Why we killed Elementor and rebuilt everything in Astro

Seven reasons we don’t use Elementor anymore: it’s slow; plugin dependency hell; updates break things; the editor crawls at scale; no real version control; hidden licence and hosting costs; total vendor lock-in.

There’s a specific kind of Monday morning that finally broke us. A client rang at 9:14am. Their site was down. Nothing had changed on our end — but Elementor had pushed an update overnight, one of the 14 plugins they relied on had a conflict, and the homepage was now a half-rendered grey rectangle.

Forty minutes of plugin disabling, version rollbacks and frantic Slack messages later, the site was back. We billed nothing for it because, frankly, it shouldn’t have happened in the first place.

That was the morning we decided: no more Elementor. No more WordPress. We were rebuilding our entire workflow around Astro, hand-crafted code, and AI as a coding partner. Six months later, we don’t recognise our old life.

Here’s why we did it — and why, if you’re running an Elementor site, you should at least be thinking about it.

The case against Elementor

Let’s just list it. No marketing softeners.

1. It’s slow. Genuinely slow.

Elementor pages ship megabytes of CSS, JavaScript, and inline styles to render a section that should be 20 lines of HTML. Core Web Vitals on a typical Elementor site sit somewhere between “yellow” and “red.” Google notices, and so do your visitors. If you’re investing in SEO, you’re fighting the tool before you write a word of copy.

2. The plugin tower of cards.

Elementor on its own doesn’t do enough, so you install Essential Addons. Then Ultimate Addons. Then Crocoblock for dynamic content. Then a form plugin. Then a caching plugin to undo the damage. The average Elementor site we audit has 22 plugins. Every one is an update, a security risk, and a future conflict.

3. Updates break things constantly.

You don’t update. You can’t update. Because every time you do, the carousel breaks, the spacing shifts, or the contact form silently stops sending emails. So you sit on outdated versions, which then become security risks. Lose-lose.

4. The editor is painful at scale.

Anyone who’s built a 30-page Elementor site knows the moment the editor starts taking 12 seconds to save. Move a column, wait. Add a heading, wait. Now do that 200 times to finish the homepage.

5. There is no real version control.

Something broke last Tuesday? Good luck. There’s no git log to scan, no diff to read. The revision system inside WordPress is patchy and often doesn’t capture template-level changes. Real engineering teams don’t work this way for a reason.

6. It’s expensive in places you don’t notice.

Elementor Pro, addon licences, a beefier host (because shared hosting can’t cope), Cloudflare Pro to mask the bloat, premium caching plugins. Each one is a small bill. Together they’re hundreds of pounds a year, per site.

7. You don’t own anything.

Try exporting an Elementor site to anything else. You can’t. The whole site is locked inside shortcodes and JSON blobs in the database. The day you want to leave is the day you discover you can’t.

What we moved to

Astro. Plus AI as a coding partner. That’s the whole stack.

Astro ships zero JavaScript by default. It outputs clean, static HTML — the same kind of HTML the web was built on — and only loads JS when a component actually needs it. The result is a site that scores 100/100 on Lighthouse without us trying.

We build components in real code: HTML, CSS, a touch of TypeScript when needed. AI handles the heavy lifting on the repetitive bits — scaffolding a new section, writing the Tailwind classes, generating content variations — but everything is real, readable code we can version, audit and fix.

There is no database. There are no plugins. There is no admin panel that gets hacked at 3am. The site lives in a Git repository, gets deployed to Cloudflare or Vercel in seconds, and runs on hosting that’s often free.

For the longer technical comparison — including where we still use WordPress — read Elementor to Astro: why we changed our web stack.

What actually changed

Site speed. Pages that took 4–6 seconds to load on Elementor now load in under a second. Every site, every time.

Bug count. We used to keep a Notion database tracking recurring “WordPress fires.” Since switching, that database has had two entries — both content typos, both fixed in a single commit.

Build time. A typical small business website design project that took us three weeks in Elementor (most of it fighting the tool) now takes us four to five days. AI does the typing. We do the thinking.

Client maintenance. Down by roughly 90%. No more “Elementor broke my site” emails. No more updating 22 plugins a week. The site we shipped is the site that’s still running, byte-for-byte. Ongoing work lives in care plans — content and analytics, not emergency plugin surgery.

Hosting cost. From £25–£40/month per site to effectively zero on Cloudflare Pages. That saving gets passed on to clients.

Security. No PHP, no database, no login page to brute-force. The attack surface is almost nothing.

Should you move?

If you’re a business owner reading this and your site is on Elementor, ask yourself three questions:

1. Has it gone down in the last 12 months because of an update?
2. Does it score below 70 on Google PageSpeed Insights for mobile?
3. Are you paying for plugins and hosting that feel like they keep creeping up?

Two yeses and it’s probably time. The migration isn’t as scary as it sounds — we can usually rebuild a 10-page Elementor site in under a week, and the difference is something you feel the moment it goes live.

We’re not anti-WordPress. We’re anti-friction. Astro removed the friction, and we’re not going back.