Better Move

Better Move is a furniture-moving company in Kuwait. For a local-services business like this, the website has exactly one job: convince a stressed customer mid-move that the company is professional, and get them into a WhatsApp conversation or a phone call before they open a competitor’s site.

The existing site ran on Gatsby with a separate Strapi CMS behind it. It worked, but it was two systems to host, secure, and update, and every content change had to travel through a webhook-triggered rebuild before it appeared. The design had aged as well — and in a market where most customers arrive on a phone, a dated, slow site quietly costs leads.

I redesigned the site and rebuilt it end to end: a modern layout, a faster stack, and a bilingual English/Arabic experience — Arabic being a first-class requirement in Kuwait, not an afterthought.

Goals & Success Criteria

“Done and good” meant four things: a redesign that reads as trustworthy and professional on a phone; a full Arabic experience with proper right-to-left layout, not a machine-translated bolt-on; content edits that go live immediately without a developer or a rebuild pipeline; and one system to run instead of two.

Approach & Key Decisions

Payload over keeping Strapi. Strapi ran as its own server beside the site — separate hosting, separate updates, separate auth. Payload lives inside the Next.js app itself: one codebase, one deployment, and a content schema defined in TypeScript that the frontend shares types with. The trade-off I accepted was a real migration cost — content had to move out of Strapi’s relational structure into Payload collections, and the client had to learn a new admin.

Next.js over Gatsby. Gatsby’s static builds were the root of the publishing lag: every edit queued a rebuild before appearing. Next.js server rendering with targeted revalidation means content changes are visible immediately while high-traffic pages still serve from cache. The trade-off: a server runtime to operate instead of purely static files — acceptable for a site whose content changes weekly, not yearly.

Arabic as a first-class locale. The English and Arabic sites share one component system, with layout built on CSS logical properties so the whole interface mirrors correctly in RTL. Content is localized in the CMS rather than duplicated, so the two languages can’t drift out of sync structurally.

WhatsApp-first conversion. Kuwaiti customers book movers over WhatsApp, so the redesign treats click-to-chat as the primary call to action — persistent on every page — with phone and email as secondary paths, rather than burying contact behind a form nobody fills in.

Challenges

RTL is the detail that separates a real bilingual site from a translated one. Mirroring isn’t just text direction: icons, carousels, spacing, and typography all need auditing in both directions, and Arabic type wants different line heights and font pairing than the Latin design was built around. Building on logical properties from the start kept this a review pass instead of a parallel stylesheet.

The migration had to be invisible to the business. The old site was live and producing leads, so the rebuild ran in parallel — content migrated and verified from Strapi, redesign approved against real content rather than lorem ipsum — and the switch happened at the DNS level once the new stack was proven.

Results & Impact

The redesigned site is live at bettermoveco.com on a single Payload + Next.js codebase. Content edits publish instantly instead of waiting on rebuilds, the client operates one system instead of two, and the bilingual experience serves both the English-speaking expat market and Arabic-speaking customers — each with a WhatsApp conversation one tap away.

What I’d Do Differently

I would lock the Arabic typography earlier. The design was validated in English first and adjusting type scale and line heights for Arabic afterwards forced a second review of nearly every section — designing both locales side by side from the first mockup would have been cheaper.