runs past 50 ms on the browser's main thread, the stretch in which the page cannot respond to a tap or a keypress.
Check your store's homepage for accessibility issues, free. We open it like any visitor and only read the page.
Free. No email needed. It usually takes about a minute.
Want it fixed? Our AI agent fixes accessibility issues in your Shopify theme's code within 48 hours of you connecting your store, on an unpublished copy you preview and publish yourself. $390 until 30 Nov (Black Friday), then $490 one time. See the store service.
Key takeaways
- We measured one thing: the time Lighthouse attributes to the accessibility widget's own script, per store.
- Median across the 10 stores: 756 ms blocking time and 1,049 ms main-thread time.
- Store by store, the widget's blocking time ranged from 49 ms to 1,962 ms, so the median hides a wide spread.
- This is a lab test on one PC with 10 stores. It is not field data from real visitors.
- It says nothing about whether a widget helps or harms accessibility.
What we measured
- Tool: Lighthouse 12.8.2, run from the command line (npm package lighthouse@12.8.2), performance category only.
- Browser: Google Chrome 157.0.8103.51, headless.
- Device and network: Lighthouse's simulated mobile setting (mobile screen emulation, simulated throttling), not a real phone.
- Runs: 3 runs per store, each started only after the previous run had ended plus a gap of at least 10 seconds. The number for each store is the median of its 3 runs.
- When: 15 November 2026, from 15:37 to 18:00 UTC.
- What we read: in Lighthouse's third-party summary, the blocking time and main-thread time attributed to the widget's own entity (the company that serves the widget script).
- How stores were chosen: public Shopify homepages. We confirmed Shopify from each page's public HTML on 15 November 2026. We detected the widget from the script the homepage loads. We did not sign in, add anything to a cart or submit any form.
- Size bucket: from the store's public products list. S is 1 to 50 products, M is 51 to 250, and L is 251 or more.
- Across stores: the headline figures are the median of the 10 per-store medians.
All 30 runs for these stores finished normally, and none was left out.
The results
Widget's own cost as Lighthouse attributes it, median of 3 runs per store:
| Store | Size bucket | Widget | Widget blocking time (median ms) | Widget main-thread time (median ms) |
|---|---|---|---|---|
| Store A | S | UserWay | 214 | 767 |
| Store B | S | UserWay | 49 | 158 |
| Store C | S | UserWay | 1,962 | 3,426 |
| Store D | M | accessiBe | 901 | 964 |
| Store E | M | UserWay | 87 | 621 |
| Store F | L | accessiBe | 1,651 | 2,741 |
| Store G | L | UserWay | 370 | 1,350 |
| Store H | L | AudioEye | 1,546 | 12,835 |
| Store I | L | accessiBe | 945 | 1,002 |
| Store J | L | UserWay | 612 | 1,095 |
| Median of the 10 stores | 756 | 1,049 |
The widget in each row is the one the store's homepage loads. The sample is too small to compare vendors, so we make no comparison between them.
Limits of this test
- It is a lab run on one PC, not field data from real visitors.
- The machine's CPU was under load from other work during the runs, which can move the numbers.
- Single runs varied a lot. For Store A, the widget's blocking time across its three runs was 135, 214 and 516 ms.
- 10 stores is a small sample.
- Lighthouse attributes time to a script's entity, so code that several parties share may be counted differently from how a store owner would count it.
- This measures one cost only. It does not show whether a widget helps or harms accessibility.
What blocking time means for shoppers
web.dev defines blocking time per task: "The blocking time of a given long task is its duration in excess of 50 milliseconds." A long task is one that runs on the main thread for more than 50 milliseconds. If someone taps or types while a long task is running, the browser must wait for the task to finish before it can respond. The figures here are the share of that blocking time Lighthouse assigns to the widget's script in a simulated mobile run, which is a lab estimate and not what any one visitor experienced.
Check your store's homepage for accessibility issues, free. We open it like any visitor and only read the page.
Free. No email needed. It usually takes about a minute.
Want it fixed? Our AI agent fixes accessibility issues in your Shopify theme's code within 48 hours of you connecting your store, on an unpublished copy you preview and publish yourself. $390 until 30 Nov (Black Friday), then $490 one time. See the store service.
FAQ
What is blocking time?
It is the part of each long task on the browser's main thread that runs past 50 ms. web.dev's Total Blocking Time metric adds these up for the stretch after First Contentful Paint. Here we report only the amount Lighthouse attributes to the accessibility widget's script.
Why is main-thread time bigger than blocking time?
Main-thread time counts all the time the widget's script spent running on the main thread. Blocking time counts only the portion of each task beyond 50 ms, so it is the smaller of the two.
Does this show whether an accessibility widget is worth having?
No. We measured one cost, the script's own time on the main thread. We did not test what the widget does for visitors, and the test cannot tell you that.
How VibeFix works on a store
Once you choose to go ahead and connect your store, our AI agent works through the issues in your report on an unpublished copy of your theme. We review each change before you preview it, and you publish it yourself. Nothing goes live until you do. The report lists anything still open. How it works for Shopify stores
Want us to check your store?
Enter your store address in our free Shopify homepage accessibility checker. No email needed, and your report usually appears on the page in about a minute. Your email is optional; if you give it, we send you the link too. It is an automated, read-only check of your store's public homepage, not a full audit with a person testing by keyboard and screen reader: we don't sign in, submit anything or change anything. It tests only what software can test of WCAG 2.0 and 2.1, Levels A and AA. It doesn't test the newer WCAG 2.2 criteria, such as target size. It finds some accessibility issues, not all of them. The report is not legal advice.
Sources
- web.dev, "Total Blocking Time (TBT)", published 7 November 2019, last updated 15 October 2025: https://web.dev/articles/tbt. Quote: "The blocking time of a given long task is its duration in excess of 50 milliseconds."
- Chrome for Developers, Lighthouse, "Reduce the impact of third-party code", last updated 2 May 2019: https://developer.chrome.com/docs/lighthouse/performance/third-party-summary
- Chrome for Developers, Lighthouse, "Total Blocking Time", last updated 9 October 2019: https://developer.chrome.com/docs/lighthouse/performance/lighthouse-total-blocking-time
- VibeFix measurement, 15 Nov 2026: Lighthouse 12.8.2 lab runs on 10 public Shopify homepages, method in "What we measured" above.
Want it fixed?
Our AI agent fixes accessibility issues in your Shopify theme's code within 48 hours of you connecting your store, on an unpublished copy you preview and publish yourself. $390 until 30 Nov (Black Friday), then $490 one time. See the store service.
It starts with a free check of your public homepage. No email needed.
More guides · Shopify store accessibility · About · Your privacy choices