Every Twin-Ex discovery call with an Israeli SMB starts the same way: "we have a WordPress site, it is slow, the SEO is mediocre, and our developer disappeared two years ago." WordPress is not the villain in that story — it is the inertia. The villain is the plugin sprawl, the unmaintained themes, and the hosting bill that grows with every visitor. The question is when to stop fixing and when to rebuild.
What WordPress is still good for.
WordPress is excellent when content updates are the daily activity. A magazine, a newsroom, a recipe blog, a wedding photographer with 200 galleries — WordPress lets a non-technical owner publish without thinking about deploys, builds, or markdown. The Block Editor is mature. The plugin ecosystem covers payments, bookings, and translations. If your team writes 5+ posts a month and the spend is under ₪400 a month, do not switch.
Where it breaks down.
WordPress breaks down on three things at once: Core Web Vitals, security surface, and total cost of ownership. The average WordPress site scores 40-60 on Lighthouse mobile because the rendering path is server-side PHP with 8 active plugins each adding 200ms. Security: the platform is the most-attacked CMS on the internet — the patch cadence is monthly minimum, and the attack surface is every plugin you installed. Cost: managed hosting that performs respectably (Kinsta, WP Engine) starts at $35/month — and goes up linearly with traffic. By month 18 you have paid more in hosting than a Next.js rebuild would have cost.
Where Next.js wins.
Next.js wins when the site is a product surface for the business, not a content firehose. A marketing site, a service catalog, a pricing page, a booking funnel, a SaaS landing page — these benefit from Static Site Generation (SSG) + Incremental Static Regeneration (ISR). The pages render at the edge in under 100ms. SEO is solved by the framework (canonical URLs, structured data, sitemap, RSS, hreflang). Hosting at Vercel is free up to 100GB/month bandwidth and $20/month after — significantly cheaper than managed WordPress at the same traffic.
The hybrid that nobody talks about.
You can have both. Run your marketing site on Next.js for speed, SEO, and a low hosting bill. Run a headless WordPress as the CMS behind the blog, exposing posts via the REST or GraphQL API. The editorial team keeps the Block Editor they know. The marketing site stays at 95+ Lighthouse. This is what we do for clients who genuinely publish weekly content but also need a fast funnel — and it costs the same as just running Next.js, because the headless WordPress instance can run on a $5/month Hetzner box.
Practical rebuild signals.
Three signals tell you it is time to leave WordPress: (1) your mobile Lighthouse score is below 50 and a developer has tried twice to fix it; (2) you have been hacked once or you got the "your version is outdated" email from your host more than three times this year; (3) your monthly hosting + plugin license bill is over ₪400 and the site is still not fast. Any one of those, on a marketing-first site, justifies the rebuild. Two or more, definitely.
What a Next.js rebuild actually looks like.
For a typical Israeli SMB with a 12-page marketing site, the rebuild is 3-4 weeks. Week 1: discovery, sitemap, content audit, picking what stays + what gets dropped. Weeks 2-3: design + build, Hebrew + English from day one. Week 4: SEO setup, GA4, sitemap, redirects from the old WordPress URLs to preserve the rankings you already have. We charge ₪5,300 for a Tier-1 site at this scope, including the redirects map (the part most agencies skip). After the rebuild the hosting bill is $0-$20 a month and the site is fast on every device.
WordPress in 2026 is a content tool, not a business stack. If your site is your product, build it on the product stack.