Softomate Solutions is a London-based software development studio. The question we are asked more than any other during initial conversations is: how much does a web application cost? The honest answer requires understanding what drives cost, what is frequently omitted from proposals, and what the difference is between a well-scoped £40,000 project and a poorly scoped £15,000 project that ultimately costs £80,000 to finish. This guide gives you the full picture.
Web application development costs in the UK follow a reasonably predictable pattern when you understand what drives complexity. The following ranges reflect work commissioned from competent UK-based development teams in 2024, including design, development, testing, and project management - but excluding ongoing hosting, maintenance, and third-party service fees.
Internal Tools and Simple Portals: £15,000 to £40,000
This tier covers focused applications with a limited number of user roles, straightforward data models, and minimal integration requirements. Examples: a staff intranet with a directory and document library; an internal job management system for a trades business; a simple customer portal showing order status and invoice history; a basic booking system for a single-location service business.
At this tier, you typically get three to five core workflows, one or two user roles, basic reporting, email notifications, and a responsive design. The team is usually two to three people: a backend developer, a frontend developer, and a part-time project lead. Timeline runs eight to sixteen weeks for a focused brief.
Customer-Facing Portals and Business Applications: £30,000 to £80,000
This tier covers applications with multiple user roles, richer functionality, payment processing, integrations with third-party services, and a need for polished UX design because external customers will use the product daily. Examples: a multi-tenant customer portal for a professional services firm; a property management platform; a subscription service with member accounts; a complex booking system with resource management and automated communications.
Teams at this tier grow to four to six people and typically include a UX designer, a backend lead, a frontend developer, a QA engineer, and a project manager. Timeline is four to nine months. The increase in cost comes from design complexity, a more intricate data model, integration with payment providers (Stripe, GoCardless), CRM systems, and external APIs, plus the additional testing overhead when external users are involved.
SaaS Products and Multi-Tenant Platforms: £50,000 to £200,000+
This tier covers software that is the product. Multi-tenancy (multiple client organisations sharing infrastructure but with complete data isolation), subscription billing, onboarding flows, a self-service admin panel, usage analytics, API access for customers, documentation, and the engineering discipline to support growth without architectural rewrites. Examples: a B2B SaaS platform for a vertical market; a marketplace connecting buyers and sellers; a data reporting platform with white-label capability.
At this tier you are building a product, not a project. The engineering decisions made in months one and two determine whether the platform can scale to 10 customers or 10,000. Multi-tenancy requires careful database architecture (row-level security or separate schemas), infrastructure that can scale horizontally, a robust deployment pipeline, and monitoring that gives you visibility into per-tenant performance. Timeline for a serious v1 is six to fourteen months.
Enterprise Applications: £200,000+
Complex integrations with legacy enterprise systems, compliance with regulated industry standards, sophisticated access control models, custom reporting engines, and the need for formal change management and documented audit trails push projects into enterprise territory. Timeline is twelve months or more. These projects require formal project governance, regular steering committee sign-offs, and a dedicated delivery team.
Headline scope (feature count) is the most obvious cost driver, but several others are equally significant:
Authentication complexity. Single login with email and password is straightforward. Adding MFA, social login, single sign-on (SSO) via SAML or OIDC for enterprise clients, or fine-grained role-based access control adds significant engineering time. SSO integration alone commonly adds £5,000 to £15,000 to a project budget because of the testing overhead across identity providers.
Third-party integrations. Every integration with an external system adds risk and cost. APIs that are well-documented, reliable, and consistent (Stripe, Twilio, SendGrid, Xero) are cheaper to integrate than those with incomplete documentation, inconsistent behaviour, or rate limiting that requires sophisticated retry logic. Legacy system integrations via SOAP or flat-file exchange are substantially more expensive than modern REST APIs.
Real-time features. Live notifications, real-time collaboration, live dashboards, and chat functionality require WebSocket infrastructure, connection management, and careful handling of reconnection and state synchronisation. This is materially more complex than request-response application patterns. Budget an additional £5,000 to £20,000 for real-time features depending on sophistication.
File handling. Allowing users to upload documents or images requires virus scanning, format validation, size limits, secure storage, and often image processing (resizing, format conversion). Large file handling, PDF generation, and document preview functionality each add measurable development time.
Reporting and analytics. Basic data tables are inexpensive to build. Sophisticated reporting with filters, aggregations, date range comparisons, export to Excel/PDF, and saved report configurations is a substantial engineering undertaking. Budget-grade reporting takes a week. Production-quality reporting that business users will actually use daily takes six to ten weeks.
Mobile requirements. A responsive web application that works well on mobile browsers is a baseline expectation, not a premium feature. A native mobile application (iOS and Android) is a separate product with a separate budget - typically £30,000 to £80,000 for a focused mobile app development project on top of the web application budget.
Compliance and security requirements. FCA regulation, NHS Data Security Standards, ISO 27001 alignment, PCI DSS for payment processing, and WCAG 2.1 AA accessibility all add engineering and documentation overhead. These are legitimate costs that should appear explicitly in proposals rather than being absorbed invisibly (which usually means they have been ignored).
The build cost is the number in the proposal. The total cost of ownership over three years often doubles it. Hidden cost categories that UK businesses consistently underestimate:
Infrastructure and hosting. A properly architected web application on AWS, GCP, or Azure with appropriate redundancy, monitoring, and security configuration costs £300 to £2,000 per month depending on scale. This includes the application servers, database (managed PostgreSQL on AWS RDS typically starts at £100 to £300/month), file storage, CDN, SSL management, and outbound email infrastructure. Many proposals include no hosting budget at all, leaving clients surprised by their first cloud invoice.
Third-party API costs. Payment processing (Stripe charges 1.4% to 2.9% per transaction plus 30p), SMS notifications (Twilio: approximately 4p per UK SMS), outbound email (SendGrid: free up to 100 emails/day, £50 to £200/month thereafter at scale), mapping (Google Maps API: £5 per 1,000 requests above the free tier), and identity verification services (£0.50 to £2.50 per verification) all compound with usage. Model these against your expected user volumes before signing development contracts.
Monitoring and observability tools. Error tracking (Sentry: £20 to £80/month), uptime monitoring (Pingdom, BetterUptime), application performance monitoring (Datadog: £100 to £500/month at scale), and log aggregation are essential for running a production application reliably. Budget £100 to £500 per month for a reasonable observability stack.
Ongoing development and bug fixes. Software requires maintenance. Libraries need updating to address security vulnerabilities. Browsers change behaviour. Users discover edge cases. Feature requests arrive from real usage. A maintenance retainer with a UK development team costs £1,500 to £4,000 per month for a responsive relationship. Without a retainer, you are in the queue when something breaks.
Content, onboarding, and operational overhead. Someone needs to write user documentation, create onboarding flows, handle user queries, and manage system configuration. These are real costs even if they do not appear on a development invoice.
Cost reduction in bespoke software development is about scope discipline, not penny-pinching on engineering quality. Cutting corners on security, testing, or architecture creates technical debt that costs more to fix than it saved to incur.
Legitimate ways to reduce cost without reducing quality:
Phase your delivery. A focused phase one covering the three to five features that deliver the most value, followed by phases two and three informed by real user feedback, produces better software at the same total budget. Every project that tries to build everything in phase one either runs over budget or compromises quality on the features added late under deadline pressure.
Use open source components intelligently. Authentication (Laravel Sanctum, Devise, Auth0), file upload handling, PDF generation, and email templating are all solved problems with mature open source solutions. Paying a developer to build these from scratch adds cost without adding value. Good developers know when to reach for a library and when to build custom.
Start with server-rendered pages rather than a full SPA. A server-rendered application (Next.js, Laravel with Inertia, Django with HTMX) delivers a simpler architecture, better SEO, and faster initial page loads for a fraction of the infrastructure complexity of a full single-page application. Reserve complex client-side interactivity for the parts of the application where it genuinely improves the user experience.
Resist feature creep during development. Every 'small addition' during development costs two to five times what the same feature would cost to add in a later phase, because it requires rethinking in-progress work. Document requests, evaluate them at milestone reviews, and add them to a future phase backlog rather than expanding scope mid-sprint.
When you receive multiple proposals for the same project, comparison is only meaningful if the proposals cover the same scope. Before comparing headline figures, verify that each proposal includes:
Discovery and architecture. A proposal that jumps straight to development without a defined discovery phase is proposing to build first and understand later. Discovery should be a named, costed line item. Some agencies charge separately for a pre-build discovery engagement (£3,000 to £8,000 for two to four weeks), which is reasonable and often worth paying before committing to the full build.
Design.UX research, wireframing, and interactive prototype are development inputs, not deliverables. A proposal covering only visual design without user experience research will produce something that looks acceptable and works badly. Meaningful design work for a customer-facing application takes four to eight weeks.
Testing. Automated test coverage (unit tests, integration tests), manual QA, cross-browser and device testing, and performance testing should all appear in the proposal. A fixed-price proposal with no testing budget is proposing to deliver untested software.
Infrastructure setup. Who provisions the production environment? Who configures monitoring and alerting? Who handles the deployment pipeline? These are engineering tasks that take time and should appear in the budget.
Post-launch support terms. What happens when a bug appears on day two of production? Is there a warranty period? What is the response time commitment? A proposal that is silent on post-launch support is a proposal from a team that has not thought about your long-term relationship.
A proposal significantly below the market rate for the stated scope is not a bargain. It is a signal that something is wrong:
Proposals vary because scope definitions vary. Two agencies responding to the same brief can interpret 'customer portal' to mean anything from a three-page static site with a login button to a full multi-tenant application with role-based access and API integrations. The only way to get comparable proposals is to write a detailed brief that defines the specific features, user roles, data to be handled, integrations required, and non-functional requirements including security, performance, and accessibility standards. Vague briefs produce incomparable proposals.
A thorough proposal should include: a restatement of the requirement demonstrating understanding; the proposed technical approach and architecture; a team structure naming the people who will actually do the work; a project timeline with milestones; a detailed cost breakdown by phase or by role; a payment schedule; a definition of the acceptance criteria for each deliverable; the change request process; and post-launch support terms. Proposals missing more than one of these elements require clarification before you can make an informed decision.
Fixed-price contracts provide budget certainty but only when requirements are thoroughly defined. If requirements are vague, agencies must price in risk and you will pay a premium for certainty that is not really certain. Time-and-materials contracts provide flexibility and transparency but require discipline on both sides to manage scope and budget. For well-defined projects with stable requirements, fixed price is reasonable. For complex or novel applications where requirements will evolve, time-and-materials with a defined monthly budget gives you flexibility without losing control. Milestone-based fixed-price with defined acceptance criteria for each phase is often the best of both approaches.
A focused internal tool takes eight to sixteen weeks. A customer-facing portal with moderate complexity takes four to six months. A SaaS product with multi-tenancy, subscription billing, and a polished onboarding flow takes six to fourteen months. These timelines assume a well-defined brief, adequate client availability for feedback, and a competent development team. Timeline overruns are most commonly caused by unclear requirements, slow client sign-off cycles, and scope expansion during development rather than developer underperformance.
Yes, and you should be cautious of agencies that will not do this. A paid discovery phase (typically £3,000 to £8,000 for a two-to-four-week engagement) produces a detailed specification, architecture recommendation, and realistic cost estimate before you commit to the full build budget. It protects you from committing £80,000 to a project that was inadequately understood, and it gives you a document you can use to seek alternative proposals if you choose not to proceed with that agency. Agencies that resist a paid discovery phase in favour of proceeding directly to a build are optimising for their contract value, not for your project's success.
More from the blog
Let us help
Talk to our London-based team about how we can build the AI software, automation, or bespoke development tailored to your needs.
Softomate Solutions Limited
Deen Dayal Yadav
Online
