WordPress vs custom build — an honest decision matrix

WordPress vs custom build is the most argued-about website platform choice in the local market, and the argument is almost always framed badly. It isn't WordPress versus "better." It's WordPress versus the right tool for a specific business. Sometimes WordPress wins. Sometimes it's actively the wrong call. This is the honest matrix.

What WordPress is really good at

Before going anywhere near the criticism: WordPress runs roughly 43% of the web and didn't get there by accident.

Mature ecosystem. Every plugin you might need exists. Forms, SEO, e-commerce, multilingual, caching, security — all have multiple competing solutions.
Cheap labour pool. A WordPress developer in Baku costs less than a Next.js or Astro developer, and there are five times as many of them.
Editor familiarity. Most marketing people have used WordPress somewhere. The Gutenberg editor in 2026 is genuinely good for content teams.
Easy handoff. Your studio leaves; the next studio picks it up. There's almost zero "we use a proprietary thing" lock-in at the framework level.
Real CMS, by default. Custom posts, custom fields, taxonomies, user roles — all out of the box.

For a business with a marketing team that publishes weekly, a multilingual content strategy, and a budget that prefers low ongoing engineering cost, WordPress is genuinely a sensible choice.

Where WordPress quietly costs more

The launch quote is the trap. WordPress sites are cheaper at launch and more expensive in every year after launch. The reasons:

1. The plugin tax

A typical WordPress site relies on 12–25 plugins. Each plugin is a third-party update obligation. Quarterly:

Some break against the latest WordPress core.
Some are abandoned by their author.
Some change their pricing model from one-time to subscription.
Some develop security vulnerabilities that need urgent patching.

The realistic ongoing cost: 200–800 AZN per quarter in maintenance, depending on how loaded the site is. Over three years that's 2,400–9,600 AZN — often more than the original build.

2. Performance

A vanilla WordPress site on shared hosting opens at 3–5 seconds. With heavy plugins, page builders, and the database round-tripping on every request, mobile Lighthouse scores in the 30–50 range are routine. Each one of those numbers translates to a real conversion loss on mobile.

WordPress can be made fast — full-page caching, a managed host, image optimisation, lazy loading, asset deferral. But "making WordPress fast" is its own line item, usually 1,500–4,000 AZN, and it has to be re-validated after every major update.

3. Security surface

The same popularity that gives WordPress its plugin ecosystem also makes it the most-attacked CMS on the internet. A small Baku business running an outdated WordPress site is a routine victim of automated exploitation — defaced homepages, injected affiliate links, hijacked admin accounts. Fixing this is reactive work no business wants to budget for.

4. Design ceiling

Most WordPress builds in Baku use Elementor or a similar page builder. The output is functional and visibly templated. Buyers don't see this immediately — it looks "fine." Customers don't articulate it — they just sense the site isn't quite premium. Over time, the brand gets dragged down to its CMS's default visual grammar.

What "custom" actually means

"Custom" doesn't mean "from scratch in raw HTML." In 2026, custom builds for a small business website usually mean:

A static site generator — Astro, Eleventy, Hugo — for marketing sites and editorial content.
A React framework — Next.js, Remix — for app-shaped sites and complex interactivity.
A headless CMS — Sanity, Strapi, Payload — when the marketing team needs an editing surface.

None of these are speculative. They're the working defaults at most studios that take design and performance seriously.

The cost shape is different. Custom builds are 30–80% more expensive at launch. They're 60–80% cheaper to maintain after launch — no plugin tax, no quarterly compatibility fights, smaller security surface.

The decision matrix

Pick WordPress if:

You have a marketing team publishing 2+ articles a week.
You need a familiar editor for non-technical staff.
The site is content-heavy with hundreds of pages.
You expect frequent layout changes managed in-house.
Your budget is constrained at launch and you can absorb maintenance later.

Pick a custom build if:

You care about performance — Lighthouse 90+, sub-2s LCP, real mobile speed.
You care about design — a real, opinionated visual direction, not a template.
You publish less often — once or twice a month is fine to push to git via MDX.
You want to minimise ongoing maintenance.
The site needs custom interactions, animations, or integrations that page builders fight.

Pick a headless setup — custom front-end with a CMS like Sanity or Payload — if you need both: an in-house editor for non-technical staff and the design/performance/maintenance benefits of a custom front-end. It's the most expensive option at launch and the most flexible long-term.

The three-year total

For a representative seven-page marketing site:

WordPress: 1,800 AZN launch + ~600 AZN/quarter maintenance = ~9,000 AZN over three years.
Custom (Astro / similar): 4,500 AZN launch + ~250 AZN/quarter light maintenance = ~7,500 AZN over three years.
Headless (custom + Sanity): 7,200 AZN launch + ~400 AZN/quarter = ~12,000 AZN over three years.

These are representative ranges, not guarantees. The lesson isn't "custom is cheaper" — it's that launch price is not total cost, and a website platform choice made on launch price alone is incomplete.

At Qaynaq, we build on Astro for marketing sites and on a headless CMS when an editing surface is needed. We've also picked up plenty of WordPress projects to maintain or migrate; if you're weighing the three-year version of this decision, write to salam@qaynaq.com with what you have now.