Milestone Reviews
Each milestone ends with a demo of working software, a written note covering what changed and what's next, and a budget check against the plan. Nothing is hidden.
How We Work
A five-stage, milestone-driven process from first enquiry to launch — typically 2 to 6 weeks, with working software and a written progress note at every step.
The Delivery Lifecycle
Five stages from first contact to launch and handover. Most small business sites run 2 to 6 weeks end to end.
Request and Discovery (1 to 2 days)
First contact is a no-obligation conversation to understand your business problem rather than jumping straight to a solution. We ask about your users, what success looks like, existing systems and constraints, budget range and realistic deadlines. Discovery is about understanding the problem deeply enough to propose the right solution, not the easiest one to build.
What you get
- Written summary of what was discussed
- Agreed problem statement
- Indication of feasibility and suitability
Proposal (2 to 3 days)
A written proposal covers scope, recommended approach and technology choices with reasons, assumptions and exclusions stated plainly. For larger builds, phasing options are provided so you can start small and expand later. The proposal includes a fixed cost and a milestone plan showing what gets delivered and when. Starting prices are published on our pricing page, so the figure in the proposal should not be a surprise. No jargon, no padding — everything is explained in plain English.
What you get
- Proposal document with scope and approach
- Milestone breakdown with deliverables
- Clear statement of what is and is not included
Refining Requirements (2 to 5 days)
Once the proposal is agreed, we turn it into something buildable: user stories with acceptance criteria, wireframes or clickable prototypes for key screens, data models and integration points, and non-functional requirements covering accessibility level, performance budgets, security and GDPR handling. Everything is prioritised in a backlog so we can adapt scope if needed without derailing the project.
What you get
- Agreed backlog of user stories with acceptance criteria
- Wireframes or prototypes for key screens
- Technical design notes covering data and integrations
- Non-functional requirements (accessibility, performance, security)
Build in Milestones (1 to 3 weeks)
Development is broken into short iterations, each ending in working software you can click through on a staging URL. Regular check-in calls review progress and unblock issues. A written progress note accompanies each milestone showing what changed, what's next, and budget burn against the plan. Change requests are handled openly — impact on cost and timeline is shown before anything is agreed. Continuous integration runs lint, type checks, tests and accessibility checks on every change, catching problems early.
What you get
- Working software each milestone on a staging environment
- Written progress notes with demos
- Updated backlog reflecting current priorities
Launch, Handover and Support (2 to 4 days)
Launch follows a pre-launch checklist: accessibility audit, performance pass, security review, analytics and consent wiring, SEO metadata, and backups configured. Deploy is controlled with rollback procedures in place. Handover includes documentation, a walkthrough session, and ownership transfer of source code and infrastructure accounts. Optional ongoing support and maintenance arrangements are discussed openly, but you are never locked in — everything is yours to take elsewhere if you choose.
What you get
- Live site or application on production infrastructure
- Handover documentation and admin walkthrough
- Source code and infrastructure ownership transferred
- Agreed support arrangement (if required)
Milestones, Not Mystery
Regular check-ins so you always know where the project stands.
Transparency
You see the staging environment at all times. You see the backlog. You see progress notes. You know where the budget stands. There are no surprises.
Adapt or Stop
Priorities change, budgets tighten, or requirements evolve. At any milestone boundary, we can re-scope or stop work — you are never locked into a contract that no longer fits.
Exit Anytime
Code and infrastructure are yours from the start. If you decide to take the project elsewhere or in-house, everything transfers cleanly without legal wrangling or ransom.
AI-Assisted and Agentic Delivery
Modern tooling to deliver more value for your budget.
AI Tooling Throughout
AI-assisted coding tools help with scaffolding, boilerplate generation, test creation, refactoring, documentation and code review assistance. This speeds up repetitive work without sacrificing quality.
Agentic Workflows
Agentic workflows handle multi-step tasks like schema migrations, bulk refactors, generating TypeScript types from API specs, and scaffolding i18n translations. Work that would previously burn billable hours is automated, freeing budget for design decisions and testing.
More Value, Less Typing
The result: more of your budget goes into requirements refinement, user experience design, and thorough testing rather than typing boilerplate that a machine can write faster and more accurately.
Human Accountability and Review
Every line of code is reviewed by a human before it ships. Nothing is merged without passing automated checks (linting, type checking, tests, accessibility scans) and manual code review. AI is a tool, not a replacement for judgment. The developer remains accountable for the output.
Data Privacy and Guardrails
AI tooling does not have access to client production data, secrets, API keys, or sensitive business logic. Confidentiality and GDPR obligations govern what can and cannot be shared with third-party services. If your project has strict data handling requirements, AI tooling is restricted or disabled accordingly.
Quality Gates
Seven automated and manual checks every project passes before it goes live.
Code Review
Every change is reviewed before merge. Code quality, logic correctness, test coverage and security implications are checked manually.
Automated Tests
Unit tests validate individual functions and components. End-to-end tests simulate real user journeys through the application. Tests run on every commit.
Type Checking
TypeScript catches type errors, null reference bugs and interface mismatches at compile time, reducing runtime errors dramatically.
Accessibility Checks
Automated accessibility scans check for WCAG 2.2 Level AA violations: missing alt text, insufficient contrast, keyboard traps, unlabelled form fields. Manual testing with screen readers validates real usability.
Performance Budgets
Core Web Vitals (LCP, INP, CLS) are measured on every build. Budgets are set for bundle size, image weight and request counts. Performance regressions are caught before they reach production.
Dependency Scanning
Automated vulnerability scanning checks for known CVEs in dependencies. Security updates are applied promptly and tested before deployment.
Staging Sign-Off
Nothing goes to production without sign-off on staging. You see and approve the working software before launch, not after.
Hosting and Environments
Cloud-first infrastructure with automated deployments and rollback.
Hosting is cloud-first on Microsoft Azure with separate development, staging and production environments. Containerised builds using Docker ensure consistency — what works in development works in production. Infrastructure as code (Bicep or Terraform) defines resources declaratively, making deployments reproducible and version-controlled.
CI/CD pipelines automate the build and deployment process: code is linted, type-checked and tested automatically; builds produce optimised production bundles; deployments push to staging first for review, then production after approval. Rollback procedures are built in — if a deployment introduces a regression, the previous version is restored within minutes.
Monitoring and logging track application health, error rates and performance metrics. Alerts notify on downtime, slow responses or spike in errors. Automated backups run daily with retention policies, and restore procedures are tested to confirm they work. The result: predictable, low-drama releases where problems are caught early and fixed fast.
Working Together
Clear communication, realistic expectations, and mutual respect.
We are UK-based and work in UK business hours (Monday to Friday, 9am–5pm GMT/BST). You have a single named point of contact — the developers building it — not an account manager acting as a relay. Communication happens on an agreed channel (email, Slack, Teams, or phone) with an agreed cadence for check-ins. Updates are written in plain English without unnecessary jargon, and we are honest when something is harder or cheaper than initially expected.
You own your code, your domain, and your hosting accounts from day one. There are no lock-in clauses, no proprietary platforms, and no ongoing fees required to keep your site running (other than standard hosting and domain costs). If you want to take the project in-house or to another developer, everything transfers cleanly with documentation and support during handover.
Mutual respect means realistic deadlines, honest communication when priorities conflict, and no surprises on either side. If we are unavailable (holiday, illness), you are informed in advance with a plan for urgent issues. If your requirements change, we discuss impact openly before committing to new work. If a deadline or a budget looks at risk, you hear about it as soon as we know, not at the next scheduled check-in.
Common Questions
Straight answers to the questions we get asked most about how projects run.
What if my requirements change mid-project?
Requirements do change — that is normal and expected. At each milestone boundary, we can re-prioritise the backlog or adjust scope. If a change is significant, we assess the impact on cost and timeline and present options before proceeding. You are never locked into a fixed scope if the business need has evolved.
How do I know progress is real and not just a status update?
Every milestone delivers working software on a staging URL that you can click through and test. Progress notes include links to the staging environment and describe what was built, what changed, and what is next. You see the actual application evolving, not a Gantt chart or a percentage-complete figure.
Can we start small and expand later?
Yes. For larger projects, we propose phasing options where the first phase delivers a working minimum viable product (MVP) or a specific high-value feature set. You launch with something useful and proven, then expand in later phases as budget allows and business priorities become clearer. This reduces risk and allows early feedback from real users.
Who owns the code and the accounts?
You do. Source code is stored in a Git repository that you own (GitHub, GitLab, Azure Repos). Hosting accounts, domain registrations, and third-party service accounts (analytics, payment gateways) are created in your name or transferred to you at handover. There are no lock-in clauses or proprietary platforms. If you decide to take the project elsewhere, everything transfers cleanly.
What happens after launch?
Post-launch support options are discussed openly but never mandatory. We offer ongoing maintenance retainers covering bug fixes, dependency updates, content changes, and small feature additions. Alternatively, you can handle maintenance in-house or engage another developer — the codebase is documented, and we provide a handover session to ensure continuity. You are never locked in.
How is AI used, and is my data safe?
AI tools assist with scaffolding, boilerplate, refactoring, test generation and documentation. Every line is reviewed by a human and tested before shipping. AI does not have access to production data, secrets, API keys, or sensitive client information. If your project has strict data handling requirements, AI tooling is restricted or disabled. GDPR and confidentiality obligations are always respected.
How long does it take to build a small business website?
Most small business sites launch 2 to 6 weeks after first contact. That breaks down as 1 to 2 days of discovery, 2 to 3 days to produce the proposal, 2 to 5 days refining requirements, 1 to 3 weeks of build in weekly milestones, and 2 to 4 days for launch and handover. The single biggest variable is how quickly you can get us your content — photos, service descriptions, opening hours. Projects that stall almost always stall there.
Should I use a small studio or a large agency?
A large agency gives you depth of bench: designers, copywriters, project managers and developers under one roof, which matters on a big multi-team build. You pay for that structure, and you usually talk to an account manager rather than the person doing the work. A small studio like ours gives you a direct line to the developers, lower overheads in the quote, and faster decisions — at the cost of that bench depth. For a small business website, an online shop or a booking system, the agency structure is overhead you are paying for and rarely using.
Ready to Get Your Business Online?
Based in Bridgend and serving businesses across South Wales. Get a free, no-obligation quote today with clear and honest advice.