Fajar“How much does a custom web app cost?” is usually the first question a founder asks, and a single number is the least useful answer. There is no published market price for custom software. What we can give you is our own rough estimate of where typical projects land: for the products startups and smaller companies usually commission, somewhere between $5,000 and $40,000 or more. Treat those figures as our working bands, not as market data. Rates, regions and teams vary widely, and other firms will quote differently. The useful part is knowing which decisions move a project along the range. There are about eight of them, and several tend to be settled casually in the first meeting.
Three Rough Bands
These bands are how we think about size when someone describes a product to us. They are estimates of typical ranges, and a written scope is the only thing that produces a real quote. At the small end, our Fixed-Price Projects typically start from $3,000 for a small, clearly defined piece of work, such as a single feature or integration. A complete application sits above that.
About $5,000 to $12,000: A Focused Tool
One user role, one core workflow, 5 to 10 screens, sign-in from an off-the-shelf provider, one or two third-party integrations (Stripe, a CRM, an email service), and a hosted admin tool instead of a custom back office. Examples: a client portal for a services business, an internal approval workflow, a booking tool, a calculator with checkout. This band assumes sensible defaults: a component library styled to your brand, and a proven stack with no exotic infrastructure. For us that usually means Next.js, PostgreSQL and a managed host.
About $12,000 to $25,000: A SaaS MVP
Two user roles (say, customers and staff), 10 to 20 screens, subscription billing with plan logic, role-based permissions, business logic that is specific to you (matching, scoring, scheduling, quoting), file handling, transactional email, and a modest custom admin view. Custom design goes on the screens that matter, with standard components everywhere else. A first version of a B2B SaaS product often fits here, and typically takes 8 to 12 weeks. Our guide to scoping an MVP explains how to keep it in this band.
About $25,000 to $40,000 and Up: Platforms and Regulated Products
Three or more user roles, marketplace mechanics or a multi-tenant architecture, real-time features, integrations with legacy or poorly documented systems, compliance work (audit logs, data residency, tooling for HIPAA or GDPR obligations), heavy reporting, or a serious data migration. The upper end is open. A two-sided marketplace with held payments and dispute flows, or a healthcare product, can pass $40,000 with nothing gold-plated. If that is your situation, our article on software that supports HIPAA compliance shows where the extra work comes from.
The Eight Decisions That Move the Number
- Number of user roles. This is the biggest multiplier and the one most often underestimated. Each role brings its own flows, permissions, empty states and edge cases. A second role adds far more than a second set of screens.
- Authentication and permissions. “Users log in” is cheap. “Organizations have teams, teams have roles, roles have granular permissions, and enterprise customers need single sign-on” is a project inside your project. Properly tested SAML single sign-on alone can add several thousand dollars.
- The quality of the systems you integrate with. Stripe takes a day or two because its API and documentation are excellent. A regional logistics provider with a SOAP API, no sandbox and a PDF for documentation can take two weeks. We estimate integrations by the quality of the other side, not by how many there are, and we ask to see the API documentation before quoting.
- Real-time requirements. Once anything must update live (chat, presence, collaborative editing, order tracking) you add WebSocket infrastructure, reconnection handling and a class of bugs that appear only under concurrency. Ask whether refreshing every 30 seconds would do. Often it would.
- Design maturity on arrival. Arriving with validated designs instead of a verbal description can change a quote by thousands of dollars. Design is not padding. Products without a design get designed during development, through rework, and that is the most expensive way to do it.
- Reporting and dashboards. “A simple dashboard” seldom is. Aggregations, filters, date ranges, exports and charts over large datasets need real back-end work and query tuning. A useful analytics section can cost as much as a core feature.
- Data migration. If your users and history live in spreadsheets or a legacy system, moving them cleanly, with validation, deduplication and a rollback plan, is a workstream of its own. It is also the item most often missing from a competing quote that looks low.
- Compliance and audit requirements. Audit trails, retention policies, access logging and tooling for deletion requests (the GDPR right to erasure, for example) are all buildable. They touch every feature, so they cost much less at the start than as a retrofit.
Where Teams Overspend
Much of a good scoping conversation is about what not to build yet.
- Custom admin panels before product-market fit. A hosted admin tool or a direct database UI covers the first year for most products. Internal screens that two staff members open twice a week should not take a large share of an MVP budget.
- Rebuilding commodity features: custom authentication, file storage, hand-rolled billing logic, in-house analytics. Each has a mature service that costs tens of dollars a month. Building your own buys liability and no differentiation. Build what makes you different, and buy the rest.
- Scale engineering for a product with no users: microservices, Kubernetes, multi-region databases. A well-built monolith on managed infrastructure carries most products a long way.
- A full design language, animation system and dark mode before anyone has confirmed that users want the product. Spend design money on the two or three screens where your value lands.
- Native mobile apps at launch, alongside the web app. That means two more codebases, store review cycles and release overhead, for reach a responsive web app already has. Unless the product depends on the camera or on notifications, mobile apps are a second-version decision.
The Costs Nobody Puts in the Proposal
A quote is not the cost of owning software. Plan for three more things.
Maintenance comes first. A common industry rule of thumb is 15 to 20% of the build cost per year for dependency updates, security patches, framework upgrades and small fixes. The exact figure depends on the product, but zero is the wrong number. An app that skips two years of updates often ends up paying for a partial rewrite.
Third-party services come second: hosting, database, email, authentication, monitoring, file storage and the APIs you integrate. For a typical small SaaS we would expect a few hundred dollars a month in total. Check each vendor’s current pricing page, because these bills grow with usage.
The third is your own time. A custom build needs a decision-maker who reviews progress every week and answers questions within a day or so. When the client goes quiet for three weeks, the team has to guess at requirements, and guessed requirements get rebuilt.
A Hypothetical Example
The following project is invented to show how the decisions work. It is not a client story, and the figures are illustrative.
Suppose a logistics company wants a customer portal. Clients log in, create shipment requests, upload documents, track status and receive email updates. Staff review requests and update statuses. Described as “a portal for our clients”, it sounds like the first band. The eight decisions say otherwise: two user roles, a status workflow with notifications at each stage, document uploads with validation, and an integration with a transport management system whose only documentation is an old email thread. A realistic estimate for that scope might be $18,000 to $24,000 over 10 to 12 weeks.
Now suppose the team cuts deliberately. Staff work from a hosted admin view instead of a custom dashboard. Notifications drop from six trigger points to the two that customers care about: request received and shipment delivered. The legacy integration moves to phase two and is replaced at launch by a CSV import that one staff member runs twice a day. The revised build might come to around $13,000 to $14,000 in about eight weeks. More important, the company would have real usage data before paying for the expensive integration. That data could easily cancel one planned feature and suggest another nobody had thought of.
The sequence is what matters: scope carefully, cut deliberately, and let production data decide where the next dollar goes.
How to Lower the Cost Without Lowering the Quality
First, cut scope and keep quality. A smaller product built properly beats a full product built badly, because the second one tends to get rebuilt, and then you pay twice.
Second, phase the delivery. Ship the core workflow first and let real usage reorder the backlog for phase two. It usually does.
Third, arrive with decisions made. A written scope, a few examples of products you like and a named decision-maker will do more for your budget than negotiating rates. It also helps to choose the right contract shape, and fixed price or a monthly team compares the options.
A warning in the other direction: if one quote for a multi-role platform with integrations and billing arrives at a fraction of the others, read it as a description of what you will receive. For a checklist of what a careful build should cover, the OWASP Application Security Verification Standard is a good public reference to hold any quote against.
If you want a number for your project instead of a range, our web development team will scope it with the decisions, bands and trade-offs laid out. Get a quote and bring whatever you have, even a rough sketch. We reply within 24 hours on working days.
