Someone hovers a product link, pauses, then clicks. The browser still starts from zero: DNS, HTML, critical CSS, the first paint of the next page. Full-page cache helps the server answer fast. It does not start the next navigation while the pointer is still deciding. Speculative loading closes that gap. The page behind a link begins fetching while interest is already visible, so by the time the click lands, navigation feels already underway.
Lightify includes speculative loading: prefetch on hover, focus, or touch start, using the Speculation Rules API where the browser supports it, with exclusions for carts, logout, and admin URLs. Full-page cache and resource hints still matter as the wider performance stack. This post stays on speculative loading itself.
Lightify is the DaftPlug performance plugin. This post is only about WordPress speculative loading prefetch before the click. Not unused CSS strip, not delay or defer JavaScript deep dives, not YouTube placeholders, not the WebP pipeline, not database cleanup, not a Core Web Vitals dashboard essay, not emoji or dashicons cleanup. Those live elsewhere. For the hint layer that prepares origins and critical assets without guessing the next document, keep WordPress resource hints (preconnect, prefetch, preload) nearby.
What Lightify actually ships here
Stay inside the real feature set:
- Speculative loading so the next page can begin fetching before the click commits
- Prefetch triggers on hover, focus, or touch start
- The Speculation Rules API where the browser supports it
- Exclusions for carts, logout, and admin URLs so personalized or sensitive destinations are not prefetched as if they were brochure pages
- Works alongside full-page cache and resource hints as context (those are not this article’s deep how-to)
This post does not invent undocumented prerender percentages Lightify does not publish, per-template speculation allowlists that are not on the marketing pages, or promises that every browser will prerender the full next document identically. The product promise is speculative loading with hover, focus, or touch-start prefetch, Speculation Rules where supported, and cart, logout, and admin exclusions. Stay there. Unused CSS, delay/defer JS essays, YouTube placeholders, WebP, database cleanup, CWV dashboards, and emoji or dashicons cleanup are parallel tools, not this article’s job.
Why speculative loading matters on WordPress
Clicks are late signals. By the time someone clicks, they have already hovered or focused. Waiting until click to start HTML work leaves free intent on the table. Prefetch on hover, focus, or touch start uses that earlier signal.
Full-page cache and prefetch solve different moments. Cache makes the first anonymous byte of a URL cheap when the server is asked. Speculative loading asks earlier, while the pointer is still on the link. You want both on shareable content URLs. Neither replaces the other.
Resource hints prepare origins and files. Speculation prepares documents. Preconnect and preload (see the related post) warm connections and critical assets. Speculative loading targets the next page navigation itself. Keep the jobs distinct in your mental model so you do not disable one thinking the other already covers it.
Commerce and account URLs must stay out of the guess. Prefetching a cart, logout, or wp-admin URL can create odd session side effects or wasted work. Lightify’s exclusions for carts, logout, and admin URLs keep speculation honest. That is the same spirit as keeping dynamic pages out of full-page cache: speed where HTML can be shared safely, caution where identity matters.
Mobile uses touch start, not hover. Phones do not hover. A speculative loading path that only understood mouse hover would miss most real traffic. Touch start support matters for WordPress sites with a mobile majority.
Browser support is uneven, and that is fine. Speculation Rules API where supported means capable browsers get the modern path. Older browsers simply navigate as they always did. You are not required to polyfill a science project.
How to enable speculative loading in wp-admin
Assume Lightify is installed and you are comfortable testing on a staging copy or a low-traffic window first. Speculative loading is a front-end navigation feature. You will verify with DevTools or a careful click path, not only with a lab score screenshot.
1. Decide which journeys should feel instant
Catalog to product, blog index to article, and hub page to key landing URLs are good candidates. Cart, checkout, logout, and wp-admin are not. Write the happy path you care about in one sentence: “From shop archive to product, prefetch should start on hover or touch.” That sentence becomes your acceptance test.
2. Confirm full-page cache is already sane on shareable URLs
If anonymous content pages are still cold and uncached, speculative loading will prefetch slow HTML. Fix or confirm Lightify full-page cache behavior on the URLs you expect to prefetch. Remember cart, checkout, and account exclusions on the cache side are a separate (already documented) concern. This article assumes shareable pages can be served quickly when asked.
3. Open Lightify and enable speculative loading
In wp-admin, open Lightify. Find the speculative loading controls Lightify exposes. Enable speculative loading. Keep the first pass boring: speculation on, trust the built-in exclusions for carts, logout, and admin URLs, leave unused CSS strip, delay/defer JS experiments, YouTube placeholders, WebP pipelines, database cleanup, and CWV dashboard tuning alone until prefetch behavior is verified.
4. Respect hover, focus, and touch start as the trigger set
You do not need a custom JavaScript hover library. Lightify’s feature is built around prefetch on hover, focus, or touch start, with Speculation Rules API where supported. Mentally accept that mouse users, keyboard users, and touch users get appropriate signals. Do not invent undocumented delay sliders or “aggressiveness” percentages that are not on the marketing pages.
5. Keep cart, logout, and admin URLs excluded
Confirm the exclusion story is active in your understanding even if you never build a custom list on day one. Do not add speculative loading and then manually point marketing experiments at logout links. If you use a custom membership logout URL outside the usual patterns, treat that as an ops checklist item: verify it is not being prefetched in DevTools before you call the rollout done.
6. Smoke-test the happy path
- Open a public hub URL over HTTPS in a Chromium-based browser that supports Speculation Rules when possible.
- Open DevTools Network (or the browser’s speculation/prefetch instrumentation you already use).
- Hover or focus a content or product link you expect to prefetch. On a phone, begin a touch on the link without completing navigation if your test harness allows observing the request.
- Confirm a prefetch (or speculation) request starts for the destination before a full click navigation.
- Complete the click and confirm the navigation feels immediate relative to a cold control load in a fresh profile.
- Hover or focus a cart, logout, or admin URL and confirm it is not treated like a normal brochure prefetch target.
- Spot-check that resource hints (preconnect/prefetch/preload) you already rely on still make sense alongside speculation. They are complementary, not rivals.
If nothing prefetches, confirm the feature is enabled, confirm the browser supports the path you are testing, confirm you are not on an excluded URL, and confirm a second performance plugin is not stripping Lightify’s front-end output.
7. Document the enablement for the next teammate
Write one line in your runbook: “Lightify speculative loading enabled on [date]. Prefetch on hover/focus/touch start. Cart, logout, and admin excluded. Re-test after major menu redesigns and after adding new custom logout or membership endpoints.” Speculation gets blamed when a new mega menu prefetches dozens of junk URLs, or when a host-level experiment blocks the Speculation Rules response.
Who this is for / when to skip
Content and brochure WordPress sites where the next article or service page is an obvious hover target.
WooCommerce catalogs that want product pages to feel ready after someone pauses on a card, while cart and logout stay excluded.
Agencies optimizing client sites that already have full-page cache and still feel a gap between click and paint.
Publishers with dense internal linking where readers scan many titles before committing.
Skip or defer if the site is almost entirely authenticated app UI with no public document navigations worth prefetching, or if you are mid-migration and URLs are unstable. Also pause if you have not yet excluded dynamic commerce paths from full-page cache: fix cache honesty first, then add speculation.
What good looks like
- A visitor who hovers, focuses, or touch-starts a normal content link triggers a prefetch toward that URL.
- Speculation Rules API is used where the browser supports it.
- Cart, logout, and admin URLs are excluded from the speculative path.
- Full-page cache still serves shareable pages quickly when prefetch asks for them.
- Resource hints remain in place for origins and critical assets (see the related post).
- Unused CSS strip, delay/defer JS deep dives, YouTube placeholders, WebP, database cleanup, CWV dashboards, and emoji or dashicons cleanup were not required for this win.
- You did not invent custom prerender percentages or claim every browser prerenders identically.
If prefetch never fires, confirm enablement, browser support, exclusions, and that another optimization plugin is not removing the speculation output. Cold cache on the destination can make a successful prefetch still feel slow. Warm or cache the shareable URL, then re-test.
Common mistakes
Prefetching cart or logout “to make checkout faster.” Those destinations are personal or destructive. Trust Lightify’s exclusions. Do not fight them for a demo.
Expecting speculation to fix an uncached origin. If every HTML response is slow to build, prefetching early still waits on a slow build. Pair with full-page cache on shareable URLs.
Confusing resource hints with speculative loading. Preconnect and preload help files and connections. Speculative loading targets the next document navigation. Keep both; do not delete one thinking it duplicates the other.
Enabling every Lightify toggle on the same afternoon. If unused CSS, delay JS, and speculation all land together, debugging regressions becomes guesswork. Enable speculation, verify, then schedule the next experiment.
Judging only from a mobile Lighthouse screenshot. Lab tools may not simulate hover or Speculation Rules the way a real Chromium session does. Use a real hover/touch path.
Ignoring keyboard focus. Focus is part of the trigger set. Keyboard users deserve the same head start as mouse users.
Testing only admin-bar sessions. Logged-in views can differ from anonymous public HTML. Spot-check a logged-out profile on public URLs.
Practical checklist before you call it done
- Confirm shareable content URLs are eligible for fast HTML (full-page cache in good shape).
- In Lightify, enable speculative loading.
- Confirm your mental model includes prefetch on hover, focus, or touch start.
- Confirm carts, logout, and admin URLs stay excluded.
- In a capable browser, hover a product or article link and confirm prefetch activity.
- Complete the click and compare feel against a cold navigation in a fresh profile.
- Negative-test a cart or logout link so it is not treated like a brochure prefetch.
- Keep resource hints configured sensibly (related post) without treating them as a substitute.
- Leave unused CSS, delay/defer JS deep dives, YouTube placeholders, WebP, database cleanup, CWV dashboard projects, and emoji or dashicons cleanup for separate passes.
- Re-test after major menu or theme link changes.
- Write the enablement date and happy-path URL in the team runbook.
FAQ
Is speculative loading the same as full-page cache?
No. Full-page cache stores and serves HTML quickly when requested. Speculative loading starts that request earlier, on hover, focus, or touch start, so the click feels closer to instant. Use both on shareable pages.
Will this prefetch my cart?
Lightify’s speculative loading excludes carts, logout, and admin URLs. That is intentional. Personalized and sensitive destinations should not be treated like brochure prefetch targets.
Do I still need resource hints?
Yes, for their own job. Resource hints (preconnect, prefetch, preload) prepare origins and critical assets. Speculative loading prepares the next page navigation. See the resource hints post for that layer.
What if the browser does not support Speculation Rules?
Lightify uses the Speculation Rules API where supported. Unsupported browsers continue with normal navigation. You are not required to maintain a separate polyfill project for this feature.
Can speculative loading replace delay JavaScript or unused CSS work?
No. Those are different performance jobs. This post stays on prefetch-before-click. Schedule asset experiments separately so you can measure what actually changed.
Takeaway
WordPress visitors often reveal the next URL before they click. Speculative loading uses that signal. Lightify prefetches on hover, focus, or touch start, leans on Speculation Rules where supported, and keeps carts, logout, and admin URLs out of the guess, while full-page cache and resource hints continue to do their own jobs.
Lightify handles speculative loading alongside the rest of the performance stack, without asking you to maintain a hand-rolled prefetch script that ignores commerce exclusions. Enable it once. Verify with a real hover or touch path. Keep sibling optimizations as separate enablement work.
One Pro membership unlocks Lightify plus Progressify, Generatify, Rankify, and Snapshotify. New Pro plugins join existing memberships. Plans differ by site count only (Pro $37.99 for 1 site, Pro Plus $74.99 for 5, Pro Max $138.99 for 100), with a 14-day money-back guarantee. See pricing when the next page should start loading before the click lands.
Turn speculative loading on. Exclude what must stay dynamic. When a real product or article link is already fetching before the click, navigation finally matches how people browse.