Use Shopify. For most stores it is faster, cheaper and more reliable than anything a custom build would produce, and the monthly fee costs less than the engineering you would spend recreating what it gives you free.
Four situations genuinely justify building your own platform. If none of them describes you, the rest of this article will save you several lakh.
What you are actually choosing between
The debate is usually framed as two options. There are three, and the middle one is where most growing brands should land.
| Shopify with a theme | Headless storefront | Your own platform | |
|---|---|---|---|
| Time to launch | Weeks | 10 to 16 weeks | 20 weeks and up |
| Typical cost | 2 to 6 lakh | 8 to 20 lakh | 20 to 60 lakh |
| Who runs payments and tax | Shopify | Shopify | You |
| Ongoing fees | Plan plus transaction fees | Plan plus hosting | Hosting and your team |
| You control | Theme and apps | The entire front end | Everything, including the bugs |
| Right for | Most stores | Brands where speed and content matter | Rules no product can express |
Headless means a storefront you fully control sitting in front of a hosted commerce backend that keeps handling payments, taxes and the parts you gain nothing from rebuilding. It is underrated, and it is the answer more often than either extreme.
The four reasons that justify building
Any one of these can override everything else. None of them is "we want more control", which is what people usually say.
- Pricing that is per customer. B2B with negotiated rates, slab and quantity breaks, contract prices, credit limits and approval flows, where the catalogue itself differs by who is logged in. This is the most common legitimate reason, and it is the one products handle worst.
- Many sellers. Onboarding, KYC, commissions, split payouts, order routing across sellers, disputes. A marketplace is a different product wearing a store's clothes.
- Fulfilment rules of your own. Multi-warehouse allocation, made to order, partial shipments, serial number tracking.
- A catalogue that is genuinely unusual. Configurable products, dimension-based pricing, rentals, or subscriptions with real variation.
If you read those and recognised nothing, that is your answer, and it is a good one.
What "more control" actually costs
Owning the platform means owning the roadmap, the upkeep and the on-call. Shopify quietly handles a long list you inherit the moment you leave: PCI compliance, payment gateway changes, fraud rules, tax updates, uptime on sale days, and every security patch.
None of that appears in a build quote. All of it appears in your second year.
The honest test: if a payment gateway changes its API on a Tuesday, who fixes it, and by when? If the answer is "we would call someone", the subscription is doing more for you than you think.
Three-year cost, which is the comparison that matters
A build is quoted once and a subscription is paid forever, so single-year comparisons flatter whichever side the person quoting prefers. Here is the honest three-year shape for a store doing meaningful volume.
| Shopify with a theme | Headless | Own platform | |
|---|---|---|---|
| Build | 2 to 6 lakh | 8 to 20 lakh | 20 to 60 lakh |
| Platform fees, 3 years | 1.5 to 9 lakh depending on plan | same, plus hosting | hosting only, 1.5 to 6 lakh |
| Transaction fees | Gateway, plus Shopify's cut unless using Shopify Payments | same | gateway only |
| Maintenance and upkeep | Theme and app updates, small | Front end is yours | 15 to 25 percent of build, per year |
| Three-year total, roughly | 5 to 18 lakh | 12 to 32 lakh | 30 to 90 lakh |
The build column is the only one where the number keeps growing after launch through maintenance rather than fees. That is the trade being made: you stop renting and start employing.
The app tax nobody budgets for
Shopify stores rarely run on Shopify alone. A typical Indian store adds apps for reviews, wishlists, COD verification, shipping, WhatsApp notifications, loyalty and upsells, and each carries a monthly fee of a few dollars to a few tens of dollars.
Two costs follow. The obvious one is that fifteen apps at $15 becomes a meaningful monthly line. The less obvious one is performance: many apps inject scripts on every page, and a store that was fast on launch day becomes slow after a year of additions. We have audited stores where removing four unused apps did more for load time than any code change.
Before concluding that Shopify is expensive, count your apps and ask which three you would actually miss. Before concluding a custom build avoids this, remember that you will build equivalents for the ones you keep.
What each route needs from your team
The choice commits you to a staffing shape, and this is usually decided implicitly rather than deliberately.
Shopify with a theme needs someone who can edit content and manage apps. Occasional agency help for theme work. No engineers.
Headless needs a front-end developer available, or a retained agency. Something will break on a platform update and it will not fix itself.
Your own platform needs engineering capacity permanently. Not a person who built it and left, but someone on call when checkout fails during a sale. If you cannot name that person or that retainer, the build will decay regardless of how well it was written.
The most expensive mistake in this article is a custom platform built by an agency and then left unstaffed. It is worse than Shopify in every dimension within two years.
What does not decide it
Page speed. A well-built Shopify theme is fast, and a badly built custom store is slow. If your current store is slow, the cause is almost certainly images, apps and third-party tags rather than the platform. That is fixable without leaving.
SEO. Both can rank. Shopify's URL structure has quirks, none of them fatal, and a migration done badly will cost you far more traffic than the structure ever did.
Transaction fees. Real, and worth modelling at your volume, but rarely decisive on their own until you are large enough that the other three reasons apply anyway.
B2B on Shopify, which has changed
The strongest historical reason to leave Shopify was B2B: customer-specific pricing, price lists per account, minimum order quantities, purchase orders and net terms. Shopify has since shipped B2B features on its higher plans, and for a distributor with straightforward account pricing they may now be enough.
The test is how far your rules go beyond a price list. Per-account catalogues where products are hidden entirely, slab pricing that changes within a single order, approval chains before an order is placed, credit limits enforced at checkout, and contract prices that expire on a date: those still exceed what the platform expresses, and they remain the clearest reason to build.
Check the current capability before assuming either way. This is the area where the answer has moved most in three years, and an agency quoting you a B2B build without asking whether the platform now covers it is not doing you a service.
Migrating without losing your traffic
Migrations go wrong in one predictable way: the store looks better and organic traffic halves. It is worth stating plainly because it is the single largest hidden cost of changing platform.
Four rules, in order of how much damage skipping them causes. Map every old URL to its new one before launch, including collection and filter pages that rank. Keep the category and product structures that already earn traffic, even where a cleaner structure is tempting. Move order history and customer accounts, or you break returns and lose repeat-purchase data. Watch rankings daily for a month, because the window to catch a mistake is short.
If a vendor's migration plan does not include a URL map, that is the whole conversation.
Built for how India buys, either way
Whichever route you take, the same requirements hold, and an overseas template ignores most of them:
- Cash on delivery, still a large share of orders in many categories, which changes address verification, fraud rules, partial prepayment and courier reconciliation.
- UPI first, so the checkout opens with the method most customers actually use.
- GST invoices with correct place of supply and HSN codes, generated at order time.
- Courier reality: serviceability by pincode, weight and dimension rules, and integrations that make tracking automatic.
Shopify handles these through apps and Indian payment gateways. A custom build handles them because you built them. Neither handles them by accident.
Questions people ask
Is Shopify customisable enough?
For the storefront, yes: themes are fully editable and a headless front end removes the limit entirely. The real constraints are in checkout and in the data model, which is where the four reasons above bite. If your restriction is visual, you do not need to leave.
What are the best Shopify alternatives in India?
WooCommerce if you are already on WordPress and have someone to maintain it, or a headless setup on a commerce API when you want control of the front end without owning payments. The alternative worth avoiding is a cheap custom build, which gives you the maintenance burden of ownership with none of the maturity.
Can Shopify handle Indian payment methods and GST?
Yes, through Indian gateways and apps: UPI, cards, wallets, net banking and cash on delivery, with GST invoicing available through apps or your accounting integration. COD in particular needs an app for address verification and partial prepayment rules, which is one of the apps worth paying for.
Is WooCommerce a cheaper alternative?
The licence is free and the total cost is not. WooCommerce needs hosting, updates, security patching and someone who owns all three. It is a sound choice when you already run WordPress and have that person. It is a poor choice when "free" is the reason for choosing it.
How much does an e-commerce website cost in India?
A Shopify store with a custom theme is typically 2 to 6 lakh, a headless storefront 8 to 20 lakh, and a full custom platform 20 to 60 lakh. The drivers are in our app development cost guide, and the platform detail is on our e-commerce platforms page.
Should we migrate off Shopify to grow?
Only for one of the four reasons. Migrations go wrong in a predictable way: the store looks better and organic traffic halves. If you do move, map every old URL before launch, keep the structures that already rank, and watch rankings daily through the switch.
Scale, and what actually breaks on a sale day
Founders worry about whether a platform can handle their volume. For almost every Indian store, it can, and the things that break on a sale day are rarely the platform.
What actually fails: payment gateways rate-limiting or timing out under a spike, inventory going negative because two systems both think they own stock, courier serviceability checks slowing every checkout, and apps whose scripts collapse under load. On a custom build, add your own infrastructure to that list, which is precisely the reason the build column carries an on-call requirement.
The practical preparation is the same either way. Load-test the checkout, not the homepage. Agree in advance who watches the store during the sale and what they are authorised to switch off. Decide beforehand which non-essential features get disabled if things slow down, because deciding that at 11pm during a campaign produces worse choices.
A worked example of the decision
A distributor selling to 400 retail accounts, each with negotiated pricing, minimum order quantities and 30-day credit terms, doing roughly 2,000 orders a month.
Reason one applies immediately: pricing is per customer, and it goes beyond a price list because credit limits must be enforced at checkout and some accounts see a restricted catalogue. Reason three partially applies, since orders ship from two warehouses with allocation rules.
Shopify B2B might cover the price lists. It will not cover credit enforcement at checkout or the restricted catalogue without workarounds that become their own maintenance burden. Three-year cost of a build lands somewhere in the 30 to 50 lakh range including upkeep, against maybe 12 lakh of platform and app fees plus the cost of the manual work the workarounds create.
The decision turns on whether that manual work costs more than 20 lakh over three years. For 2,000 orders a month with credit terms, it usually does. For 200 orders a month, it usually does not, and the same business should stay on the platform and accept the friction.
That is the shape of the calculation. Do it with your own numbers rather than accepting anyone's rule of thumb, including ours.
How to decide in ten minutes
Write down the pricing rule, the fulfilment rule and the catalogue shape that make your business work. If a product manager at Shopify would recognise all three as normal, stay. If any one of them makes you say "well, ours is different because", that is the sentence a platform build exists for.
We say "use Shopify" more often than an agency is supposed to, because a platform build is an expensive way to get a store. When it genuinely is the right answer, our e-commerce platforms page covers what that build contains, and e-commerce development covers the Shopify and headless routes.