WordPress modernisation,
without the SEO hit.
Migrate from slow, fragile WordPress to a fast, secure Cloudflare-edge stack — with the redirects, content, and search rankings intact.
The WordPress debt trap.
We’ve watched this get worse, not better. Here’s the data from 2024.
Most Australian businesses running WordPress aren’t running it so much as managing it. Weekly plugin updates, intermittent slowdowns, the creeping anxiety that something will break on a Friday afternoon — these are the background noise of a platform that’s showing its age.
The performance ceiling is structural, not configurable away. HTTP Archive data puts WordPress at a 45% mobile Core Web Vitals pass rate — the lowest of any major CMS. Server-rendered PHP, database round-trips per page load, and 20–30 plugin stacks compound over time. Wordfence blocked over 9 billion cross-site scripting attempts against WordPress sites in 2024 alone.
Then there’s the hidden cost: WordPress maintenance in Australia runs $90–$500/month for a basic site, before plugin licence renewals ($600–$1,500/year is typical), developer call-out fees when something breaks, and the hosting bill on top. Many businesses have no idea they’re paying this much.
What we migrate to.
Every component chosen for longevity, performance, and low ongoing cost. No exotic choices — these are the most well-resourced tools in the category.
Cloudflare Pages + Workers
Static assets served from 300+ global edge locations. Workers for dynamic logic — forms, redirects, API proxying — with no cold starts and no per-GB egress fees. Typical cost: $0–$60/year for an SME site.
Astro (default) or Next.js
Astro ships zero JavaScript by default and scores 95–100 on Lighthouse for content-heavy pages without tuning. Next.js when the project genuinely needs SSR or authenticated routes. Astro 6 has first-class Cloudflare bindings following the January 2026 acquisition.
Cloudflare acquired Astro, Jan 2026 — deep platform integration
Payload, Sanity, or Decap
Chosen per client. Sanity for marketing sites and collaborative teams. Payload CMS for developer-managed builds with tight type safety and self-hosting — acquired by Figma in June 2025, remains fully open source. Decap for git-based, developer-managed blogs.
Figma acquired Payload, June 2025 — well-resourced, open source
Pagefind or Vectorize
Pagefind is MIT-licensed and free — indexes from rendered HTML at build time and ships as static files. Around 50KB entry script. Cloudflare Vectorize for AI-driven semantic search: let visitors ask your site a question rather than keyword-match against it.
Workers + Resend
Contact forms routed through a Cloudflare Worker, with transactional email via Resend. Form submissions stored in D1 if needed. No PHP, no WP plugins, no contact form spam surface.
D1, KV, R2
Cloudflare D1 (SQLite-compatible serverless database) for structured data, KV for sessions and config, R2 for media storage. All accessible from Workers with zero cold starts and no public-facing database endpoint.
Our three tiers.
Fixed-price, scoped before we start. The Assessment tier also serves WordPress modernisation — it’s the right place to begin if you’re unsure about scope or cost.
Modernisation Assessment
What you get
- Security audit: plugin inventory, CVE exposure, TTFB baseline
- Content inventory: post types, volume, custom fields, media
- Architecture recommendation: CMS, framework, hosting, search
- SEO baseline: rankings, traffic, meta coverage, redirect risk
- Migration scope estimate with itemised complexity factors
- Fixed-price quote for Standard or Complex migration
No build commitment. Output is yours regardless of what you decide.
Standard Migration
What you get
- New site on agreed stack (Astro + Sanity/Payload, Cloudflare Pages)
- Content migration with AI-assisted schema generation
- Full redirect map — 301 rules deployed at Cloudflare edge
- SEO meta migration: titles, descriptions, OG tags, structured data
- Contact form and email flow (Workers + Resend)
- Target: 90+ Lighthouse, <400ms TTFB on Cloudflare PoPs
- Editorial training (1-hour session for content team)
- Handover documentation + 30-day hypercare support
For brochure sites, marketing sites, and content-focused blogs up to approximately 50 pages.
Complex Migration
What you get
- All Standard Migration deliverables
- Custom content modelling for complex post types and taxonomies
- WooCommerce migration (scoped per project)
- Multilingual content with AI-assisted translation review
- Third-party API integrations: CRM, booking, payments
- Custom Workers replacing plugin functionality
- Full QA cycle with staging environment
- Extended editorial training + 60-day hypercare
For mid-market businesses, e-commerce, multilingual sites, or complex plugin dependencies.
How we protect your rankings during cutover.
Industry data shows 30–60% organic traffic drops immediately post-migration, with an average recovery period of 523 days. Seventeen per cent of migrations never recover to pre-migration traffic levels. We engineer for the opposite — but we tell you this upfront because the risk is real and deserves to be scoped, not footnoted.
The AI-augmented migration method.
AI compresses the expensive, manual phases of a migration. What would take weeks of developer time — content extraction, schema mapping, redirect inventory — can be accelerated without cutting corners on architecture or testing. Our engineers own every decision; AI is the tool.
Content extraction and schema inference
WordPress XML export and REST API output is analysed to propose a structured CMS schema — post types, custom fields, taxonomies — in Sanity, Payload, or Decap format. A developer reviews and refines; AI handles the volume work.
Image optimisation and alt text
Automated WebP/AVIF conversion for the media library. AI-generated alt text for images missing accessibility attributes — improving both accessibility and image search coverage without a manual pass over hundreds of files.
Redirect mapping at scale
WordPress URL structures — date-based paths, category prefixes, custom slugs — need 1:1 301 mapping to the new structure. AI handles edge cases: multiple old URLs consolidating to one page, URL normalisation, trailing slash rules. Output is a deployable Cloudflare redirect config.
Content modernisation pass
Optional but often valuable: an AI pass over content during migration to update stale statistics, tighten prose, add internal links, and ensure Australian English consistency. Adds scope; improves what lands on the new platform. Quoted separately.
AI accelerates the slow parts of a migration. It does not make content model design decisions or replace the developer judgment calls about what to keep versus cut. Those are ours.
When we’ll tell you not to migrate.
A migration that doesn’t make sense for your situation is a migration we’ll say no to. Here are the clearest cases.
-
Deep Elementor or Divi dependency. Elementor stores content as serialised JSON interleaved with layout data — there is no automated path to separate them cleanly. If the entire site was built in Elementor and the team thinks in widgets, migration is labour-intensive and the training overhead is real. We’ll scope it honestly.
-
Very small sites under five pages. If the site is a single-page brochure or a site with minimal content and low traffic, the migration overhead doesn’t pay off. We’ll say so.
-
Bespoke WooCommerce extensions where rebuild cost exceeds value. Complex e-commerce — multiple payment gateways, subscription logic, custom product variations, fulfilment integrations — can cost $40,000–$80,000+ to replicate headless. Unless there’s a clear revenue and security justification, the ROI case often doesn’t hold. Shopify or BigCommerce migration is sometimes the better call.
-
Large non-technical editorial teams without a training budget. The productivity dip post-migration is predictable. If the organisation can’t support a training and handover period as a formal deliverable, the migration will frustrate the people it’s supposed to help.
If the answer is no after the Assessment, you still get the full scope document and security audit — useful regardless of what you decide to do next.
Outcomes you should expect.
These figures reflect real-world evidence from documented migrations. “Properly implemented” is load-bearing — headless is not automatically faster; it requires deliberate choices.
Common questions.
-
Not necessarily identical — and usually that’s a good thing. Migrations are often an opportunity to refresh the design and tighten the content structure. If you want the same look reproduced faithfully, we can do that too. The Assessment tier will document the current design and scope what fidelity level is right for your situation.
-
Editors move from WordPress (or Elementor) to Sanity, Payload, or Decap — depending on which CMS suits the project. Sanity in particular has a polished editorial interface with real-time collaboration and structured content models that prevent editors from publishing malformed content. There is a learning curve; we scope editorial training as a formal deliverable, not an afterthought. Most editors find the new systems easier to use after the first week.
-
For Standard migrations, DNS cutover takes 15–30 minutes. We prepare staging ahead of time and only cut DNS when the new site is fully validated. For larger sites we stage the cutover over a low-traffic period. Redirects and robots configuration are tested on staging, not live. Your old WordPress install remains available as a backup for at least 30 days post-cutover.
-
Plugin functionality needs a replacement decision for each item: a third-party SaaS equivalent, a custom Worker build, or an embedded hosted service (Cal.com for booking, Stripe for payments, etc.). The Assessment tier maps every plugin and scopes the replacement before you commit to a build. Some plugin features are trivial to replace; others need scoping honestly. We’ll tell you which is which.
-
Yes, but they sit firmly in the Complex Migration tier and are scoped case by case. Depending on the complexity, we may recommend migrating to Shopify or BigCommerce headless rather than a full custom rebuild. Standard WooCommerce (products, cart, basic checkout) is manageable. Multi-gateway, subscription, or heavily customised checkout flows need honest scoping. The Assessment will tell you which category you’re in before you commit.
-
Yes. A common pattern is to migrate the public-facing site first (the part that’s slow and vulnerable) while keeping WordPress as the CMS backend via headless API during the transition. This reduces risk and lets the editorial team adapt gradually. Full decoupling then happens in a subsequent phase once the team is comfortable. The Assessment will recommend whether a phased approach makes sense for your situation.
Start with an assessment.
A fixed-price, 1–2 week engagement that scopes your migration, assesses the SEO risk, and gives you a clear quote before anything is built. No obligation to proceed.