This is for businesses ready to sell online who want the platform decision made on catalogue size, margin and order volume rather than on whatever the agency happens to build. It is not for you if you have no product photography, no stock system, and no one to pack orders — a store is an operations commitment before it is a website, and launching without that side ready produces refunds and bad reviews rather than revenue.
The decision
Ecommerce website development in India turns on three numbers, not on which platform an agency happens to enjoy building. Most agencies build one platform and recommend it universally. The honest version is that the choice turns on three figures: how many SKUs you carry, how many orders you expect a month, and what your margin per order is. Those three decide it more reliably than any feature comparison.
A hosted platform wins below roughly 500 orders a month. You pay a subscription and a small cut per order, and in exchange somebody else handles hosting, security patching and PCI compliance. At that volume the platform fee is smaller than the cost of maintaining infrastructure yourself, and the faster launch is worth real money.
Self-hosted wins above that, and wins decisively where margins are thin. No platform cut, full database access, and unrestricted checkout logic — at the cost of owning updates, backups and security. A custom build is the right answer far less often than it is proposed: essentially only when your selling model genuinely does not fit either platform, such as configurable industrial products or complex multi-vendor logic.
Compare
| Hosted (Shopify) | Self-hosted (WooCommerce) | Custom build | |
|---|---|---|---|
| Best below/above | Under ~500 orders/mo | Over ~500 orders/mo | Model does not fit either |
| Build cost from | ₹34,999 | ₹34,999 | Quoted per scope |
| Monthly platform cost | Subscription + per-order fee | Hosting only | Hosting + engineering |
| Time to launch | 3–6 weeks | 5–8 weeks | 12 weeks and up |
| Who patches security | Platform vendor | You or your agency | You or your agency |
| Checkout flexibility | Constrained | Unrestricted | Unrestricted |
| B2B / per-customer pricing | Awkward, app-dependent | Native | Native |
| Right for most first stores | Usually yes | Sometimes | Rarely |
What's included
The India-specific configuration below is the part that decides whether a store works commercially. It ships on all three routes.
Razorpay, PayU or Cashfree integrated with UPI enabled, tested with live transactions including the refund path. Gateway KYC runs in parallel with the build because bank verification typically takes 3 to 7 working days on its own.
Pincode serviceability checked at the product page rather than at checkout, COD eligibility limited by order value, and partial-prepaid options in categories where return-to-origin eats the margin. RTO is the quiet killer of Indian D2C unit economics.
GSTIN, HSN codes per product, and the correct intra-state versus inter-state tax split on a document your accountant and your B2B customers will both accept. Sequential numbering configured properly.
Shiprocket, Delhivery or your existing courier connected so labels print, tracking updates flow, and customers get status emails without anyone copying numbers between systems. Rate rules by weight and zone.
Built to load on a 4G connection in a tier-3 town, not just on office broadband. Images converted to WebP, scripts kept minimal, and the product page checked against a throttled mobile connection before launch.
GA4 ecommerce events and Google Tag Manager configured so you can see which channel produced which order, not just how many people visited. Feeds directly into whether ad spend is working.
How we work
Length depends on the route. The sequence does not.
Catalogue size, expected order volume and margin reviewed, and the platform chosen on those numbers. You get the reasoning in writing, including why the other two were rejected.
Store architecture, catalogue import, theme or template development. Reviewed on a live preview link as it is built rather than revealed at the end.
Gateway, COD rules, GST invoicing, shipping and analytics. This runs partly in parallel with the build because gateway KYC has its own clock.
Live transaction testing, redirect verification if migrating, Search Console submission, and a handover session covering order management and refunds.
Who this suits
It suits businesses already selling — through a shop, a WhatsApp catalogue, Instagram DMs, or a marketplace — who want to stop paying a marketplace commission and own the customer relationship. Existing demand makes the store an efficiency gain rather than a gamble.
It suits brands with a defensible product and repeat purchase potential, typically anywhere from pre-revenue up to around ₹5 crore a year. Below that, a lean hosted store is almost always right. Above it, the platform question deserves the full arithmetic.
It suits manufacturers and distributors moving to B2B online ordering, where per-customer pricing and bulk tiers matter more than a pretty product page — see manufacturing and export marketing for that variant.
It suits badly if the operations are not ready. Stock accuracy, packing capacity, a returns policy and someone to answer questions all need to exist before launch. A store that takes an order it cannot fulfil damages the brand more than not having a store at all — ecommerce website development is an operations decision wearing a website's clothes.
Transparent pricing
Fixed-price for the build. Platform subscription, hosting and gateway charges are billed to you by those vendors directly.
All prices in INR, exclusive of GST. Gateway charges of roughly 2% per transaction are deducted at settlement by the gateway.
Common questions
The build starts at ₹34,999 for a starter catalogue and ₹59,999 for a full one with filters and logistics integration. Custom builds are quoted per scope and start considerably higher. Running costs sit on top: platform subscription or hosting, roughly 2% gateway charges per transaction, and any paid apps. We give the full three-year picture in the quote, not just the build figure.
Three to five weeks for a starter store, five to seven for a full catalogue, twelve or more for a custom build. The two things that reliably push it later are product data and photography arriving late, and payment gateway KYC, which sits with your bank. We start gateway onboarding in week one so it is not the bottleneck.
Whichever your numbers point to. Under roughly 500 orders a month, a hosted platform almost always costs less once maintenance is priced properly. Above that, or where margins are thin, self-hosted wins. Custom is right far less often than it is proposed. You get the reasoning in writing, including why the other two were rejected.
In most Indian categories, yes — but it needs controlling. Unrestricted COD invites return-to-origin, and RTO quietly destroys margin because you pay shipping both ways on a sale that never happened. We limit COD by order value and pincode, check serviceability on the product page, and offer partial prepaid in high-RTO categories.
Yes. The product data is the tedious half and the URL redirect map is the half that costs money if skipped. Every platform enforces its own URL structure, so product addresses change on migration. We crawl the existing store first, map every old address to its new one, and verify each redirect resolves in a single hop after go-live.
That is a design constraint from the start, not a later optimisation. Images are converted to WebP, scripts kept minimal, and the product and checkout pages tested on a throttled mobile connection before launch. A store built and only ever checked on office broadband will disappoint a large share of Indian buyers.
We will say so. If stock accuracy, packing capacity, photography or returns handling are not in place, a store creates problems rather than revenue. A catalogue website with WhatsApp enquiry is often the right intermediate step, and considerably cheaper. That advice costs you nothing and saves a bad launch.
Related
Let's talk
Tell us your catalogue size, expected monthly orders and typical margin. You will get a straight recommendation and a fixed quote — even if the answer is that you are not ready yet.
Or drop your details — we reply within 4 hours
No spam. No commitment. Response within 4 working hours.