Compare · Headless vs Bitrix - One more cache layer on Bitrix or a headless move to Next.js?
This is the fork every mid‑market e‑com owner on 1C‑Bitrix faces in 2026. We compare the two paths across 10 criteria — from mobile LCP to 3‑year TCO. No marketing: with numbers and real projects that went through both.
The local‑maximum trap - Why 'one more cache layer' keeps getting worse
Trying to squeeze performance out of the monolith through Varnish, lazy loading and minification gives you linear improvement that tops out at 2.5–3s LCP. That's the PHP + MySQL + jQuery ceiling — past that, either rewrite the stack or live with it. Every next iteration buys you less, and the debt compounds.
| Criterion | 1C‑Bitrix optimization | Headless on Frontbox |
|---|---|---|
| Time to first result | 2–4 weeks for patch releases | 2–4 weeks for a one‑category pilot |
| Achievable mobile LCP | 2.5–3.2s (ceiling of PHP + Varnish) | 1.5–2.0s (SSR/ISR + edge cache) |
| INP (Core Web Vitals) | 300–500ms due to jQuery and hacks | < 200ms on React Server Components |
| Catalog search | Sphinx or built‑in — no synonyms | Elasticsearch with morphology and synonyms |
| Scaling under peak load | Vertical, expensive, capped by MySQL | Horizontal on the frontend, edge CDN |
| Support team cost | Rare Bitrix + custom stack, expensive | Standard Next.js + TypeScript, mainstream |
| Vendor lock‑in | Custom Bitrix = strong lock‑in | Frontbox open‑core (MIT), minimal |
| SEO risk | Zero (nothing changes) | Low with 301 map and staged cutover |
| 3‑year TCO (10K‑SKU store) | $90K–$160K (Enterprise license + support) | $110K–$180K (replatform + retainer) |
| AI modules and personalization | Limited by architecture | Native: embedding search, RAG, copilots |
When Bitrix optimization is the right call - Staying isn't always worse
We're not fans of "rewrite everything". Sometimes optimization is the smarter path. If any of the below applies — don't rush into migration.
- Small store (under 1,000 SKUs). On a small catalog, the difference between LCP 2.5 and 1.5s usually doesn't pay back. CDN, image optimization and LCP bundle come first.
- No team to maintain Next.js after release. A platform without an owner dies within a year. If there's no in‑house Next.js engineer — start with our retainer or team training, not a migration.
- Redesign/rebrand planned in the next 12 months. Migration and redesign at the same time doubles the risk. Run them sequentially.
When headless is the only way - Signals that optimization isn't enough anymore
- Mobile LCP > 3s and competitors on headless. Each 100ms = ~1% of revenue (Google, 2025). At 3+ seconds you're losing mobile traffic.
- Planning AI personalization or semantic search. On Bitrix these cost huge custom layers. On headless they're native.
- Your Bitrix dev team is shrinking. The market has 3× more senior Next.js engineers than senior Bitrix ones. Trend is accelerating.
- Need omnichannel (web + mobile app + POS + kiosks). Headless gives one API contour for all channels. Monolith — a separate project per channel.
- Expansion into KZ / UAE / new markets. Frontbox ships with native Kaspi Pay / Stripe / multi‑language out of the box.
Not ready to decide yet? Start with a Catalog Probe
Upload a CSV of your catalog. In 3 days you'll see a live headless preview and measure how much better LCP and search get. Decide on facts, not slides.
What else helps - Related materials
Tell us about your project
Our offices
- Russia
Saint Petersburg, Rizhskaya st. 5, bldg. 1, office 402
+7 (967) 555-90-32 - Kazakhstan
Almaty
+7 (707) 340-29-12