A slow click feels broken.

Not technically broken. The link works. The page eventually loads. But to the customer who just tapped your product, service, booking, or contact page, the delay feels like hesitation. That hesitation is expensive when the visitor is already deciding whether they trust you.

That is why the Speculation Rules API deserves attention from small businesses. It is not another full redesign trend. It is a browser feature that lets your site tell Chrome and other Chromium-based browsers which pages a visitor is likely to open next, so the browser can start loading those pages before the click happens. MDN describes speculation rules as JSON instructions for prefetching or prerendering future navigations, and Chrome says the API can make future navigations quicker or even instant.

For a small business website, that can mean your service pages, quote form, cart, booking page, or contact page feels faster without changing your CMS or redesigning your brand.

What Speculation Rules actually do

Speculation Rules give the browser a short list of likely next pages. The browser can then do one of two things.

Prefetch downloads the HTML document for a future page. MDN explains that prefetching downloads the response body of the referenced page but does not load its subresources. That makes it lighter, safer, and easier to use across more pages.

Prerender goes further. Chrome explains that prerendering behaves like opening the future page in an invisible background tab and activating it when the user navigates. That can create a near-instant load, but it also uses more memory and bandwidth, so you should reserve it for paths where the next click is very predictable.

Think of it like a restaurant. Prefetch is the kitchen pulling the ingredients out because the waiter knows what people usually order next. Prerender is cooking the whole plate before the customer asks. One is low-risk. The other is powerful, but you do not do it for every menu item.

Why this matters for small businesses

Website speed is usually discussed like a developer problem. It is not. It is a sales problem.

Chrome says prerendering can improve Core Web Vitals with near-zero Largest Contentful Paint, reduced layout shift because load-time shifts happen before the page appears, and better Interaction to Next Paint because loading may be complete before the visitor interacts with the new page (Chrome prerendering guide). Core Web Vitals are not abstract scores. They affect how fast and stable your pages feel while visitors are comparing you against another company.

Shopify rolled out a platform-wide Speculation Rules configuration for Liquid storefronts in June 2025 and reported an average 130 millisecond improvement on desktop and 180 millisecond improvement on mobile across loading metrics in supported scenarios (Shopify Performance). Those numbers may sound small, but they happened across a huge number of real storefront journeys.

The bigger lesson is not “copy Shopify.” The lesson is that native browser preloading can make normal websites feel sharper without throwing away the platform they already run on.

Another public example is Ray-Ban. CoreWebVitals.io summarizes Ray-Ban’s Speculation Rules case study and reports mobile Largest Contentful Paint dropping from 4.69 seconds to 2.66 seconds, a 43% improvement, after prerender rules were added to product listing pages. The same summary reports mobile conversion rates increasing by 101% and desktop conversions by 156% (CoreWebVitals.io). That is an ecommerce example, but the pattern applies to any site with predictable next steps.

A visitor lands on a service page, then clicks pricing. A homeowner reads about roofing repair, then clicks the quote form. A manufacturer buyer reads a capability page, then requests a quote. These are exactly the kinds of paths where prefetching can help.

Where to use Speculation Rules first

Start with clicks that are both valuable and predictable. Do not try to speed up every page. That is how performance work turns into waste.

Good first candidates:

  • Homepage to your top service pages, especially if analytics shows the same two or three pages get most homepage clicks.
  • Service page to quote, booking, contact, financing, or pricing pages.
  • Category page to product pages, or product page to cart, when the site is ecommerce.
  • Blog post to a related service page or lead magnet, when the content has a clear commercial next step.
  • Location page to directions, phone, appointment, or estimate pages.

Chrome’s implementation guide says sites can include rules directly in HTML, inject them with JavaScript, or serve them through an HTTP header (Chrome implementation guide). For many small business sites, the easiest starting point is a small inline rule added to the site template, or a plugin if the CMS supports it.

WordPress also has a practical path. Chrome’s guide points to the Speculative Loading plugin for WordPress as one implementation option (Chrome implementation guide). That does not mean every WordPress site should blindly install it and move on. It means a small business does not need custom application architecture to test the idea.

Where not to use it

Speculation Rules are useful because they load pages before the visitor asks for them. That is also why you need guardrails.

Avoid speculating pages that change account state, trigger actions, or contain stale user-specific information. MDN’s example excludes logout links, add-to-cart query patterns, nofollow links, and links marked with a no-prerender class. That is a good mental model. If a URL does something instead of showing something, be careful.

Shopify warns against being too aggressive. Their team says aggressive eagerness increases the risk of loading pages visitors never open, wasting data and possibly slowing the current page. They also note that prerender can execute scripts, including analytics, which can lead to overreporting if analytics providers do not account for speculative loads (Shopify Performance).

Chrome gives similar guidance. Chrome says prerendering uses extra memory and network bandwidth, so site owners should avoid over-prerendering and only prerender when navigation likelihood is high.

For small businesses, that means you should usually start with prefetch, not prerender. Prefetch is less dramatic, but it is safer. Once you know the site behaves correctly, you can test prerender on a few high-confidence paths.

A practical rollout plan

Do this like a business experiment, not a science project.

First, look at analytics for the last 30 to 90 days. Find the pages where visitors often take the same next step. If 38% of people who visit your “emergency plumbing” page click the “request service” page, that is a candidate. If people scatter to twenty different pages, leave it alone for now.

Second, start with prefetch using moderate or conservative eagerness. Chrome documents the eagerness settings and explains that moderate eagerness on desktop waits for a 200 millisecond hover, while mobile can trigger on pointerdown because there is no hover event. That makes moderate a good balance for many small business sites because it waits for a signal before spending resources.

Third, exclude risky URLs. Your rules should skip login, logout, cart mutation, admin, search query, filter query, and form submission URLs unless a developer has reviewed the behavior. This is especially important on ecommerce, membership, healthcare, legal, financial, and booking sites.

Fourth, measure before and after. Chrome recommends measuring the actual performance impact in analytics rather than assuming the feature helped (Chrome prerendering guide). Track navigation timing, Core Web Vitals, conversion rate, and form starts for the pages you changed. Do not declare victory because a lab score improved on one test run.

Fifth, expand slowly. If prefetch works on service-to-contact journeys, try category-to-product. If that works, test prerender on the one or two highest-intent paths.

A simple example

Here is a stripped-down example for a service business. This rule tells the browser to prefetch links that you mark with a data attribute:

<script type="speculationrules">
{
  "prefetch": [{
    "source": "document",
    "where": {
      "selector_matches": "a[data-prefetch-next]"
    },
    "eagerness": "moderate"
  }]
}
</script>

Then your quote button could look like this:

<a href="/get-started/" data-prefetch-next>Request a Quote</a>

This approach is boring on purpose. It does not try to guess the whole website. It marks the next step you care about and lets the browser prepare it when the visitor shows intent.

If you want broader rules, you can use URL patterns. But for small businesses, selector-based rules are often easier to reason about. Your team can decide exactly which buttons and links deserve the treatment.

What about Safari and Firefox?

This is not universal browser magic. Sentry’s ecommerce guide notes that the Speculation Rules API is available in Chromium browsers, while Safari and Firefox users need fallback approaches such as framework-level prefetching when available (Sentry).

That limitation does not make the work pointless. A faster experience for Chrome and Edge users is still valuable, especially if those browsers represent a large share of your traffic. But your foundation still matters: compressed images, lean JavaScript, good hosting, cached pages, clean templates, and sensible third-party scripts. The 2025 Web Almanac page weight chapter explains that heavier pages require more data transfer, more CPU work, and more device memory, which can hurt performance, accessibility, and SEO (HTTP Archive Web Almanac). Speculation Rules can make the next click faster, but they do not excuse a bloated page.

The analytics catch

If you use prerender, analytics can get weird.

Chrome’s 2026 post on the “prerender until script” origin trial explains that full prerender can execute JavaScript before the user actually navigates, and that analytics may fire early if the provider does not account for speculation (Chrome Developers Blog). That matters because bad analytics lead to bad decisions.

This is another reason to start with prefetch. Prefetch does not execute page JavaScript according to MDN’s explanation of prefetch behavior. For most small businesses, fewer measurement surprises beat a slightly faster but harder-to-audit setup.

If you do test prerender, confirm that your analytics, heatmaps, chat widgets, call tracking, and form tools behave correctly. Watch for inflated pageviews, fake form impressions, chat popups firing too early, stale cart counts, or booking widgets loading in a hidden state.

The bottom line

Speculation Rules are not a replacement for fixing a slow website. They are a smart layer on top of a decent one.

If your site takes eight seconds to load because the homepage has oversized images and six marketing scripts, fix that first. If your mobile layout jumps around, fix that first. But if your site is reasonably healthy and your next-step paths are predictable, Speculation Rules can make important clicks feel faster without a full rebuild.

That is the kind of performance work small businesses need more of: targeted, measurable, and tied to revenue paths.

If you want help finding the highest-impact speed improvements on your website, start here. We’ll review the pages that actually drive leads and show you what is worth fixing first.