©Code byMilanParmar

Should you keep WordPress, or rebuild a custom website?

I get a lot of rebuild quotes because the site is WordPress and slow. Half the time I cut plugins. The other half, the theme is the whole site and custom is cheaper than fighting it.

Updated 5 min read

A metal door hinge lying on a desk

Someone sent a rebuild quote because the site is WordPress and slow. Should you keep WordPress, or rebuild a custom website? I get this brief a lot. Half the time I cut plugins and compress images. The other half, the theme is the whole site, and custom is cheaper than another year of fighting it.

This is website work, usually a takeover, not a blank-page rebuild so the stack looks current. If the number is the problem, read how much a custom website costs. A rebuild quote that's still a theme is the cheap template. If the wait is the problem, start with a slow website is a product problem.

What most WordPress rebuild briefs actually mean

They mean start over. New stack, new sitemap, a month billed as discovery. The live site still has to take the call on Monday. I write a brief before I quote so the rebuild is a named job, or I cut it.

  1. 01

    It's slow, so it must be WordPress

    Sometimes the theme is packed with plugins and you will never make that fast. Sometimes it's one uncompressed hero image and a cache that was never turned on. I can tell which in an afternoon on the live URL. A rebuild quoted as the only way to a three-second page is how you pay for a new brand you didn't ask for.

  2. 02

    Too many plugins, nobody knows which

    I count the ones that take money, bookings, or search. Those can stay. The rest are leftovers from five years of 'just add this.' If the homepage only exists because a page builder drew it, the plugins aren't the mess. The theme is the site. That's when I talk about a rebuild.

  3. 03

    Nobody can change a line without a ticket

    That's a person problem. I've seen custom sites with the same ticket queue. If changing a heading means emailing a vendor and waiting a week, fix who can deploy. Rebuilding so you still can't ship a one-line fix gets you a second site you also don't own.

  4. 04

    An agency looked for an hour and said start over

    They price a blank page. You already have URLs that rank and copy people read. I don't decide from an hour in wp-admin. I decide when repair means fighting the theme on every page, and the product still has to answer the phone on Monday.

When you should keep WordPress

WordPress is a CMS a lot of teams already know. If the job is publish and get found, and the theme isn't drawing every page by itself, throwing it out dumps a year of content so a developer can own every comma. I won't do that.

  1. 01

    People inside the business write every week

    A newsroom. A services firm where someone actually updates forty pages. A blog that is the marketing. They already know WordPress. If I rebuild it into a custom CMS they won't open, the site goes stale in a month. Keep WordPress. Fix the theme if the wait is the theme.

  2. 02

    The URLs rank and the words are good

    Then I keep the paths. A rebuild that invents cleaner slugs so the IA looks neat is how you pay for the site and then pay again for the traffic you had. Same titles, same URLs. I only replace what the phone downloads.

  3. 03

    The theme is thin, and the job is a brochure

    A stranger can tell what you do and send a message. No login. No page builder sitting on twenty plugins. Speed is a pass I can do in the current install. Spending custom-website money on that job buys you a different site you don't need yet.

When a custom website rebuild is the job

I don't rebuild because WordPress is unfashionable. I rebuild when a page builder is the only way the homepage exists, or when people log in and the CMS is pretending to be a product. A content-heavy consulting siteI took from about three seconds to about 450 milliseconds. Same copy. A frontend that didn't make the phone wait. That's in selected work. I implemented the Figma. I dropped the old theme. The URLs stayed.

  1. 01

    The theme is the product

    Twenty plugins, each one a network request. A page builder owns the homepage. Change one block and three other pages jump. You can't make that fast without lying to Lighthouse. I replace the theme. I keep the words and the URLs if they rank.

  2. 02

    People log in, pay, or run a job through it

    That's a web product sitting on a CMS. WordPress will take the payment plugin. It will not enjoy the Tuesday when auth breaks. Accounts, records, money that has to move: that's custom work, or a takeover of what already runs. Calling it a WordPress rebuild gets you a theme with a login taped on.

  3. 03

    The wait is in the architecture

    Every page ships a builder, a slider library, a font kit. A faster host still sends all of that. Then I rebuild the frontend. Not a new brand. HTML that arrives before the person leaves. Same company, same URLs. I did this on a consulting site: about three seconds down to about 450 milliseconds.

Keep WordPress

People still publish in it

Paths rank. The theme is thin enough that a speed pass can land.

Custom rebuild

The theme is the whole site

Plugins are how pages exist. The phone waits. Repair fights the theme every week.

When rebuilding WordPress is the wrong first move

A rebuild is still a first versionof a new stack, even if the company is ten years old. If you haven't named whether this is a website or a web product, stop. You might be rebuilding the wrong thing.

  1. 01

    An afternoon of cuts would do it

    Images, unused scripts, a cache, plugins that do nothing. If the theme is honest, that's days, not a quarter. I see rebuilds quoted so nobody has to do the boring pass. Do the pass.

  2. 02

    You said website and meant a product

    Login, an admin, customers who come back. A prettier WordPress brochure still isn't that product. Name it. Price that job. Don't spend a custom-website budget to avoid saying SaaS.

  3. 03

    The rebuild wants new URLs for cleanliness

    If the old paths rank, new slugs are vanity. Redirects are worse than never moving. A custom site can keep the same paths. If the quote opens with a new sitemap, ask how much traffic they're willing to bet.

  4. 04

    Nobody will own it after launch

    Custom with no owner is a zip file. WordPress with no owner is a plugin update that sits for a year. Starting over doesn't hire anyone for Tuesday. Pick an owner, or I'll build fewer moving parts on purpose.

If the URLs rank, I keep them. I rebuild when a page builder is the only way the homepage exists.

If you're looking at a rebuild quote, send the live URL. Tell me what you're building, or just send the site. Someone still has to own it after launch on either stack. More of that split is on WordPress or a custom website.

FAQs

Can you keep my WordPress URLs and content in a rebuild?

Yes. If they rank, they stay. I move the copy onto a stack that can be fast. The path in the address bar shouldn't change. A 'cleaner' sitemap is how you give traffic away.

Will I lose Google ranking if I leave WordPress?

Not because you left WordPress. You lose ranking when titles, paths, and the HTML Google already understood disappear. A new look on new URLs is the gamble. The framework isn't.

Can you just make my WordPress site faster instead?

If the theme is thin, often yes. Images, scripts, cache, dead plugins. If the homepage can't render without a page builder, a faster host still sends a late page. I say which after I open the URL, not from a plugin screenshot.

What if I have WooCommerce or a heavy blog?

Staff living in a catalogue, or a team that publishes daily, is a reason to keep a CMS they know. It isn't a reason to keep a three-second theme. Sometimes I custom-build the storefront and leave the writing in WordPress. Those are two jobs. Name both.

How long does a WordPress rebuild take?

A marketing site where I keep the URLs is often weeks after the brief, not a quarter of workshops. Content-heavy takes longer because the layouts are the work. I quote after I see what can stay. Copy and paths first.