Rebuild when the platform is the problem, not when you are bored of the colours. Four second page loads, plugin conflicts, an Elementor build only one freelancer understands, a mobile PageSpeed score in the forties. Those are platform problems. A tired design is a redesign, and you can do that on the stack you already have.
Should you rebuild
The honest test is whether the current stack is costing you money or time you cannot get back. Slow pages lose search visibility and conversions. A fragile build means every small change is a risk and a bill. If neither is true, a redesign inside WordPress is cheaper and lower risk than a full rebuild.
WordPress hosting, premium plugins and maintenance for a serious business site run roughly USD 3,000 to 12,000 a year. A Next.js or Astro site on a modern host costs USD 0 to 240 a year with no plugin overhead. And moving mobile PageSpeed from around 45 to around 95 typically tracks a 15 to 40 percent lift in organic search visibility over three to six months. On a site that earns money, the rebuild often pays for itself inside a year.
What a rebuild costs in 2026
| Scope | Cost | Timeline |
|---|---|---|
| Simple site, brochure or blog | from around USD 2,000 | 2 to 6 weeks |
| Content site with custom layouts | USD 10,000 to 25,000 | 6 to 12 weeks |
| WooCommerce or complex functionality | USD 10,000 to 50,000 and up | 6 weeks to 6 months |
| Do it yourself, your own time | hosting only, USD 0 to 240 a year | weeks of evenings |
The single biggest cost driver is template count, not page count. A fifty page site with twenty two unique Elementor layouts is twenty two React components to design, build and test, not fifty pages of content to paste. Before anyone quotes you, audit your layouts and count the genuinely distinct ones.
The seven step framework
- 1Inventory. Every URL, every unique layout, every form and every third party integration. This list is the real scope, and it is almost always bigger than the page count suggests.
- 2Decide the stack. Next.js or Astro for the front end. Keep WordPress running headless as the editor if the team writes in it happily, or move to a headless CMS like Sanity, Contentful or Strapi. Do not rebuild the CMS if nobody is complaining about it.
- 3Rebuild layouts as components. One component per unique template, not per page. Content flows in from the CMS.
- 4Migrate content. Export, transform, import. Move images to a CDN or the new host's asset pipeline.
- 5Map redirects. Every old URL to its new equivalent, as a 301. This is the step where search rankings are kept or lost, so give it real time.
- 6Cut over. Deploy, switch DNS, and keep the old site reachable on a subdomain for a week as a reference while you check the new one.
- 7Monitor. Watch Search Console coverage, Core Web Vitals and 404s. Expect to be fixing redirects for a month after launch.
Keep your SEO through the move
- Redirect every old URL with a 301, including trailing slash and letter case variants, to its exact new equivalent.
- Keep titles, meta descriptions and canonical tags identical, unless you are deliberately improving a specific page.
- Own the sitemap. Generate it from the new site and submit it in Search Console on cutover day.
- Preserve structured data. Article, Product, Breadcrumb and Organization markup should carry over.
- Expect a small ranking dip for one to three weeks while Google recrawls. A clean migration recovers within weeks.
- If you can avoid it, do not change the URL structure and the platform in the same release. Do one, let it settle, then do the other.
Do it yourself or delegate
The grid is simple. On one axis, how much the site earns. On the other, how badly a botched redirect map or a week of downtime would hurt. Low earnings and low risk, do it yourself and pocket the hosting savings. Real revenue and real risk, pay someone who can show you SEO safe migrations they have done before, with the redirect maps and the before and after Search Console data to prove it.
Ask to see a redirect map from a past project, and the Search Console coverage graph for the eight weeks after that cutover. Anyone who has done this well will have both. Anyone who hesitates has not.
Case study, rebuilding startdigitalnomad.com
This site is a rebuild in progress. Legacy PHP content and an old theme, moved onto Next.js with the App Router, with the historical database kept as the source of truth and articles rendered on the server. The lessons as they land: redirect discipline matters more than any framework choice, keeping the editing experience familiar means nobody has to be retrained, and moving pages off forced dynamic rendering onto incremental static regeneration cut server load and improved time to first byte. This section will grow as the migration continues, with real numbers.
- Next.js or Astro?
- Astro if the site is mostly content and marketing pages. It ships less JavaScript and is simpler to maintain. Next.js if you have app like interactivity, authentication, or a dashboard.
- Keep WordPress or move CMS?
- Keep it headless if the team writes in it happily. You get a fast front end without retraining anyone. Move only if WordPress itself is the source of the pain.
- How long until rankings recover?
- With a clean migration and complete redirects, one to three weeks of wobble, back to baseline within a few weeks, then the speed gains compound over three to six months.
- Cheapest way to rebuild?
- Do it yourself on Astro or Next.js with a static host and headless WordPress. The cash cost approaches zero. The real cost is your evenings.
- Can I rebuild while the site stays live?
- Yes. Build on a staging domain, test the redirects against a copy of the URL list, then cut over in a single DNS change with the old site kept on a subdomain for a week.