What a Mobile Banking App Really Costs: Build, Run and Certify Over Three Years
A build quote usually covers less than half of what a banking app costs over its first three years. We cost one hypothetical bank three ways (custom, low-code and white-label) with onshore and offshore teams.
Why the build quote covers less than half the mobile banking app development cost
Boards usually approve a mobile banking app based on the build quote. In the model below, that quote covers less than half of what the bank spends over three years. The rest goes on security testing that has to be repeated, accessibility and app-store work that returns with every operating-system cycle, hosting, and a release team that never really winds down. So the honest answer on mobile banking app development cost is a three-year number, not a launch number.
We take one hypothetical bank down three routes: a custom build, a low-code digital banking platform and a white-label app. Each is costed from discovery to the end of year three. Every figure is a typical 2026 market range in USD. Actual numbers vary by contract, region and scope, and local pricing can differ sharply, so confirm them with vendors and integrators before anything goes into a budget.
Our position: for most regional banks without a large in-house engineering function, a low-code platform gives the best balance of cost, control and risk. Custom only pays off when the app itself is central to the bank's strategy and the bank will fund a permanent team. White-label is the cheapest way to launch, and the bank pays for that later in lost flexibility.
The hypothetical bank and the app it wants to ship
Hypothetical example. Harbor Regional Bank is a fictitious bank with 300,000 retail customers and 12,000 small-business customers in one country. Its core banking system comes from an established vendor and has a partial API layer. The target is 120,000 monthly active mobile users by the end of year three.
The scope is identical on all three routes:
- Native-quality iOS and Android apps, published under the bank's own developer accounts
- Digital onboarding with document capture, liveness checks and eKYC
- Accounts, transactions, statements and downloadable documents
- Domestic transfers, instant payments where the local scheme supports them, and saved payees
- Card controls: freeze, limits, PIN view, and travel or merchant-category blocks
- Biometric login, device binding and step-up authentication for high-risk actions
- Light SME features: multiple users and two-person payment approval
- Push notifications and handoff from in-app chat to the contact center
Assumptions: the three-year clock starts at contract signature. After launch there are four major and roughly 20 minor releases a year. The bank reuses its existing fraud engine and identity platform. Core replacement, card-processor changes and marketing are out of scope.
One variable outweighs the rest. If the core can't expose modern APIs, add something like $300,000–$1,000,000 of middleware and integration work to whichever route you choose. Price that gap first. No app vendor can make it disappear.
Custom build: line by line for Harbor
Most banks build custom with an external delivery partner, or a partner working alongside in-house staff. A realistic build team is about 14 people: a delivery lead, a product owner, a designer, two iOS and two Android engineers, three backend and integration engineers, two QA engineers, a DevOps engineer and a security engineer. Discovery and design take around 10 weeks, then roughly 10 months to production launch.
| Cost line | Onshore team | Blended offshore team |
|---|---|---|
| Discovery, service design, architecture (one-time) | $180k–$300k | $100k–$180k |
| Build, integration and QA (one-time) | $2.2M–$3.2M | $1.1M–$1.7M |
| Pre-launch penetration test and remediation retest | $40k–$80k | $40k–$80k |
| Accessibility audit and app-store readiness | $25k–$60k | $25k–$60k |
| App shielding / runtime protection licence (per year) | $40k–$100k | $40k–$100k |
| Cloud hosting, environments, observability (per year) | $150k–$300k | $150k–$300k |
| Run and release team, 7–9 people (per year) | $1.1M–$1.6M | $0.5M–$0.8M |
| Recurring pen tests and accessibility re-audits (per year) | $50k–$120k | $50k–$120k |
Look at which lines stay put when the team moves offshore. Security testing, tooling and hosting cost about the same wherever the engineers sit. Labor is the only large line that shrinks.
A cross-platform framework such as Flutter or React Native can trim the mobile engineering line, often by around a fifth to a quarter. It does nothing for integration, QA or security, so the saving on the whole build is closer to 10%. The bigger planning error is budgeting run costs at 15–20% of the build each year. That rule of thumb comes from internal enterprise software. A banking app has to absorb annual iOS and Android releases, payment-scheme changes, new fraud patterns and regulator findings. In Harbor's model, run costs come to roughly half the build cost every year, and that's the figure we'd plan with.
Low-code digital banking platform: paying for a head start
Digital banking platforms supply a pre-built front end, journey orchestration and a library of banking components, configured through low-code tooling. Backbase, Temenos Infinity and Infosys Finacle's digital engagement suite sell internationally. Q2 and Alkami sell mainly in North America. A systems integrator or the vendor's own services arm usually handles implementation.
Typical 2026 ranges for a bank Harbor's size:
- Platform licence: $350k–$800k a year, priced by tiers of registered or active digital users and by module (retail, SME, onboarding). Billing usually starts at signature, not launch.
- Implementation: $900k–$1.8M with an onshore integrator, or $500k–$1.0M through an offshore delivery center. Expect seven to nine months to launch.
- Hosting: often bundled when the platform runs as vendor-hosted SaaS. Self-hosting in your own cloud tenancy runs $100k–$200k a year.
- Run team: 3–5 people configuring journeys and integrations, at $500k–$800k a year onshore or $250k–$400k blended.
- Major version upgrade: $150k–$300k, once in three years. Vendors call it optional. It isn't, because support for old versions runs out.
This is where buyers get caught. Over three years, low-code isn't dramatically cheaper than custom. What it buys is a launch roughly four to five months sooner, a smaller permanent team, and a vendor that absorbs OS updates and most component-level accessibility fixes. The weakness: the licence is a fixed cost that offshore labor can't reduce, and annual escalators, commonly in the low-to-mid single digits, compound.
White-label mobile banking app pricing and what it leaves out
White-label apps come either from core banking vendors bundling a front end with their core or from specialist app providers. The bank configures branding, product rules and feature toggles but doesn't touch the underlying code. For Harbor, white-label mobile banking app pricing typically looks like this:
- Setup and configuration fee: $150k–$400k
- Bank-side integration to the core, card processor and identity systems: $150k–$400k, depending on whether the vendor already has a connector for your core
- Subscription: $300k–$650k a year, tiered by active users. Some vendors price per active user per month instead; at Harbor's target volume, roughly $0.20–$0.45 per user lands in the same band
- Change requests: $100k–$250k a year at vendor day rates
- Internal team: a product owner plus one or two analysts or testers, $250k–$400k a year onshore
Launch takes four to five months. The vendor hosts the app and pushes platform updates on its own schedule, which saves money and costs you control. Anything off the vendor's roadmap either waits or becomes a change request. Read the subscription definitions line by line. Some contracts count an "active user" as anyone who logs in once a quarter. Others count every registered user, and on that basis Harbor's bill could roughly double while engagement stays flat.
Offshore labor barely matters here because most of the spend is subscription. Moving bank-side integration and testing offshore typically trims the three-year total by only 10–15%.
Penetration testing, app-store rules and accessibility: the lines that keep coming back
These are the costs proposals most often leave out, and they recur on every route.
Penetration testing and app protection
A credible mobile pen test covers the app binary, the APIs behind it and the authentication flows, usually against the OWASP Mobile Application Security Verification Standard (MASVS). Budget $40k–$80k for a full-scope test plus a retest after fixes. Repeat it at least annually and after any major release touching authentication or payments. Supervisors in many markets expect regular testing as part of their technology-risk rules: in the EU under DORA, in Singapore under the MAS technology risk management guidelines, and in India under the RBI's IT governance directions, among others. Even when a white-label vendor shares its own test reports, regulators generally expect the bank to commission independent testing of its own deployment.
App-store compliance
Developer account fees are trivial. The work isn't. Apple's review guidelines expect apps in regulated fields such as banking to be submitted by the institution offering the service, so the bank needs its own verified organization accounts on both the App Store and Google Play. Apple's privacy labels and Google Play's data safety and financial-features declarations need updating whenever data flows change. Plan for two to four weeks of effort a year, plus review delays around launch.
Accessibility
WCAG 2.1 or 2.2 at level AA is the de facto benchmark, but its legal force varies by country. The European Accessibility Act has applied to consumer banking services in the EU since June 2025. In the US, accessibility claims are brought under disability discrimination law. The UK, Canada, Australia, India and others have their own equality or disability statutes. An expert audit typically costs $15k–$40k, with annual re-audits of $10k–$25k. Ask platform and white-label vendors for a current conformance report, not a marketing claim. On this and the other regulatory points, get advice from local counsel.
Three-year totals side by side, onshore versus offshore
These totals combine the ranges above over 36 months from signature. Run costs start at launch, so a faster route also starts paying them sooner. This is a hypothetical worked example for comparing routes, not a quote.
| Route (Harbor, hypothetical) | Time to launch | Three-year total, onshore | Three-year total, blended offshore | Share that is recurring |
|---|---|---|---|---|
| Custom build | 12–13 months | $5.2M–$7.8M | $3.0M–$4.4M | About 50% |
| Low-code platform | 7–9 months | $4.3M–$6.4M | $3.2M–$4.7M | About 70% |
| White-label app | 4–5 months | $2.6M–$4.3M | $2.3M–$3.8M | About 85% |
On rates, typical 2026 blended engineering rates sit around $110–$200 an hour in the US, UK, Western Europe, the Nordics and Australia, and Gulf-based onshore teams often price in a similar band. Nearshore teams in Central and Eastern Europe or Latin America commonly run $45–$95, and teams in India, Vietnam and the Philippines $25–$55. "Offshore" is relative. A bank in Mumbai or Manila already pays the lower rate, so the offshore column is its baseline, and the platform licence becomes a much bigger share of its total.
Offshore delivery has costs the hourly rate hides. The EBA outsourcing guidelines, DORA's register of third-party arrangements and data-residency rules in markets such as Saudi Arabia, Indonesia and India add vendor oversight, audit rights and sometimes in-country hosting. Set aside 5–10% of offshore labor spend for governance.
The cost to build a banking app is only the opening figure. The recurring share tells you more about year four. A white-label deal that's 85% subscription keeps costing roughly the same indefinitely. A custom build that's been paid for can be scaled down.
Red flags in proposals and a decision rule for choosing a route
Send a proposal back before it reaches the board if it shows any of these:
- A build or setup quote with no run-cost estimate. That alone hides most of the digital banking app maintenance cost.
- Licence escalators above 5% a year, or no cap at all.
- User-based pricing tied to registered rather than active users, or no definition of "active".
- One pen test at launch and nothing after.
- "Accessibility-ready" claims with no conformance report or audit evidence.
- The app published under the vendor's developer account instead of the bank's.
- No source-code escrow, no data-export terms or no stated exit fee. In our model, leaving a white-label vendor could cost $300k–$800k once rebuilding and re-onboarding users is included.
- Change requests at open-ended day rates, with no annual pool or cap.
The decision rule we'd apply at a bank like Harbor:
- Need to be live within six months, and the app is a hygiene factor rather than a differentiator? Take white-label, and negotiate exit terms before you sign.
- Want real control over journeys but can't permanently staff more than five engineers? Choose a low-code platform. For most regional banks, this is the right answer.
- Choose custom only if you can commit a run team of seven or more for at least five years, and the board can name the features competitors won't be able to copy.
Next step: send every shortlisted vendor the scope list above and require a 36-month cost schedule broken out by year. Missing lines and buried escalators are much easier to spot when every bidder fills in the same rows.
Quick answers
How much does it cost to build a mobile banking app for a mid-sized bank?
For a regional bank with full retail features and core integration, a custom build typically costs about $1.3M–$3.6M before launch, depending largely on where the team is based. Over three years, once security testing, hosting and a release team are added, the total usually lands between $3M and $8M. These are typical 2026 ranges that vary by contract and region, so confirm them with vendors.
Is a white-label banking app always the cheapest option?
It's usually the cheapest over the first three years and the fastest to launch, often in four to five months. But 80–90% of the spend is recurring subscription, so the gap narrows after year three, and leaving the vendor can be expensive. Model at least five years and price the exit before treating white-label as the low-cost route.
How often should a banking app be penetration tested?
At minimum, test fully before launch, then at least annually and after any major release that changes authentication, payments or onboarding. Many supervisors expect regular independent testing under their technology-risk rules, even when a vendor supplies its own reports. Confirm the exact requirements in your jurisdiction with your compliance team or a licensed adviser.
Does using Flutter or React Native make a banking app much cheaper?
Less than most buyers expect. A shared codebase can reduce mobile engineering effort meaningfully, but integration, QA, security testing and hosting stay the same, so the saving on the total build is typically around 10%. It's worth considering for maintainability, but it won't change which delivery route is right for you.
Written by the Trustdeskapp editorial team. Prices, features and rules change often - confirm details with the vendor or a licensed professional before you decide.