The honest answer first
A custom Shopify app costs anywhere from a few thousand francs to well into five figures, and the range is that wide because "app" describes everything from a single Shopify Function to a full ERP integration. Anyone who quotes a precise number before understanding your scope is guessing — or selling you a template. As a planning anchor: a focused build sprint typically runs from around CHF 4,000 to 15,000+, and a multi-system integration goes up from there.
The useful question isn't "what's the price." It's "what drives the price," because once you understand the cost drivers you can shape scope to fit a budget instead of being surprised by it.
What actually drives cost
Scope and edge cases. The happy path is cheap. The cost lives in the exceptions — the customer type that's priced differently, the order that splits across warehouses, the discount that can't stack. A tight, written scope is the single biggest lever on cost. A vague one is the single biggest cause of overrun.
Integrations. Every external system you touch — an ERP, a WMS, an accounting tool, a payment provider — multiplies the work, because each one has its own auth, rate limits, data model, and failure modes. One system is a build. Three systems talking to each other is a different category of project, which is exactly why Shopify ERP integration is priced and planned as its own discipline.
Data model. If the app just reads and writes Shopify objects, the model is given. The moment you need your own database — for state Shopify doesn't store, for reporting, for a B2B catalogue — you're designing schema, migrations, and sync, and that's real engineering time.
Shopify Functions and checkout logic. Server-side logic for discounts, shipping, payment customization, or bundle validation is powerful and fast, but it's specialized work with its own constraints. It often replaces a paid app and removes storefront scripts, which is why it can pay for itself — more on the mechanics in Shopify Functions explained for founders.
Maintenance. This is the cost most cheap quotes ignore entirely. Shopify ships API versions on a schedule, dependencies need patching, and webhooks need monitoring. Budget for it up front or pay for it later as emergencies.
Custom vs public app cost at scale
The comparison that matters isn't build cost versus a monthly subscription today — it's total cost at the volume you're aiming for.
Many public apps charge per order or per usage. At low volume that's cheap and correct. At high volume it becomes a percentage tax on every order, forever. A custom build has a higher upfront cost and a roughly flat run cost, so the two lines cross somewhere. Find out where, using your projected volume rather than today's. If you're already feeling the per-app tax, the pattern is laid out in the hidden cost of too many plugins and again in why your store doesn't need more apps.
The decision is rarely all-or-nothing. The strongest setups keep public apps for commodity features and build a thin custom layer only where generic tools force a compromise.
How to reduce the cost
You don't reduce cost by finding a cheaper developer. You reduce it by scoping better.
- Start with an audit or a small sprint, not a big build. Diagnose the actual problem before committing to a solution. Often the fix is removing apps and repairing a webhook, not new software. A Shopify CRO and tech audit ranks the expensive problems first.
- Cut scope to the differentiating part. Build the one workflow that's genuinely yours; buy the rest. Necessary-but-common features should stay as public apps.
- Phase it. Ship the core that pays for itself, then extend. A working slice in production beats a perfect spec that's still being argued over.
- Write the spec down. Ambiguity is the most expensive thing in any build. Every undocumented edge case is a future change request.
What cheap quotes hide
A quote far below the ranges above is usually omitting the parts that make software survive contact with production:
- No error handling or retries — so a single API hiccup silently drops an order.
- No monitoring — so failures are discovered by customers, not by you.
- No idempotency — so a retried webhook double-charges or double-fulfils.
- No handover or documentation — so you can't maintain or move what you paid for.
- A junior building against one happy-path assumption — fine in a demo, fragile at volume.
These aren't gold-plating. They're the difference between an app and a liability. The unglamorous infrastructure — queues, idempotency, logging — is most of what separates a CHF 4,000 demo from a CHF 12,000 system you can trust with revenue.
Getting a number you can trust
The fastest path to a real figure is a real scope. A paid Shopify audit turns "we think we need an app" into a written diagnosis with ranked problems and a scoped recommendation — so the cost is predictable before anyone commits. If you're earlier than that, run the free Shopify audit first to see whether a build is even the right move.