When headless WordPress is worth it, and when it isn’t

A decision matrix from four real engagements. The honest answer is “rarely” — but the cases where it wins are specific.

What “headless WordPress” actually means in 2026

WordPress as a content API (REST or GraphQL), a separate front-end framework (Next.js, Astro, SvelteKit) rendering the public site, the editorial team working in WP-Admin as before. The framework has matured: Vercel’s preview branches just work, image CDNs handle the responsive-srcset work, ISR makes “static + dynamic” the default. The tooling is no longer the problem.

The four engagements

Project A — fintech marketing site, headless was right

120k monthly visits, regulated copy that needs branch-and-merge editorial workflow, a marketing team comfortable with Git. Built on WP + Next.js + Vercel. LCP sub-500ms globally, editorial preview branches, structured-data round-tripping, AB tests via edge-config. Worth the $32k migration; worth the $1,800/mo Vercel + CMS hosting.

Project B — nonprofit, headless was wrong

4k monthly visits, a single volunteer editor, no Git fluency on the team. We migrated to headless because “headless is the future.” Six months later: the volunteer team can’t preview content changes without filing a ticket. We migrated back to classic WP. The headless rebuild cost £18k; the rollback cost another £6k. Don’t be us.

Project C — B2B SaaS marketing site, headless was right (barely)

30k monthly visits, marketing team of three engineers (yes, really), need for sub-1s LCP on global edge, multi-language with locale-specific routing. Built on WP + Next.js + Cloudflare Pages. The decision was close — classic WP with aggressive caching could have hit the performance target. What pushed us headless was the marketing team’s engineering depth; they wanted to ship campaign branches via Git.

Project D — agency landing pages, headless was wrong

One-off campaign pages, three-week shelf life each, no maintenance plan. Classic WP + Elementor would have shipped these in a day. Headless added a week of front-end build per campaign with no measurable user-facing benefit. The agency switched back after two cycles.

The decision matrix

Headless WordPress wins when:

• Your marketing team can edit Git OR has dedicated engineering support for the front-end.
• Performance budget is non-negotiable and classic WP can’t meet it (sub-500ms LCP, edge-rendered, globally consistent).
• You need preview branches, AB tests at the edge, or structured-data round-tripping that’s painful in classic WP.
• You’re willing to pay 1.5 — 3× the operational cost (hosting + monitoring + maintenance).

Headless WordPress loses when:

• Your editorial team isn’t engineering-native and can’t reliably preview their own content.
• The performance target is sub-1s LCP — classic WP with aggressive caching gets you there.
• You’re below 50k monthly visits — the operational overhead doesn’t amortise.
• You don’t have $30k+ of migration budget and a multi-year horizon.

The right default in 2026

Classic WordPress, custom theme, aggressive caching, edge-distributed via Cloudflare. This is the right answer for 80% of B2B marketing sites I see. Headless is the right answer for the remaining 20%, and the 20% is specific enough that you’ll know when you’re in it.