Ask us what a custom Shopify app costs and the honest answer is a range: roughly ₹1.5–4 lakh for a single-purpose private app, ₹5–12 lakh for something with an admin interface, webhooks and its own billing logic, and ₹12–30 lakh for middleware that sits between Shopify and an ERP, WMS or a subscription engine. But the build price is the least interesting number in the decision. What matters is the 36-month total, because that's the horizon over which an off-the-shelf app quietly bills you ₹8–15 lakh and a custom app quietly bills you for hosting, API upkeep and the developer who understands it. Model both, and the answer usually flips at least once while you're doing the arithmetic.
Below is the model we actually use when a founder asks whether to build. It has five lines, and skipping any of them is how people end up with a ₹9 lakh app that does what a $49 subscription did.
The five lines in a three-year custom Shopify app development cost model
For the custom side:
- Build. One-off. Scope, design, development, QA, Shopify app review if you're listing it publicly.
- Infrastructure. Anything that runs outside Shopify. A single-purpose app with a couple of webhooks might run for ₹1,500–3,000 a month on a small instance. Add queues, workers, a Postgres instance and a Redis cache for a sync app and you're at ₹12,000–20,000.
- Maintenance. Budget 15–25% of the build cost per year. This is not padding. Shopify ships a new Admin API version every quarter and supports each one for a limited window, so an app you built in 2025 needs deliberate attention in 2026 or it starts throwing deprecation errors at the worst possible moment.
- Change requests. Every custom app grows. A new warehouse, a new marketplace, a COD partner, a GST rule change.
- Bus factor. The cost of documentation and handover so the app survives one developer leaving. Most teams price this at zero and pay it later at 3x.
For the off-the-shelf side there are only two lines that people remember, and three they forget: the subscription, the usage or revenue component, the second app you need because the first one doesn't do one thing, the annual price increase, and the migration cost when the vendor sunsets a plan you're grandfathered into.
One tax note before the numbers. Foreign SaaS billed to an Indian business attracts 18% GST, and Indian development services do too. If you're GST-registered and claiming input credit, compare both sides pre-tax so you're not double-counting. If you're not registered, add 18% to both columns and the ratio stays roughly the same.
Worked example one: the case where the App Store wins outright
A D2C brand wants subscriptions. Coffee, supplements, pet food, the usual suspects. Recurring revenue through the programme is ₹25 lakh a month and growing about 40% a year.
Off-the-shelf subscription apps commonly charge a base fee plus a percentage of subscription revenue. Take $99 base plus 1%. At ₹85 to the dollar:
- Year 1: ₹25,000 (1% of ₹25L) + ₹8,415 base = ₹33,415/month, or ₹4,00,980
- Year 2 at ₹35L/month: ₹35,000 + ₹8,415 = ₹43,415/month, or ₹5,20,980
- Year 3 at ₹49L/month: ₹49,000 + ₹8,415 = ₹57,415/month, or ₹6,88,980
Three-year total: ₹16,10,940. That is a genuinely uncomfortable number, and it's the number that sends founders to us asking for a build.
Now the custom side. Subscriptions on Shopify mean selling plan groups, subscription contracts, the Customer Account extension for self-serve pause and skip, a payment mandate that survives card expiry, dunning logic, and retry behaviour that works with Razorpay's recurring flows as well as Shopify Payments. We'd quote ₹11 lakh for that and we would not enjoy quoting it lower. Hosting with billing workers and retry queues: ₹6,000/month, so ₹2,16,000 over 36 months. Maintenance on something that touches money: ₹25,000/month retainer, ₹9,00,000.
Total: ₹22,16,000. Six lakh more expensive than the app you were annoyed about, and you now own the pager duty for failed renewals.
Build subscriptions yourself only when the recurring revenue is large enough that 1% dwarfs a maintenance retainer, which in practice means north of ₹1 crore a month, or when your commercial model genuinely doesn't fit any vendor. Until then, negotiate the percentage down. Vendors move on rev-share more often than they admit.
Worked example two: where the custom app development cost actually pays back
Different store. High order volume, low AOV, 25,000 orders a month, and inventory that has to stay in sync across Shopify, a 3PL's WMS, and two marketplaces. Off-the-shelf integration platforms price this per order or per task.
Assume $499/month base including 10,000 orders, then $0.02 per order after that.
- Year 1 at 25,000 orders/month: $499 + (15,000 × $0.02) = $799/month ≈ ₹67,915, or ₹8,14,980
- Year 2 at 35,000: $499 + $500 = $999/month ≈ ₹84,915, or ₹10,18,980
- Year 3 at 50,000: $499 + $800 = $1,299/month ≈ ₹1,10,415, or ₹13,24,980
Three-year total: ₹31,58,940.
Custom middleware for the same job: ₹14 lakh build, ₹15,000/month infrastructure (₹5,40,000 over three years), ₹20,000/month maintenance (₹7,20,000). Total ₹26,60,000.
Custom is about ₹5 lakh cheaper. Roughly 16%. Here's the part nobody says out loud: 16% over three years is not a good enough reason to build. Estimation error on a ₹14 lakh build is easily 20%. If the only argument for custom is a modest cost saving, buy the connector and spend your management attention elsewhere.
The real argument in this scenario is different. When the same middleware has to handle COD reconciliation against a courier's remittance file, split a single Shopify order across three fulfilment locations by pincode serviceability, and hold back marketplace stock during a festive sale, most connectors will do two of the three and charge you for a workaround on the last one. You build because the constraint is yours and nobody sells a product shaped like it. The cost being level is what makes the decision easy, not what drives it.
Usage-based and revenue-share pricing is the real trigger to build
Flat-fee apps almost never justify replacement. A $49/month app is ₹4,165 a month, ₹1,49,940 over three years. You cannot build anything for that. Even five flat-fee apps at $49 come to ₹7.5 lakh over 36 months, which buys a mid-complexity build with nothing left for maintenance.
Apps priced on your success are the ones to watch. Revenue share, per-order, per-SMS, per-1,000-emails, per-active-customer. Those bills grow while your build cost stays fixed, so there's a crossover point, and you can calculate roughly when you'll hit it. Plot the app's fee against your forecast, add your build plus three years of running cost as a flat line, and see where they meet. If the crossover is inside 18 months, building is worth a serious conversation. If it's at month 30, do nothing and revisit next Diwali.
The costs that never make it into the spreadsheet
API version upkeep. Shopify's quarterly API releases mean any custom app carries a standing obligation. On a small app that's a few hours a year. On an app touching orders, fulfilment and payments it's a real testing cycle. We've had first-pass upgrades break something in staging often enough that we now schedule them rather than react to them.
The speed tax on off-the-shelf apps. Public apps inject scripts. Six or seven of them and you're looking at 400–900ms of extra JavaScript execution on mobile, which shows up as a worse LCP and a worse INP. A custom app built as a theme app extension with server-side rendering and no third-party script loads nothing extra on the storefront. That's a conversion argument dressed as a cost argument, and it's one of the few genuine wins for custom that doesn't appear anywhere in a licence-fee comparison. If you want to see what your current app stack is costing you in load time before deciding anything, SwiftStore scans it and tracks the score over time.
Vendor risk both ways. App vendors get acquired, change pricing, and occasionally stop supporting a plan. Custom apps have the mirror-image risk: the person who built it moves on. Ask for the repository, the deployment pipeline and a written runbook as deliverables, not favours. If your agency can't hand those over on day one of a handover, the build was never really yours.
Before you build: the cheaper things that solve half these problems
A surprising share of custom app requests we scope turn out not to need an app.
Metafields and metaobjects handle structured content that people used to build apps for. Shopify Functions cover discount, shipping and payment customisation logic without a hosted service, so there's no monthly infrastructure line at all. Theme app extensions and app blocks let you add storefront UI without a script tag. Flow handles a lot of tagging, alerting and internal routing that used to justify a small tool.
A Function plus a couple of metaobjects might be ₹80,000 to ₹2 lakh of work with almost no running cost. That's the option to rule out first, and it's the one an agency with an app development practice is least incentivised to suggest. We suggest it anyway, because a client who spent ₹1.8 lakh instead of ₹9 lakh comes back.
Why our quoted range for custom Shopify app development is so wide
Three things move the number more than anything else.
Whether it touches money or inventory. An app that renders a size chart can ship with light QA. An app that reprices products, splits fulfilment or charges cards needs edge-case testing, idempotency, audit logging and a rollback plan. Same feature count, double the hours.
Public listing or not. A custom app installed on one store skips the App Store review process, the mandatory webhook compliance work, the billing API integration, the privacy policy and the support burden. Listing publicly adds 30–50% to the build and gives you an ongoing support commitment. We've published our own apps, so we know exactly what that tail costs. Don't list unless you intend to sell.
How clean your data is. Integration builds get expensive when SKUs don't match between systems, when the ERP has three records for one product, or when someone has been maintaining stock in a spreadsheet. That's discovery work, and it's real work. Our broader breakdown of Shopify development cost in India covers how we scope this.
A ten-minute decision rule
- Total what the incumbent apps will bill you over 36 months at your forecast volume, not today's.
- Add 18% GST if you can't claim input credit.
- Get a build estimate, then add three years of infrastructure and 20% of the build per year for maintenance.
- If custom isn't at least 40% cheaper, buy. The gap has to be big enough to absorb estimation error.
- Override step 4 only if the off-the-shelf option cannot do something that actually costs you orders. Then build, and be clear that you're buying capability, not savings.
The stores that get this wrong most often are the ones at ₹3–8 crore a year who've read that custom is more professional. It isn't. It's a fixed cost with a maintenance obligation, and it earns its place when a percentage-of-revenue fee or an unfixable constraint makes it earn its place.
If you've got a specific app bill you want to argue with, send us the app names, your monthly order volume and the one thing the app won't do. We'll build the 36-month comparison for both options and tell you if the answer is to do nothing. Our app development page covers how we scope builds, and a free audit is the fastest way to get the numbers on paper.