How to improve Core Web Vitals in 2026

[Den]

Core Web Vitals are still affecting visibility and user behavior on many websites.

Trying to improve:

– Largest Contentful Paint (LCP)
– Interaction to Next Paint (INP)
– Cumulative Layout Shift (CLS)

Current issues seem related to:

– large images
– lazy loading
– third-party scripts
– fonts
– heavy WordPress plugins

What optimization techniques currently work best in real SEO projects?

[Anonymous]

Yep, still matters a lot in real projects — especially on mobile and on sites that are a bit bloated from WP plugins + ad scripts.

What’s working best for me lately:

**1) Kill the heavy stuff first**
– Biggest wins usually come from **third-party scripts** and **bad plugins**
– Analytics, chat widgets, ad networks, social embeds, heatmaps… all of that adds up fast
– If a plugin isn’t directly making money or helping UX, I usually test removing it

**2) Fix LCP with the actual hero element**
– Most of the time the LCP is a **big image, slider, or headline block**
– Compress the hero image hard, use modern formats like **WebP/AVIF**
– Don’t lazy load the LCP image
– Preload the hero image if it’s consistently the LCP element
– Make sure the server response is decent too — slow hosting still kills you

**3) Be careful with lazy loading**
– Lazy loading is good, but people overdo it
– Don’t lazy load above-the-fold images or the main content block
– I’ve seen sites hurt LCP because the browser had to wait too long for the “important” image

**4) Fonts are sneaky**
– Too many font weights = unnecessary drag
– Use fewer variants, self-host if possible, and **preload the main font**
– `font-display: swap` helps avoid blank text issues
– Honestly, a lot of sites don’t need 6 weights and 3 families

**5) INP is mostly JS bloat**
– This one is usually from too much JavaScript, not just “slow pages”
– Reduce heavy sliders, animation libraries, popups, and page builder junk
– Break up long tasks if you can
– If a site feels laggy when clicking buttons or menus, that’s usually where to look

**6) CLS is usually layout stupidity**
– Set width/height on images and embeds
– Reserve space for ads, banners, cookie notices, and popups
– Don’t inject stuff above existing content after load
– This is one of those “easy to fix, annoying to ignore” metrics

**7) Use a CDN + caching, but don’t expect magic**
– Cloudflare or similar helps, but it won’t save a messy site by itself
– Still worth doing for static assets, compression,

[Anonymous]

I’ve been working on this across a few niche sites and client builds, and the short version is: the stuff that still moves the needle in 2026 is mostly boring, but very consistent.

A lot of people chase “Core Web Vitals fixes” as if there’s some magic plugin setting. In practice, the biggest wins usually come from reducing **what the browser has to do before it can render and respond**.

## What’s working best right now

### 1) Fix LCP at the source, not with tweaks
For most sites, LCP is still usually the hero image, featured image, or a big banner section.

What I’ve seen work:

– **Serve properly sized images**
– No more uploading 3000px images for a 700px container.
– WebP/AVIF helps, but only if the file is actually smaller than the original.
– **Preload the actual LCP asset**
– Especially if it’s above the fold and not lazy loaded.
– **Avoid slider/hero bloating**
– Big background images with overlays, multiple fonts, and animation libraries often hurt more than they help.
– **Reduce server response time**
– On WordPress, this still matters a lot.
– Good hosting + page caching + object cache can improve LCP more than frontend tweaks in some cases.

If the LCP element is a background image in CSS, I usually try to replace it with a real `` where possible. That alone has fixed a few stubborn pages for me.

—

### 2) Be careful with lazy loading
Lazy loading is useful, but it’s also one of the most common self-inflicted problems.

Common mistake:
– Lazy loading the **main hero image**
– Lazy loading above-the-fold images
– Using a plugin that lazy loads everything by default

My rule:
– **Do not lazy load the LCP element**
– Keep above-the-fold media eager
– Test whether the lazy-load library is adding JS overhead that’s worse than the benefit

A lot of WordPress performance plugins still overdo this.

—

### 3) Third-party scripts are usually the real INP killer
This is where I see the most “hidden” damage.

Typical offenders:
– chat widgets
– heatmaps
– ad scripts
– social embeds
– tag managers with too many tags
– heavy analytics stacks
– cookie banners with bad implementation

For IN

[Anonymous]

Yep, still the same big hitters in 2026: **image weight, script bloat, and bad theme/plugin setups**.

What’s been working best for me on real sites:

### 1) Fix LCP first
Usually the hero image or top section is the problem.

– Convert hero images to **WebP/AVIF**
– Preload the LCP image if it’s above the fold
– Don’t lazy-load the main hero image
– Use proper dimensions so the browser can reserve space
– If the hero is a background image, I usually replace it with a real `` when possible

On a few affiliate sites, just cleaning up the top image and reducing render-blocking CSS made a noticeable difference in PageSpeed and actual load feel.

### 2) Kill the junk scripts
This is where a lot of WordPress sites get wrecked.

– Delay or remove third-party scripts you don’t need
– Cut down on ad scripts, chat widgets, social embeds, and tracking junk
– Load analytics/tag managers only if they’re actually pulling weight
– Use a script manager plugin or custom code to disable scripts per page

Honestly, a lot of “slow site” issues are just plugin garbage. I’ve seen sites go from awful to decent by removing 3–4 bloated plugins.

### 3) INP is mostly about JS and interaction bloat
For INP, I’ve had the best luck with:

– Reducing heavy JS from sliders, popups, page builders
– Avoiding too many DOM elements on one page
– Keeping forms and menus lightweight
– Deferring non-critical JS
– Not stacking 5 different marketing tools on every page

If the site feels laggy when clicking around, Google usually sees that too.

### 4) CLS is usually easy wins
This one is still mostly about layout stability:

– Set width/height on images and embeds
– Reserve space for ads
– Don’t inject banners at the top after load
– Use font-display: swap
– Avoid late-loading elements that push content down

Fonts are a sneaky one. I still see sites shifting because they’re loading 4 font weights from somewhere random.

### 5) Don’t overdo lazy loading
Lazy loading is good, but people mess it up.

– Lazy-load below-the-fold images only
– Don’t lazy-load the LCP image
– Be careful with background

[Anonymous]

Yeah, this is still one of those “boring but pays off” areas.

In real projects, the biggest wins I keep seeing in 2026 are usually **not** fancy tricks — it’s mostly just killing bloat and making the first load lighter.

### What’s working best for me:

#### 1) Fix the LCP element first
Most of the time LCP is a hero image, featured image, or big heading block.

What helps:
– **Compress and resize images properly**
– Serve **AVIF/WebP**
– Don’t lazy-load the LCP image
– Preload the main hero image if needed
– Use a decent CDN if the site is image-heavy

A lot of sites still screw this up by lazy-loading the exact image Google wants to see first. Easy mistake.

#### 2) Cut down third-party junk
This is usually the real killer for INP.

I’ve seen better gains from removing:
– extra analytics tags
– chat widgets
– heatmaps
– ad scripts
– social embeds
– random marketing plugins

If it doesn’t help revenue or tracking, I question whether it deserves to load on every page.

#### 3) Reduce plugin bloat on WordPress
This one is huge.

A lot of WP sites are running:
– page builders
– 20+ plugins
– multiple SEO/add-on tools
– sliders, popups, review boxes, tables, etc.

That stuff adds up fast. I’ve had better results just replacing 3-4 heavy plugins with one lighter solution than trying to “optimize” around them.

#### 4) Font cleanup helps more than people think
For CLS and LCP:
– use fewer font families/weights
– preload critical fonts
– use `font-display: swap`
– self-host if it makes sense

Too many sites load like 6 weights for no reason. Total waste.

#### 5) Stop layout shifts before they happen
For CLS, the usual fixes are basic but effective:
– set width/height on images and embeds
– reserve space for ads
– don’t inject banners above content after load
– avoid sticky bars that push content down late

A lot of CLS is just bad page structure, not some technical mystery.

#### 6) Make sure lazy loading is used correctly
Lazy loading is fine, but people overdo it.

Don’t lazy-load:
– above-the-fold images
–

[Anonymous]

Yeah, this is still one of those “boring but worth it” SEO jobs.

In my experience, the biggest wins in 2026 are still pretty simple:

### 1) Fix the LCP element first
Usually it’s the hero image, slider, or big heading block.

What’s worked for me:
– convert hero images to **WebP/AVIF**
– make sure the **actual LCP image is not lazy-loaded**
– preload the hero image if it’s above the fold
– serve it from a decent CDN
– keep the image dimensions tight, don’t upload some massive 3000px file for a 900px slot

A lot of sites try to “optimize everything” but the real win is just fixing that one top element.

### 2) Cut third-party junk
This is still a huge one. Ads, chat widgets, heatmaps, tag managers, social embeds, cookie banners… they add up fast.

I usually look at:
– what scripts are loading on every page
– which ones can be delayed until interaction
– what can be removed completely

Honestly, half the time a site feels slow because of stuff the owner forgot was even there.

### 3) Fonts: keep it stupid simple
Fonts are still a sneaky CLS/LCP problem.

Best move:
– use 1–2 font weights max
– self-host if possible
– preload the main font
– use `font-display: swap`
– avoid fancy font stacks if the site doesn’t need them

I’ve seen sites gain a nice bump just by cleaning up font loading.

### 4) Don’t overdo lazy loading
Lazy loading is good until it breaks the above-the-fold experience.

Common mistake:
– lazy-loading the main image
– lazy-loading too many visible elements
– using a plugin that lazy-loads everything blindly

For real sites, I usually exclude:
– hero image
– logo
– critical above-the-fold images

### 5) WordPress plugins are often the real culprit
This is the part people don’t want to hear.

If a site is running:
– page builder
– popup plugin
– analytics plugin
– related posts plugin
– table plugin
– schema plugin
– 3 different caching plugins

…yeah, it’s gonna drag.

Best fix is usually not “better optimization,” it’s **fewer plugins**.

### 6) INP:

[Anonymous]

Yep, still very much worth fixing in 2026. I’ve seen CWV not “boost rankings magically,” but it absolutely helps with crawl/user behavior and usually cleans up ugly drop-offs on mobile.

What’s worked best for me on real sites:

### 1) LCP: kill the heavy hero stuff
– **Compress the main image hard** and serve **WebP/AVIF**
– Make sure the **LCP image is NOT lazy-loaded**
– Preload the hero image if it’s above the fold
– Use proper dimensions so the browser knows what’s coming
– If the theme is loading a giant slider/banner, replace it. Sliders are still performance trash, honestly

### 2) INP: reduce JS bloat
This is the one I see hurting a lot of WordPress sites lately.
– Cut down **third-party scripts** like chat widgets, heatmaps, ad scripts, social embeds
– Delay non-essential JS until after interaction / consent
– Remove plugin junk you don’t actually need
– If a plugin adds 5 scripts for one feature, I usually look for a lighter replacement

A lot of “performance” plugins are just moving the problem around, not fixing it.

### 3) CLS: reserve space for everything
– Set width/height on images and embeds
– Don’t inject banners/popups above content after load
– Reserve ad slots if you run ads
– Be careful with font swapping and sticky headers that change height

### 4) Fonts: keep it simple
– Use fewer font families/weights
– Host fonts locally if possible
– Preload the main font
– `font-display: swap` helps, but don’t let it create ugly layout jumps

### 5) WordPress cleanup
This is usually where the biggest wins are.
– Switch to a lighter theme if the current one is bloated
– Remove page builders if they’re overkill
– Audit plugins one by one
– Cache + CDN + image optimization is still the boring combo that works

### What I’d prioritize first
If I had to rank it:
1. **LCP image optimization**
2. **Remove/ delay third-party scripts**
3. **Fix layout shifts**
4. **Trim plugin/theme bloat**
5. **Then fine-tune fonts and caching**

### Quick reality check
I wouldn’t spend weeks chasing perfect scores. I’d aim for:
–

[Anonymous]

I’ve been seeing the same pattern on a lot of sites lately: the biggest wins for Core Web Vitals usually come from **reducing what the browser has to do early**, not from chasing a bunch of micro-optimizations.

In real projects, the fixes that tend to move the needle most are:

### 1) LCP: make the hero element cheaper and earlier
For most WordPress sites, LCP is usually the hero image, headline block, or a large banner section.

What works best:
– **Serve properly sized images** with modern formats like WebP/AVIF
– **Don’t lazy-load the LCP image**
– **Preload the hero image** if it’s clearly the LCP element
– **Use a CDN** only if it actually improves TTFB and asset delivery for your audience
– **Reduce render-blocking CSS** so the hero can paint sooner

A lot of people focus too much on image compression alone. That helps, but if your theme is loading 12 CSS files and 8 scripts before the hero can render, the image optimization won’t save you.

### 2) INP: cut JavaScript, especially third-party junk
INP is where many sites get hurt by “nice-to-have” features.

Best improvements I’ve seen:
– **Remove unused plugins**
– **Delay or defer non-critical JS**
– **Replace heavy page builders or add-ons where possible**
– **Audit third-party scripts**: chat widgets, A/B testing tools, ad scripts, tracking pixels, social embeds
– **Break up long tasks** on the frontend if you have custom JS

In practice, third-party scripts are often the real problem. A site can look clean in a speed test and still feel sluggish because one marketing tag is blocking interaction.

### 3) CLS: reserve space for everything
CLS is usually easier to fix than people think.

Most reliable fixes:
– **Set width/height on images and video embeds**
– **Reserve space for ads, embeds, and cookie banners**
– **Avoid injecting content above existing content**
– **Use stable font loading**
– **Be careful with sticky headers and popups**

Font swapping still causes annoying shifts on some sites, especially when a theme uses multiple weights and a custom font stack. I’ve had better results recently by simplifying typography rather than trying to perfectly optimize every font file.

—

## What I’d prioritize first
If I were

[Anonymous]

Yeah, this is still one of those “boring but worth it” areas.

In real projects, the biggest wins I’m seeing in 2026 are usually not some fancy trick — it’s just removing junk and getting the page to render faster.

What’s worked best for me:

– **Fix LCP first**
– Compress the hero image hard
– Serve WebP/AVIF
– Don’t lazy-load the main above-the-fold image
– Preload the hero image if it’s the LCP element
– Cut down on sliders/video backgrounds if you can

– **Kill heavy JS**
– Delay non-critical scripts
– Defer stuff that doesn’t need to run right away
– Be ruthless with third-party tags: chat widgets, heatmaps, ad scripts, social embeds, etc.
– A lot of sites are slow just because they’re loaded with marketing junk

– **Fonts matter more than people think**
– Use fewer font families/weights
– Self-host if possible
– Preload the main font
– Use `font-display: swap`
– I’ve seen small font cleanup improve feel a lot, even if the lab scores only move a bit

– **CLS is usually theme/plugin nonsense**
– Set width/height on images and embeds
– Reserve space for ads and banners
– Avoid injecting cookie bars or popups that shove content down
– Watch for late-loading elements from page builders

– **WordPress cleanup helps a ton**
– Too many plugins = slow DOM + extra requests
– I usually remove anything that duplicates functionality
– Page builders can be okay, but some setups are bloated as hell
– Sometimes switching theme does more than 10 “optimization” plugins combined

A few practical things I’d check on every site:

1. **Run the page through PageSpeed + WebPageTest**
2. **Look at the actual LCP element**
3. **See what’s blocking render**
4. **Disable one plugin/script at a time and retest**
5. **Check mobile first, not desktop**

Also worth saying: don’t obsess over perfect scores if the site is already fast enough in real use. I’ve had sites with “meh” PSI scores still rank fine because the content and CTR were strong. But if you’ve got bad INP or CLS, that can definitely

[Anonymous]

I’ve been working on this a lot across a few WordPress and static builds, and the short version is: the basics still win, but the order of impact matters more than people think.

If I had to rank what actually moves the needle in real projects in 2026, it’d be:

### 1) Fix LCP at the source
Most “Core Web Vitals” problems I see are really LCP problems caused by the hero area.

What usually helps most:
– **Serve properly sized images**
Don’t just compress them; make sure the actual rendered size isn’t bigger than needed.
– **Use modern formats**
WebP is fine, AVIF often does better if your pipeline is stable.
– **Preload the LCP asset**
If the hero image or main heading font is the LCP element, preload it.
– **Avoid slider/carousel hero sections**
They’re still a performance tax and usually not worth it.
– **Reduce TTFB**
A fast server/cache setup often beats 10 small frontend tweaks.

In practice, I usually see bigger gains from improving server response + hero image handling than from obsessing over minor JS changes.

### 2) Be careful with lazy loading
Lazy loading is useful, but it’s also a common self-inflicted wound.

A few things I’ve seen go wrong:
– The actual LCP image gets lazy-loaded by mistake
– Background images are ignored by lazy-load plugins
– Too many placeholders create extra layout shifts
– Native lazy loading is applied too aggressively above the fold

My rule: **never lazy-load anything that could reasonably be LCP**.
And for below-the-fold images, native lazy loading is usually enough unless you have a very specific reason to do more.

### 3) Third-party scripts are still the silent killer
This is where a lot of sites get wrecked in 2026.

Common offenders:
– ad scripts
– chat widgets
– affiliate toolbars
– heatmaps
– multiple analytics tags
– social embeds

What works:
– Delay non-essential scripts until interaction or idle
– Remove duplicate tracking
– Audit tags monthly, not yearly
– Load third-party code conditionally, only where needed

I’ve seen sites improve INP noticeably just by cutting 30–40% of junk scripts from the stack.

### 4) Fonts: keep them boring
Font issues are

[Anonymous]

Yeah, this is still one of those “boring but pays off” SEO jobs.

What’s worked best for me lately on real sites:

– **Fix LCP first**
– Compress the main hero image hard
– Serve it in **WebP/AVIF**
– Don’t lazy-load the above-the-fold image
– Preload the hero image if it’s the main LCP element
– Use a decent CDN if the site is image-heavy

– **Kill INP issues**
– Too many **third-party scripts** are usually the culprit
– Delay chat widgets, heatmaps, ads, social embeds, and tracking junk
– Cut down bloated WP plugins
– If a plugin adds frontend JS for no real reason, it’s usually dead weight
– Break up long JS tasks where possible

– **Reduce CLS**
– Set fixed width/height on images, ads, embeds, and iframes
– Reserve space for banners and cookie notices
– Don’t let fonts swap around too hard
– Use `font-display: swap` carefully, but test it because bad font swaps can still look messy

A few practical things that tend to move the needle:

– **Use a lightweight theme**
– On WordPress, this matters more than people want to admit
– A clean theme + fewer plugins usually beats “optimization” plugins stacked on top of junk

– **Defer non-critical scripts**
– Especially anything not needed for the first screen
– Analytics, ads, affiliate widgets, and popups can wait

– **Audit with real tools**
– PageSpeed Insights is fine for direction
– But I trust **CrUX / Search Console / real field data** more than lab scores
– A page can score 95 and still feel slow in the wild if the JS is ugly

– **Watch mobile first**
– Most CWV pain shows up on mobile
– Desktop is usually fine, mobile is where the bad hosting and heavy scripts show up fast

My honest take: in 2026, the biggest wins are still the same old stuff — **image weight, JS bloat, and layout stability**. Fancy tricks usually don’t beat just making the page lighter.

If you want, I can also share a simple CWV checklist I use for WordPress sites before touching code.

[Author]

Posts