Headless Commerce: What Changes in the First 90 Days

The headless commerce pitch is easy to like: a fast storefront, total control over design, and no platform dictating what your checkout looks like. The invoice for that freedom arrives in the first ninety days, and it lands on the operations and engineering leads at 20-200 person Indian companies who chose headless for a rebrand or a new dealer portal rather than out of a specific technical need. Here is what actually changes, week by week, and where the money and hours go.

Days 1-15: You have just taken on infrastructure

Headless means you now own the rendering layer. That is not a metaphor. It is a server, a build pipeline, a CDN configuration and a deploy process that someone has to own at 11pm on the last day of a sale.

Before you ship a single product page, most teams assemble some version of this stack:

• Commerce engine: Shopify Storefront API, Commerce Layer, Medusa, Saleor, or your existing backend exposed over REST or GraphQL.
• Frontend hosting: Vercel, Netlify, Cloudflare Pages, or a Node container on AWS or DigitalOcean. Each has a different build limit and cost curve.
• Content: Sanity, Contentful, Payload or Strapi for banners, category copy and policy pages.
• Search: Algolia, Typesense or Meilisearch, unless the commerce API exposes something usable.
• The glue: webhooks, a queue for order events, Sentry or similar, and analytics that survive client-side rendering.

That is five new vendor relationships and at least two new runbooks. None of it is hard. All of it takes calendar time.

The cost shape changes before the feature list does

On a hosted platform, your cost is mostly subscription plus apps, and it roughly tracks order volume. Headless inverts half of that: you pay in engineering hours and traffic-based hosting, both of which are less predictable.

A worked example. A Gurgaon distributor with about 4,000 SKUs and 300 dealer accounts runs on a platform plan of roughly ₹35,000 a month plus ₹20,000 in apps. Moving the storefront headless typically adds ₹15,000-₹60,000 a month across hosting, CDN, search and CMS at their traffic levels, plus the build itself. Two engineers for six weeks at a fully loaded ₹1.4 lakh a month each is about ₹4.2 lakh of internal time before you pay a single contractor. Platform fees may well fall. Total first-quarter cost almost never does, and that is the trade you are actually approving.

Weeks 3-8: Where the hours disappear

The work is not glamorous. In our experience, the first two months go into problems a hosted platform had already solved for you.

Cart and session state across a CDN is the first one. Then authentication, because B2B pricing means the catalogue looks different to every dealer, and you cannot cache a personalised page the way you cached a public one. Then the hundred small rules that used to live in plugins: pincode serviceability, minimum order values, COD eligibility by tier, and price lists that change when a contract renews.

The integration that usually decides the timeline is inventory and pricing sync between your ERP or accounting system and the commerce engine. If Tally, Zoho Books or a custom ERP is the source of truth, someone has to define what happens when stock changes mid-session, and how a price update propagates without a full rebuild. Budget more for this than for the frontend.

The compliance list is longer than you think

Headless does not remove Indian compliance requirements. It removes the plugins that used to handle them.

• GST invoice generation with correct HSN codes, place of supply and tax breakup, issued from your own backend.
• E-invoicing IRN generation and cancellation if you cross the turnover threshold.
• TDS treatment on payment gateway settlements, which your finance team will ask about in month two.
• DPDP Act consent capture and retention rules, because you are now storing customer data in more places, not fewer.
• WhatsApp order confirmations and shipping updates, usually the channel your customers actually read.

Each is a small build. Together they are two to four weeks that rarely appear in the original estimate.

Days 60-90: The payoff, measured honestly

This is when headless starts to earn its keep, but only if you agreed on metrics before day one. The usual wins: materially faster product pages on mobile, merchandising changes that ship without a developer, and per-dealer catalogues and pricing that a hosted theme simply cannot express. The usual non-wins: better conversion by itself, cheaper hosting, and less work for your team. Pick two or three numbers — mobile LCP, time to publish a campaign page, dealer checkout completion — and review them at day 90. If they have not moved, the problem was never the platform.

Where to start

Do not go headless across the whole site at once. Pick the surface with the clearest payoff, usually the category and product pages, and keep checkout on the hosted platform for the first quarter. Write down who owns hosting, builds and deploys, because 'the agency does it' stops being true the moment the retainer ends. Set a day-90 review with two or three named metrics and a real decision: extend headless to checkout, or stop and keep what you built.

If you want a second opinion before signing anything, Maysan Labs builds storefronts, dashboards and B2B portals from Gurgaon and Cardiff, and we are happy to look at your current stack and tell you whether headless is worth it at your volumes.