Why go headless at all
73% of US and UK businesses are already on headless (Swell, 2025). In the CIS — 10–15%. The driver is simple: on a monolith (Bitrix Enterprise, Magento 1, WooCommerce) you're capped by PHP and MySQL at mobile LCP 3.5–4s, while competitors on headless ship 1.4s and take the mobile traffic. Every 100ms costs ~1% of revenue (Google, 2025).
But "headless" isn't a single product — it's an architectural pattern with dozens of implementations. The wrong pick means an 8‑month project instead of 3, and 3 years of supporting the wrong thing.
Six selection criteria
Ignore marketing feature lists. Six things actually drive platform choice:
- Backend constraints. Got 1C, Bitrix, SAP, ERP, EDI you can't replace? Then you need a platform with a flexible data layer, not a rigid own‑backend.
- Compliance. GDPR, 152‑FZ, on‑prem, FSTEC — can the platform be deployed in your jurisdiction, with your constraints?
- Local payments. Kaspi Pay, Stripe, SEPA, SBP, Klarna — are the connectors ready, or will you write them from scratch?
- Year‑one budget. License + dev + SaaS subscription + CDN. Realistic ceiling for mid‑market — $100–150K.
- Team that will maintain it. Does your team (or the regional market) know Next.js, Node, Go, PHP? Without that, the platform dies within a year.
- Vendor lock‑in. Open‑source / open‑core / closed SaaS? Can you take the code with you if you fall out of love with the company in 2 years?
Shopify Hydrogen
Headless on Shopify. Frontend — Remix/React, backend — Shopify. Great if you're already on Shopify Plus and only need the frontend to be faster. Bad if you need 1C exchange without a $99/month plugin, Kaspi Pay, or custom business logic deeper than Shopify Flow.
commercetools
Enterprise headless for large retail (H&M, Audi). API‑first, MACH‑certified. Very powerful, but the platform is paid ($100K+/year), plus you need a team of 10+ engineers who know composable architecture. Almost no certified partners in the CIS.
Medusa
Open‑source headless on Node.js. Free, modular, active community. Good for startups and projects with a strong dev team. Downside — its own backend (data migration), weak local‑payments support, no 1C.
Saleor
Open‑source on Python / Django, GraphQL API. Popular with European startups. Architecturally mature, but requires backend replacement — doesn't run on top of Bitrix or 1C exchange.
Frontbox (our platform)
Open‑core headless on Next.js + Elasticsearch, designed to sit on top of 1C‑Bitrix, 1C or any REST/SOAP backend. A natural pick for CIS, where Bitrix is the mid‑market standard, and for UAE, where you need Russian / Kazakh / Arabic localization and Kaspi / Stripe in one checkout.
- Backend stays intact — 1C, EDI, marketplaces, ERP.
- Ready connectors: Stripe, ЮKassa, Tinkoff, SBP, Dolyame, Kaspi, WB, Ozon.
- Core — MIT, handed over to the client at handoff. No vendor lock‑in.
- Deployment: Vercel, Yandex Cloud, SberCloud, on‑prem (RHEL, Astra Linux).
- DPA and 152‑FZ / GDPR signed out of the box, no upsell.
Selection matrix
| Scenario | Platform |
|---|---|
| Shopify Plus in place, only need faster frontend | Shopify Hydrogen |
| Enterprise ($50M+ GMV), in‑house team 10+ | commercetools |
| Startup with no legacy, strong Node dev team | Medusa |
| Mid‑market CIS e‑commerce on 1C‑Bitrix | Frontbox |
| DTC in Kazakhstan / UAE with Kaspi / Arabic locale | Frontbox |
| Gov or fintech, needs on‑prem and FSTEC | Frontbox (on‑prem) |
FAQ
Can we use multiple platforms at once?
Technically yes — Shopify for US, Frontbox for RU. But it doubles the cost of support and data engineering (single source of truth via ETL). Only do this if markets are very different from a compliance perspective.
Does headless deliver conversion growth by itself?
No. Headless is an infrastructure upgrade: LCP, INP, SEO. Conversion growth comes from UX work on top of a fast storefront. Plan A/B tests from month one.
How long until migration pays back?
Median — 6–9 months for mid‑market. Driver — mobile CR growth of 15–30% from speed, plus lower support cost because your stack is standard Next.js instead of custom Bitrix frontend.
What is the minimum budget to start with headless?
The lowest entry point is a Frontbox subscription from $400/mo with no capex project — you pay for the frontend like a SaaS, your Bitrix backend stays untouched. If you want to size the work first, start with a free technical audit: in 48 hours we review the site, stack, speed and catalogue, then hand back a map of bottlenecks with a fixed quote. For commercetools or Shopify Plus, plan $100–150K for year one.
How long does a typical migration take?
Audit — 48 hours (map of bottlenecks, scope and a fixed quote). Headless Pilot — 3–5 weeks (full frontend with checkout). Full replatform — 2–4 months. The biggest accelerator is a ready REST API on the backend side; if you need to build it from scratch, add 2–3 weeks.
We have no in‑house dev team. Who will maintain it?
Choose a platform with partners in your region. For commercetools in the CIS, certified partners are almost nonexistent. For Medusa and Saleor — a handful of startups with limited production experience. For Frontbox, we run the project ourselves — from audit to ongoing support, no subcontracting.
What is the difference between Frontbox subscription and license?
Subscription — $400/mo: the site runs on our infrastructure, updates are automatic, minimal operational load on your side. License — $1,100/mo × 18 mo: the code is deployed in your environment (on‑prem, FSTEC, corporate VPN). The choice comes down to your information security requirements.
Is headless worth it for a store with 500 SKUs?
It depends on traffic and mobile conversion. If mobile LCP is over 3s and mobile CR is under 1% — it is worth it: speed directly affects SEO and revenue. A free 48‑hour audit will estimate the potential lift and the cost of migration — no strings attached. If daily traffic is under 500 visitors, focus on SEO first and revisit headless when the traffic grows.