How to Speed Up WordPress: 30 Tips You Can’t Ignore
To speed up WordPress, test a slow page, identify what delays it, and fix that bottleneck first. Start with public-page caching, server response time, the main image, and scripts that block rendering or interactions. Change one thing at a time and compare the same URL under the same test conditions.
A brochure page, a WooCommerce checkout and a logged-in course page have different caching requirements. Back up your site and use a staging copy for changes that affect scripts, plugins or server settings.
Measure before changing your site
Run PageSpeed Insights on your homepage and at least one important article or service page. Test mobile and desktop separately. Record the URL, date, device setting and results, then repeat runs to see whether the result is consistent.
Core Web Vitals cover loading, responsiveness and visual stability. The good thresholds are LCP at or below 2.5 seconds, INP at or below 200 milliseconds and CLS at or below 0.1, assessed at the 75th percentile of real visits. Google’s Web Vitals guide explains the measurements. A Lighthouse lab run can help diagnose a problem, but its Total Blocking Time is not a measurement of real-user INP.
If the report says there is not enough real-user data, record that limitation. Do not describe a good lab result as a real-user Core Web Vitals pass. Good performance supports visitors, but it does not guarantee a particular search position.
A measured example from How-To WP
On October 8, 2026, a refreshed Ubersuggest PageSpeed test for How-To WP reported mobile LCP of 2.7 seconds, CLS of 0 and Total Blocking Time of 271 milliseconds. Desktop LCP was 0.691 seconds, CLS was 0 and Total Blocking Time was 10 milliseconds. These are a single audit’s lab results, not results from the LocalWP lab or a promised improvement. Real-user INP was unavailable.
That result points toward investigating mobile rendering and script work. It does not justify rebuilding the whole site or buying a different host without checking the request waterfall, cache behavior and actual visitor experience.
30 practical tips to speed up WordPress
1. Choose hosting for your actual workload
A slow server response can delay everything that follows. Check the hosting plan’s CPU, PHP worker, memory and storage limits, the server location, and whether public pages reach a cache. Shared hosting can work well for a small site; a VPS is not automatically faster and may require server administration.
Compare total renewal cost, backups, support and the resources your site needs. Avoid choosing a plan from a historical price screenshot or an uptime claim alone. Our WordPress hosting guide explains what to compare.
2. Configure one public-page cache
A page cache reuses generated HTML so WordPress does not rebuild the same public page for every anonymous visit. Check whether your host already provides one before installing a second cache plugin.
Exclude logged-in or personalized pages as required by your application. Cart, checkout, account and protected course pages must not expose one visitor’s content to another. Test the exclusions, forms and content updates, not just the homepage speed score.
3. Serve appropriately sized images
Resize images for their displayed slots and provide responsive image candidates. Use WebP or AVIF where your tools and browsers support them, but keep screenshots sharp enough to read. Keep originals for later editing.
Check both srcset and sizes: a small card that declares it occupies the full viewport may download an unnecessarily large image. Supply width and height to reserve space and retain the intended aspect ratio.
4. Keep WordPress, themes and plugins maintained
Updates can include security, compatibility and performance fixes, but they are not a guaranteed speed increase. Back up first, review release notes and test important workflows on staging.
After an update, check forms, search, navigation, login and any store or course flows. Retest the same pages and retain a recovery path if the update causes problems.
5. Remove unnecessary plugin work
Review what each plugin does and where its assets load. A single expensive query or third-party integration can matter more than the number of installed plugins. Do not assume every small plugin makes the site slow.
Disable an unused feature on staging, check the results, then remove a plugin only when you know its data and integrations are no longer needed. Avoid installing several tools that perform the same optimization.
6. Use a CDN where it helps
A CDN can serve cached assets closer to visitors. Test from regions your readers actually use and check whether your CDN is caching the relevant resources.
An asset CDN cannot automatically fix a slow database or an uncached personalized response. If you enable edge HTML caching, verify cache keys, bypass rules and purging for your applications.
7. Choose a theme by testing representative pages
Test the layout you actually use, including its fonts, menus, blocks and mobile behavior. A bare demo with no content does not represent a busy production page.
Keep useful design features and remove unused effects where measurement shows a cost. Check the resulting page visually before publishing.
8. Schedule heavy background tasks sensibly
Backups, imports, scans and scheduled jobs can compete for server resources. Look for a pattern between slow responses and scheduled work.
Stagger expensive jobs and ask your host about a reliable scheduled-task setup if needed. Do not disable WordPress cron without replacing the tasks your site relies on.
9. Keep archive pages focused
Use concise excerpts and appropriately sized thumbnails instead of loading many full articles on one archive page. Provide normal crawlable links to older pages.
Keep helpful article summaries and full thumbnail titles visible. Check that an infinite-scroll feature, if used, has accessible navigation and does not load a large archive immediately.
10. Paginate unusually large comment threads
On articles with many comments, test whether comment markup and avatars contribute materially to transfer or rendering time. WordPress discussion settings can split comments into pages.
Preserve navigation to existing comments and check direct comment links after changing the setting. Articles with only a few comments usually do not need this change.
11. Load video players when they are useful
A third-party video embed can load substantial scripts before a visitor presses play. Consider a lightweight preview that loads the real player on interaction, while keeping an accessible way to watch.
Avoid autoplaying large videos. Self-hosted video needs appropriate encoding, delivery and bandwidth; uploading a large unoptimized file is not a shortcut. See embedding videos in WordPress.
12. Reduce requests that do not serve the page
Use the network waterfall to find unused widgets, duplicate trackers, extra font weights and integrations loaded on pages that do not need them. Remove or conditionally load confirmed unnecessary assets.
Do not chase a fixed request count. Modern protocols can handle multiple requests efficiently, and combining every file can make caching or execution worse.
13. Investigate expensive database queries
Use a profiling tool on staging or your host’s application monitoring to identify repeated or slow queries. Review the feature generating the query before making database changes.
Persistent object caching can help some database-heavy workloads when supported and configured correctly. It complements page caching; it does not make every query or frontend asset faster.
14. Keep a sensible revision history
Post revisions are useful for recovering edits. A large revision history is not automatically the cause of a slow public page, so measure before deleting it.
If you choose to limit retained revisions, back up the database and agree on a useful number with editors. Avoid bulk cleanup tools that remove recoverable content without a review.
15. Investigate unwanted asset traffic
If logs show other sites consuming significant bandwidth by linking directly to your images, ask your CDN or host about targeted hotlink controls.
Test legitimate image access, feeds, search previews and your own CDN domains. Broad referer rules can block legitimate requests and are not a universal performance fix.
16. Use bot protection for abusive traffic
A firewall or rate limit may reduce abusive requests before they consume application resources. Confirm the traffic pattern in logs and choose rules that fit it.
Do not block legitimate search crawlers, payment callbacks or users simply to improve a synthetic score. Security rules need their own functional checks.
17. Fix mixed-content and HTTPS configuration at the source
Use HTTPS URLs for assets and internal links, and correct the WordPress/site configuration when moving domains. Avoid repeated redirect or replacement work on every request where a safe source correction is possible.
Back up before a database search-and-replace and use a tool that understands serialized WordPress data. Keep the HTTP-to-HTTPS redirect.
18. Use a supported, compatible PHP version
Your host controls the PHP runtime. Check the current version under Tools > Site Health > Info > Server and consult the PHP support schedule. Test a supported target with your theme and plugins before changing production.

Our PHP update guide covers preparation and rollback. A newer runtime can help, but it does not replace checking inefficient application code.
19. Check the server configuration, not just its name
Apache, Nginx and LiteSpeed can all support fast WordPress sites when configured appropriately. The server label alone does not tell you whether a particular plan will meet your needs.
Look at caching, PHP capacity, compression and your measured workload. Use configuration instructions specific to your host instead of copying an Apache rule to an Nginx server.
20. Diagnose Time to First Byte
TTFB includes work before the first response byte reaches the browser. Network distance, redirects, connection setup, cache misses and backend processing can contribute.
web.dev describes 0.8 seconds or less as a rough good TTFB target; TTFB itself is not a Core Web Vital. Compare several anonymous requests and a logged-in workflow rather than treating one response as conclusive.
21. Measure DNS before changing providers
DNS lookups are one part of the connection process and are often cached. Check whether DNS time is consistently a meaningful part of the delay.
A paid DNS plan is not automatically required to make WordPress fast. Reliability and correct records matter; plan a provider change carefully to avoid disrupting email or the website.
22. Protect login traffic without confusing it with authentication
Automated login traffic may consume resources, but knowing a username or the default login URL is not enough to log in. Use strong unique passwords, suitable multi-factor authentication and appropriate rate limits.
A custom login URL can reduce some unwanted requests, but it does not replace those controls. Test account recovery and integrations before changing the login route.
Google Tag Manager organizes tag deployment. It does not automatically combine all tracking requests or make third-party scripts free to run.
Remove duplicate tags, use appropriate triggers and respect consent requirements. Confirm required analytics events still fire after changes. See our Tag Manager setup guide, then audit the tags themselves.
24. Check response compression
Your host or CDN may already compress text responses using Brotli or gzip. Inspect response headers and transferred sizes before adding server rules.
Compress HTML, CSS and JavaScript through the supported server configuration. Already-compressed image/video formats usually offer little benefit from another compression layer.
25. Link directly to final URLs
Update internal links that take unnecessary detours through old paths or domains. Preserve useful permanent redirects for external links and old bookmarks.
Keep required canonical and HTTPS redirects. Avoid replacing a correct redirect with duplicate public pages merely to remove one request.
26. Cache static assets without freezing personalized HTML
Browser caching and server-side page caching are different. A long browser lifetime can suit versioned static files because a changed filename or URL lets visitors fetch a new version.
Do not apply a blanket 15-day Expires rule to text/html. It can leave visitors with stale pages and is inappropriate as generic advice for account, cart, checkout or protected content. Ask your host to define HTML caching and bypass rules for the application, and test both logged-out and logged-in behavior.
On Apache, Expires rules require the relevant module; Nginx does not use .htaccess. Prefer your host’s documented controls and verify the actual response headers instead of pasting a universal snippet.
27. Prioritize the main mobile content
Make the main heading and useful content available without waiting for nonessential features. Do not lazy-load the image that is the largest visible content element near the top of the page.
WordPress added native image lazy loading in version 5.5 in 2020. Current WordPress also includes loading optimizations, so inspect the output before adding another lazy-loading system. Use lazy loading for suitable below-fold images and verify the result on a narrow screen.
28. Use modern HTTP support supplied by your host
HTTPS and HTTP/2 or HTTP/3 support can improve delivery characteristics, but the benefit depends on the network and workload. Ask your host what is available.
Do not make a protocol change a substitute for reducing excessive JavaScript or correcting a slow backend. Measure the full page experience.
29. Evaluate AMP only for a specific requirement
AMP is an optional publishing approach. It is not required for a fast mobile site or as a general WordPress SEO step.
First optimize the standard responsive page. A separate AMP version introduces templates, feature compatibility and canonical relationships that need ongoing maintenance.
30. Defer only scripts that can safely wait
Identify render-blocking resources in a waterfall and preserve script dependencies. WordPress supports dependency-aware loading strategies; blindly adding async to everything can break execution order.
After changing loading behavior, test the menu, search, table of contents, forms and consent controls. Test logged-in, checkout or course behavior separately when affected. Keep the improvement only if the measured result and functionality both hold up.
Retest and keep a record
Repeat your mobile and desktop tests using the same URLs and settings. Compare several runs and note which change you made. Test the visible page and key actions as well as the metrics. If performance is unchanged, undo an unnecessary change and investigate the next measured bottleneck.
For broader guidance, see the WordPress optimization handbook. If you need help tracing a specific issue, my WordPress repair service starts with the symptoms and an agreed scope.
Frequently asked questions
Will a perfect speed score put my site first in Google?
No. Search relevance, content quality and many other signals still matter. Use performance measurements to improve the reader’s experience, not as a ranking guarantee.
Should I install two caching plugins?
Usually, start with one coordinated caching setup and any host-provided cache. Overlapping tools can create conflicting rules and make troubleshooting harder. Confirm which layer caches which resources.
Can I cache a WooCommerce or LearnDash site?
Public pages may be cacheable, but personalized and protected pages need suitable exclusions. Follow the application and host guidance, and test different user states before enabling broad HTML caching.
What should I change first?
Start with the biggest repeatable bottleneck in the measured page: for example a slow uncached server response, an oversized main image or heavy scripts. There is no universal plugin or setting that is the first fix for every site.
If you would prefer help investigating a recurring WordPress problem, you can request a WordPress repair assessment. Share the affected page and the measurements you collected so we can agree on the next step.

What a thorough post-every blogger should pay attention to this. I’m Pinning it!
Thank you so much, Carolyn! 🙂