# Review widgets and page speed

> Do review apps slow down your store? How to measure what any review widget costs your product pages with Lighthouse, and how Reytd keeps its widget fast.

Every script a review app adds to your product pages has to be downloaded, run and drawn, and slow product pages lose shoppers. This guide shows how to check what any review widget costs your store, and how Reytd keeps its own to about 9 KB, with the stars showing in less than 50 ms.

_Figure: The average review app: 172.7 KB. Reytd: 8.7 KB._

About 19 times less code. The average of 5 popular review apps, each measured on its stores' product pages (14 pages), on 27 October 2026: scripts and styles as the browser downloads them, compressed, with an empty cache. Reytd measured the same way, on the live widgets in these docs.

## Why review widgets slow pages down

- **Large scripts.** A review app's script is loaded on every product page, often with a framework, styles and features the page doesn't use. The browser has to download and run all of it, which competes with your own theme on slower phones.
- **Scripts that block the page.** A script added without `async` or `defer` stops the browser from reading the rest of the page until it has downloaded and run.
- **Waiting on the app's servers.** A widget that asks the app's own servers for each product's reviews waits while they look them up, for every shopper on every page.
- **Content that jumps.** When stars or reviews appear after the page has drawn, with no space kept for them, the text below them moves down. Google counts this as layout shift (CLS), one of its Core Web Vitals.
- **Busy work while shoppers tap.** A heavy script keeps the browser busy, so taps and clicks answer late. Google measures that too, as Interaction to Next Paint (INP).

## Check any review app yourself

You don't have to take any app's word for its speed, ours included. Chrome has everything you need, and it doesn't take sides.

### With Lighthouse

1. Open one of your product pages in Chrome, in an incognito window, so your extensions don't count (unless you've allowed some in incognito).
2. Open the developer tools (`Cmd+Option+I` on a Mac, `F12` on Windows), then the **Lighthouse** tab.
3. Choose **Mobile** and **Performance**, and click **Analyze page load**.
4. Note **Total Blocking Time**, **Largest Contentful Paint** and **Cumulative Layout Shift**. Under **Reduce unused JavaScript** and **Avoid enormous network payloads**, look for the review app's scripts.
5. To see what the app costs, block it and measure again: in the **Network** tab, right-click its script and choose **Block request** (**Block request URL** in older Chrome), then run Lighthouse once more. The difference is the app.

_Figure: The Lighthouse panel with Mobile and Performance chosen, and the Analyze page load button_

In Chrome's developer tools: Mobile, only Performance, then Analyze page load.

Run each measurement a few times, as the results vary from run to run.

### With the Network tab

1. In the developer tools' **Network** tab, tick **Disable cache**, so the browser downloads everything as a first-time shopper would, and type the review app's domain in the filter, for example `reytd.com`.
2. Reload the page, and scroll to the end: many widgets, Reytd's reviews widget included, only load when a shopper gets near them.
3. Read the number of requests and the **Transferred** size at the bottom: that's what shoppers download, compressed. It includes each response's headers, a few hundred bytes a request, so it comes out a little above the file sizes on this page, which leave them out.

| Name | Type | Size |
| --- | --- | --- |
| reytd.js | script | 8.7 kB |
| id-MTAwMQ.json | fetch | 0.1 kB |
| settings.json | fetch | 0.7 kB |

3 requests, 9.4 kB, Reytd's star rating on a product page: each file's compressed size, without the headers the Network tab adds

For your store's real visitors rather than a test, [PageSpeed Insights](https://pagespeed.web.dev) shows Chrome's data from real shoppers, for sites with enough traffic.

## How Reytd keeps its widget fast

- **One small script.** About 9 KB compressed, written without a framework. Review carousels and helpful votes have small scripts of their own, loaded only where they're used. Our build fails if the script grows past the size this page says, so the number stays true.
- **Out of your page's way.** The script tag has `async`, so the browser keeps drawing your page while it downloads, and the widgets draw once your page is ready.
- **Files near your shoppers.** A product's rating and reviews are written as files whenever they change, and a CDN serves them from a location near each shopper. Showing the stars and a product's latest reviews never waits for a database, so a traffic peak on your store is the CDN's job, not ours. Reytd's own server only answers for reviews past a product's latest 100, sorted or filtered lists of products with more, helpful votes and the usage counts, which go out in the background.
- **3 small requests for the stars.** The script, then your store's settings and the product's rating at the same time.
- **Reviews when they're needed.** The reviews widget usually sits far down the product page, so it loads and draws itself only when a shopper is about to scroll to it.
- **Stars that stay put.** Each star badge's line is kept before the script runs (see [Layout shift](#layout-shift)).
- **No cookies or tracking libraries.** The widgets send anonymous usage counts in the background, without cookies, batched so they never compete with your page.

_Figure: The shopper's browser loads the script and the review files from a CDN near them. Reytd's app writes the files to the CDN when reviews change, and isn't asked when a shopper opens a page._

A shopper's page view never waits for Reytd's servers: the widget reads files a CDN already has, written ahead whenever a review is approved, hidden or replied to.

## How we measure it

"Less than 50 ms" is the time from the browser asking for the script until the star rating is drawn, on a good connection (10 ms to the server, 100 Mbps), measured 20 times with the real script and files in a test setup:

- **A shopper's first page** on your store, with nothing cached: less than 50 ms.
- **Their next product page,** with the script already in the browser's cache: faster still.

Not counted: opening the first connection to `widgets.reytd.com`, which a script from any other address needs on a shopper's first page. On a phone on a slow mobile connection, everything takes longer, and there the network sets the pace rather than the widget.

## Layout shift

Nothing on your page moves when Reytd's star badges appear. The install snippet has a small `<style>` line beside the script that keeps each star badge's line from the moment the page is drawn, before the script has even arrived. A product without reviews keeps its line too, just without stars, so the text below never jumps up or down.

_Figure: Without a reserved line, the price and the button move down when the stars appear. With the line reserved, nothing moves._

On our demo page, on a slowed mobile connection so the stars arrive after the page has been drawn, the Cumulative Layout Shift is 0.

If your script tag was added without the snippet's `<style>` line, add this to your theme's stylesheet (or, inside a `<style>` tag, to the `<head>`):

```css
.reytd-stars {
    min-height: 1.5em;
}
```

The reviews widget usually sits near the end of the product page and draws itself before a shopper scrolls to it, so nothing in view moves when it appears. Placed near the top instead, it draws right away and can push what's below it down, as can a carousel: give their spot a height of its own if they sit above the fold.

## Next

- [Add review widgets to your store](https://reytd.com/docs/widgets.md): the script and the widgets, step by step.
- [Star rating badge](https://reytd.com/docs/widgets/star-rating.md) and [Reviews widget](https://reytd.com/docs/widgets/reviews.md): where to place them.
- [Start free](https://reytd.com/register): the widgets are just as fast on every plan.
