How to Speed Up a WordPress Website: 9 Fixes in the Order That Matters

To speed up a WordPress website, work in this order: compress and resize your images, turn on page caching, check your hosting is not the bottleneck, then clean up plugins and the database. That order is not a preference — images typically account for the majority of a slow page's weight, and caching removes the per-visit server work, so the first two fixes routinely deliver more than the other seven combined. Everything after them is refinement.

Speed matters for unglamorous, commercial reasons: visitors abandon slow pages, Google uses page experience among its ranking signals, and every campaign you run pays for clicks that a slow page then wastes. This guide assumes a typical small-business WordPress site and no developer on staff — each fix is either a setting, a reputable plugin decision, or a clear 'ask your host/developer for exactly this' instruction. Measure first (next section), fix in order, measure again.

Step 0: Measure Honestly Before Touching Anything

Run your homepage and your most important inner page through PageSpeed Insights and note the mobile scores and the top listed problems — mobile is the stricter, more representative test, because most of your visitors are on phones and mid-range ones at that. Record the numbers somewhere; without a baseline you cannot prove a fix worked, and 'feels faster on my office Wi-Fi' is not evidence.

Read the diagnosis in plain terms: if the report complains about image sizes, that is Fix 1. 'Reduce server response time' points at hosting or caching (Fixes 2–3). Long lists of unused CSS/JavaScript point at theme and plugin bloat (Fixes 4–5). The report is your prioritised shopping list — this guide simply translates it.

Reading PageSpeed Like an Owner: The Three Numbers That Matter

PageSpeed reports dozens of metrics; three decide whether visitors experience your site as fast, and they map cleanly onto this guide's fixes:

• LCP (Largest Contentful Paint) — how long until the main content (usually the hero image or headline) appears. Target: under 2.5 seconds on mobile. Improved by Fix 1 (that hero image is often the LCP element itself), Fix 2 and Fix 3.
• INP (Interaction to Next Paint) — how quickly the page responds when a visitor taps or clicks. Target: under 200 milliseconds. Hurt by heavy JavaScript and plugin bloat; improved by Fixes 4 and 8.
• CLS (Cumulative Layout Shift) — whether content jumps around as it loads (the button you were about to tap moves). Target: under 0.1. Caused by images without dimensions, late-loading banners and font swaps; improved by Fix 6 and by themes that reserve space properly.

You do not need to memorise thresholds — you need the habit of reading these three, on mobile, for your homepage and your money page, before and after any fix. Everything else in the report is supporting detail.

Fix 1: Images — Usually Half the Problem

A page carrying 4 MB of photos loads slowly on any connection, and WordPress sites accumulate full-size images because uploading the camera original 'worked'. The fix has three parts: resize images to the size they are actually displayed (a hero shown at 1,600 px wide does not need a 5,000 px file), compress them (quality 80–85 is visually identical for photos), and serve modern formats — WebP is typically 25–35% smaller than the same photo as JPG.

The full workflow, with settings and worked examples, is in our guide to compressing images without losing quality, and converting an existing library is covered in converting images to WebP. Doing this one fix properly is the single highest-return hour most site owners can spend.

Fix 2: Page Caching — Stop Rebuilding Every Page

By default, WordPress rebuilds each page from the database for every visitor. Caching stores a finished copy and serves that instead, which is why a good caching setup can cut load times dramatically on content pages. Many quality hosts include caching; otherwise a well-regarded caching plugin configured with page cache (and browser cache) enabled is the standard route. After enabling, browse your own site logged-out — cached sites occasionally cache something they should not, like a logged-in view, and it is better that you see it first.

Shops and membership sites need care here: carts, checkouts and account pages must be excluded from caching, which reputable caching tools handle when told the site is a store. If your site sells, treat 'enable caching' as 'enable caching with store exclusions verified by a test order'.

Fix 3: Hosting — The Ceiling You Cannot Optimise Past

If server response time stays poor after caching, the host is the ceiling. The cheapest shared hosting tiers pack many sites onto weak servers, and no plugin fixes physics. Signs hosting is your problem: the admin dashboard itself is sluggish, speed tests blame 'server response time' rather than content, and the site slows at busy times of day. Moving to a reputable WordPress-focused or higher-tier plan is unglamorous and, at that point, the honest fix.

Cost context: hosting is a modest, predictable line compared with what slow pages cost commercially — the maintenance budgeting in our US maintenance cost guide shows where hosting sits in a sensible annual budget.

Fixes 4–6: Plugins, Database and Fonts

Fix 4 — Plugin audit. Every active plugin is code loaded into your site's life. Deactivate and delete anything unused, replace heavy all-in-one tools doing one small job with lighter single-purpose ones, and never keep two plugins doing the same job (two SEO or two security plugins is a classic). Fix 5 — Database cleanup. Years of post revisions, trashed comments, expired transients and orphaned data bloat the database WordPress queries on every uncached request; a reputable optimisation plugin cleans this safely — take a backup first, always. Fix 6 — Fonts. Web fonts are heavy: limit the site to one or two families and only the weights you actually use, and prefer system-font stacks for body text where brand allows.

Worked through in that order, fixes 4–6 rarely transform a site alone, but they compound — and they remove the slow creep that makes a healthy site sluggish again within a year.

Fixes 7–9: CDN, Updates and What Not to Do

Fix 7 — A CDN (content delivery network) serves your static files from locations near each visitor; for audiences spread across countries it helps meaningfully, for a purely local business on good hosting it is optional. Fix 8 — Keep everything updated: current WordPress, theme and plugin versions are often faster as well as safer, and running a currently supported PHP version at your host is free speed. Fix 9 — Lazy-load below-the-fold media (now largely built into WordPress) so the browser fetches what is visible first.

And what not to do: do not install three 'speed booster' plugins that overlap and fight; do not chase a perfect 100 score at the cost of breaking your booking or checkout (a working 85 beats a broken 100); and do not pay for 'speed optimisation' that reports only desktop scores afterwards. Measure mobile, on your money pages, before and after — that discipline is the whole method. More performance and maintenance writing lives on the Asad Webs blog.

Worked Example: A Slow Homepage, Fixed in Order

An illustrative walk-through, in the order this guide prescribes, on a typical small-business homepage — a services firm, WordPress, four years of accumulated uploads. Starting measurement (mobile): page weight 6.8 MB, main content appearing at about 5.9 seconds.

After Fix 1 (images, about 50 minutes): the hero was a 2.9 MB camera original displayed at 1,600 px wide; resized and compressed, it became 210 KB. The six gallery images got the same treatment and the library switched to WebP. Page weight fell to 2.4 MB and the main content time to about 3.1 seconds — the largest single improvement of the whole exercise, exactly as the chart above predicts.

After Fix 2 (caching, 20 minutes): the host's page cache was enabled with browser caching, verified logged-out. Repeat and cached-page responses became near-instant; the mobile score climbed again, mainly on server-response diagnostics.

After Fixes 4–6 (an afternoon): eleven unused plugins deleted (including a second, forgotten SEO plugin), the database cleaned of years of revisions, and the font stack cut from four weights of two families to two weights of one. Final state: page weight about 1.1 MB, main content at roughly 1.8 seconds on the same mobile test. Fix 3 (hosting) was assessed and left alone — with caching on, server response was no longer the bottleneck. Total effort: one focused day, no rebuild, no new theme. Most slow small-business sites hold a similar day of wins; the order is what makes it a day instead of a month of random tweaking.

WooCommerce Speed Notes

Stores feel every fix differently, because a shop's most important pages are the ones caching cannot simply freeze. Three notes that save store owners pain. First, the product page is your homepage: most store traffic lands there, so apply Fix 1 ruthlessly to product galleries — the five-shot discipline and weight targets in our product photo guide are a speed policy as much as a sales one. Second, cart and checkout stay dynamic: they must be excluded from page caching (reputable tools do this when told the site is a store), and their speed is governed by hosting quality and plugin count — which is why Fix 3 matters more for stores than for brochure sites. Third, count your extensions like plugins: every WooCommerce add-on is code on every page; the audit in Fix 4 applies with interest, especially to marketing pop-ups and live-chat tools loaded sitewide for a feature used on one page. Test with a real order after every speed change — a fast store that cannot take payment is not fast, it is closed.

Speed Triage: What to Do With an Hour, a Day, or Nothing to Spend

One hour, $0: measure (Step 0), then resize and compress the images on your five most-visited pages and delete every plugin you are not using. This is the highest-value hour in this guide. One day, $0–small: the full sequence in the worked example — whole-library image pass, caching configured and verified, database cleanup, font tidy. Most small-business sites finish the day in a different performance class. Ongoing, the honest spend: if server response is still the bottleneck after caching, the money goes to hosting (Fix 3) before it goes to any optimisation service — paying experts to polish a site on a weak server buys improvements the server then caps. Whichever rung you are on, end the same way: measure again, on mobile, and write the numbers down next to the first set.

The 9 Fixes at a Glance

Fix	What it does	Typical payoff	Effort
1. Images	Resize to display size, compress, serve WebP	Usually the largest single win	An hour on your top pages
2. Page caching	Serves stored copies instead of rebuilding each page	Large on content pages	A setting or one plugin
3. Hosting	Removes a slow-server ceiling	Decisive when server response is the bottleneck	A plan change or migration
4. Plugin audit	Deletes unused and duplicate plugins	Moderate, and stops future creep	30 minutes of decisions
5. Database cleanup	Clears revisions, transients and orphaned data	Small to moderate	One optimisation plugin, backup first
6. Fonts	Fewer families and weights, lighter loading	Small to moderate	A design tidy-up
7. CDN	Serves files from locations near each visitor	Meaningful for spread-out audiences	Setup plus DNS change
8. Updates & PHP version	Current WordPress, theme, plugins and PHP	Steady background gains, plus security	Routine maintenance
9. Lazy-loading	Loads below-the-fold media only when needed	Mostly built into WordPress now	Usually already on

Frequently Asked Questions

What is the fastest way to speed up a WordPress site?

Fix images first — resize to display size, compress at quality 80–85 and use WebP — then enable page caching. For most small-business sites those two changes deliver the majority of the available improvement.

How fast should my WordPress site load?

Aim for pages that feel instant on a mid-range phone: key content visible within about 2.5 seconds on a mobile connection (the Largest Contentful Paint benchmark Google uses). Test the mobile result, not desktop.

Will a caching plugin break my online store?

Configured properly, no — reputable caching tools exclude cart, checkout and account pages. Enable it, then place a test order logged-out to prove the store flow still works before you consider the job done.

How many plugins is too many?

There is no magic number — twenty light, well-built plugins can run faster than five heavy ones. The real rule: every plugin must earn its place, unused ones get deleted (not just deactivated), and no two do the same job.

My speed score is 90+ on desktop but poor on mobile. Why?

Mobile tests simulate slower processors and connections, so heavy images and JavaScript hurt far more. Your visitors mostly arrive on phones, so the mobile number is the one to fix — start with images and plugin weight.

How often should I re-check my site's speed?

Quarterly, and after any big change — new theme, new plugin batch, a marketing tool added sitewide. Speed decays by accumulation, not by accident: each small addition is justified, and the total quietly is not. A calendar reminder and the before/after habit from Step 0 keep the decay visible.

Should I pay for a premium performance plugin?

Sometimes — a good premium tool can bundle caching, image optimisation and file cleanup with safer defaults than three free plugins arguing. But pay for it after the free wins: images resized, unused plugins deleted, hosting assessed. A premium plugin layered over a 6 MB homepage optimises the problem instead of removing it.