"How much does an app cost?" without context is like asking how much a house costs. The answer depends on the floor area, the materials and the plot, yet every house estimate is made of the same line items. Mobile apps work the same way: the price is the sum of hours from specific roles at specific stages, plus external costs that people tend to forget.
Below we break the budget down: who works on the project, how many hours each stage usually takes, where the costs outside development hide, and how to read a proposal so you can compare studios fairly. All hour ranges reflect our studio's experience, not an industry standard.
The budget formula
Simplified, a budget looks like this: Σ (role hours × role rate) + uncertainty contingency + external costs + post-release support. Each term moves the total independently, and most differences between proposals come from one team including contingency and support while another does not.
Estimate accuracy depends on the project stage. Steve McConnell described the "cone of uncertainty": at the idea stage an estimate can be off by several times, while after detailed requirements and design the range narrows to tens of percent. That is why an honest early proposal always gives a range, not a single number.
Project roles and their share
A mid-complexity app typically involves seven roles. Their load is uneven: the designer is busiest at the start, QA towards release, while the project manager and tech lead are present throughout.
| Role | Responsibilities | Share of project hours (studio experience) |
|---|---|---|
| Business analyst | Requirements, user flows, acceptance criteria | 5–8% |
| UX/UI designer | Prototypes, design system, state mockups | 10–15% |
| Mobile developers | iOS, Android or a cross-platform codebase | 35–45% |
| Backend developer | API, database, integrations, admin panel | 15–25% |
| QA engineer | Test plan, manual and automated testing | 12–18% |
| Project manager | Planning, communication, risk | 8–12% |
| DevOps / tech lead | CI/CD, infrastructure, architecture, code review | 3–7% |
If QA accounts for 3–5% of a proposal, or is missing altogether, testing is either done by developers (slower and worse) or it lands on you after release in the form of user reviews.
Stages and typical hours
Here is how hours break down for an MVP with 15–25 screens, authentication, a profile, a catalogue or feed, push notifications, one payment integration and a simple admin panel. Ranges are for two platforms.
| Stage | Duration | Hours | Output |
|---|---|---|---|
| Discovery | 2–4 weeks | 80–200 | Spec, user flows, prototype, refined estimate |
| UX/UI design | 3–6 weeks | 120–300 | Mockups for every screen and state, UI kit |
| Backend and admin | in parallel | 250–600 | API, database, integrations, control panel |
| Mobile development | 8–16 weeks | 600–1,400 (cross-platform) / 1,000–2,400 (two native) | iOS and Android builds |
| QA and stabilisation | in parallel + 2–3 weeks | 200–450 | Test cases, regression, reports |
| Release | 1–3 weeks | 30–80 | Publishing, store listings, monitoring |
In total that is roughly 1,300–3,000 hours for a cross-platform MVP and 1,700–4,000 for two native apps. The spread is wide precisely because of the factors below.
What moves the hours most
- Offline mode and sync. Can double a module's scope because of data conflicts and action queues.
- Real time. Chat, courier tracking or collaborative editing need separate infrastructure (WebSockets, queues) and harder testing.
- Number of roles. A customer app plus a separate provider app is effectively two products sharing a backend.
- Third-party integrations. Especially legacy ERPs or banking APIs without a sandbox: these estimates always carry contingency.
- Custom animation and components. Every "wow effect" means design, development on two platforms and testing across devices.
Native vs cross-platform
Flutter or React Native let you write most of the code once. In our experience the saving on the mobile part is 25–40% compared with two native teams, not 50%: platform integrations, testing on both operating systems and publishing still happen twice. Design, backend, analytics and management barely depend on the technology choice. What to include in version one without inflating the budget is covered in our article on the mobile app MVP.
Costs outside development
These items never appear in a studio's proposal, but they are in your P&L.
| Item | Amount or order of magnitude | Note |
|---|---|---|
| Apple Developer Program | USD 99 per year | Apple; organisations need a D-U-N-S number |
| Google Play Console | USD 25 one-off | Google; new personal accounts must run a closed test with 12 testers × 14 days |
| Store commission | 15–30% of digital sales | 15% under Apple’s Small Business Program and on the first USD 1M of revenue on Google Play |
| Hosting and backend services | tens to hundreds of USD per month for an MVP | Grows with audience, media and real-time features |
| Third-party SDKs | 0 to hundreds of USD per month | Maps, SMS verification, chat, analytics, crash reporting tiers |
| Legal | one-off | Privacy policy, terms of use, data processing agreements |
| Marketing and ASO | separate budget | Without traffic, even a perfect app goes unnoticed |
SMS verification is a frequent surprise: in some countries a single message costs several cents, and SMS pumping attacks can run up a bill of thousands of dollars overnight. Put rate limits and protection into the budget from day one.
Post-release support
Release is the halfway point. New iOS and Android versions ship every year, Google Play raises its target API requirement (since 31 August 2026 it is Android 16, API 36, per Android Developers), and third-party SDKs update and drop old versions. A common industry rule of thumb, which our experience confirms: budget 15–25% of the initial development cost per year for maintenance and small improvements. That excludes new features.
- Adapting to new OS versions and devices: 40–120 hours a year.
- Dependency updates and security patches: regularly, every month.
- Crash monitoring and critical fixes: an agreed response SLA.
- Infrastructure: backups, certificates, scaling.
Engagement models: fixed price, time & materials, dedicated team
| Model | Best when | Risk for the client |
|---|---|---|
| Fixed price | Clear spec, small scope, little uncertainty | The studio builds a 20–40% buffer into the price; any change goes through a change request |
| Time & materials | The product evolves, requirements are refined | Needs budget control: weekly reports, an hours cap per sprint |
| Dedicated team | Long-term product, steady backlog | You pay for team capacity; a strong product owner is essential |
A hybrid works well in practice: discovery at a fixed price, then development on T&M with a cap per stage. You pay for certainty when it exists and do not overpay for someone else's buffer.
How to compare studio proposals
- Ask for a breakdown by role and stage. A single total cannot be compared.
- Check that QA, PM and discovery are included. Leaving them out only makes a proposal cheaper on paper.
- Clarify assumptions. A good estimate lists them: "we assume the payment provider offers a sandbox", "the admin panel uses an off-the-shelf framework".
- Compare the definition of done. Does it include store publishing, analytics, crash reporting, API documentation?
- Ask about the warranty period. One to three months of free defect fixing after release is normal practice.
- Weigh rate against productivity. A cheaper hour multiplied by twice the hours is more expensive.
Common budgeting mistakes
- No contingency. Reserve 15–25% for uncertainty even with a detailed spec.
- The whole budget on version one. If no money is left for iterations based on user data, the MVP loses its purpose.
- Cutting state design and QA. These are the first items to be cut and the ones that decide your store rating.
- Ignoring store commission. For a subscription product, 15–30% is part of unit economics, not a rounding error.
- Comparing rates instead of outcomes. Ask about timelines, warranties and examples of similar work.
Action plan
- Define the goals and scope of version one, including a Won't list.
- Commission discovery or prepare a spec with non-functional requirements.
- Get 2–3 estimates broken down by role, stage and assumptions.
- Add external costs: developer accounts, hosting, SDKs, legal, marketing.
- Reserve 15–25% contingency and 15–25% per year for support.
- Keep 20–30% of the total budget for post-release iterations.
Sources: Apple Developer Program: enrollment; Google Play Console: registration; Apple App Store Small Business Program; Google Play: service fees; Play Console: testing requirements; Android Developers: target API level; Construx: The Cone of Uncertainty.



