Practice 01 / Web Development Updated October 2026

Web development services for platforms that have to hold.

One accountable team for the web layer of the business, from the customer experience to the systems behind it and the operating model after launch.

Web services
09
Core disciplines
04
Performance budgets
CWV
Accountable team
1

Overview

What Web Development has to solve.

Replacing a dated platform or extending a working product. The useful questions are about ownership, failure paths, and what the team can operate after launch.

What good looks like

  • Performance, accessibility, and SEO budgets agreed before design
  • Server-rendered pages with a clean hydration path
  • A content model editors can run without engineering
  • An API layer that other teams can consume safely
  • Documentation, monitoring, and handover included

What custom web development includes

Custom web development turns a business process into a system people can use, operate, and change. It includes the interface, the data model, the API layer, the deployment path, and the operational detail that keeps the whole thing healthy after launch.

The scope usually starts with one of four surfaces: a marketing and content platform, a commerce flow, a customer or partner portal, or an internal system. The technical answer changes, but the questions stay the same: who owns each entity, what happens on the exception path, and how the team will know the system is working.

  • Information architecture and content modelling
  • Server-rendered interfaces and reusable design components
  • Authentication, roles, permissions, and audit
  • API contracts, integrations, and background jobs
  • Performance, accessibility, SEO, and analytics instrumentation
  • Deployment, monitoring, and handover documentation

How we approach a web platform build

We start with the operating reality. Discovery interviews cover the people who use the platform, the people who support it, and the people who reconcile the data afterward. That is where the expensive assumptions usually surface.

The build runs in working increments with a working environment on every commit. We schedule the riskiest integration or workflow first, because a late surprise is a scope problem disguised as an engineering problem. Performance, accessibility, and analytics are tested continuously from the first release.

  • Risky integrations are scheduled before polish
  • Every release has a rollback path
  • Content editors test the publishing workflow in staging
  • Analytics events are specified with the interface from the start

What changes after launch

The first phase of real traffic shows which assumptions were wrong, which administrative tools are missing, and where the support team still does work the software should absorb.

We plan for that transition. The handover includes runbooks, architecture decisions, dashboards, and a short list of improvements worth making once the system has real usage. Some clients keep us for long-term operation, and others take the keys. Both are valid outcomes.

What a useful web development brief should settle

The brief should name the audience, the revenue or operational outcome, and the systems that already hold the truth. It should also say who owns content, identity, deployment, and support after launch.

A clear brief separates the platform from the campaign. It identifies which surfaces must be shared, which can stay independent, and which integrations are non-negotiable. That prevents a redesign from becoming an unbounded migration.

  • The primary user journeys and the failure paths that matter
  • Systems of record, identity providers, and integration owners
  • Performance, accessibility, and SEO budgets
  • Who operates the platform and how change is approved

Why teams call us

What usually goes wrong.

These are the recurring problems behind a web development brief. Naming them early changes the architecture, the estimate, and the launch plan.

01

The site is a stack of plugins

Every campaign adds another plugin, script, and owner. The team cannot explain what runs on the page or why it is slow.

02

Marketing cannot publish

A landing page needs a developer, a release, and a queue. The content model was built for a demo, and daily editors struggle with it.

03

The platform is a black box

The previous agency left without documentation, credentials, or a reliable deployment path. The business is afraid to touch it.

04

Traffic exposes the seams

Checkout, search, and account pages behave differently under load. The architecture was sized for a launch, and the business has outgrown it.

Scope

What the project includes.

Every web development project is scoped around a working system and the decisions that keep it healthy.

01

Website and platform development

Marketing sites and product surfaces built on a reusable component system. Performance and accessibility are release criteria.

02

Commerce and CMS

Catalog, checkout, merchandising, and editorial tooling designed around the people who operate them.

03

Portals and internal systems

Customer, partner, and operations interfaces with identity, permissions, audit, and the exception paths that decide whether the system works.

04

Custom web applications

Focused software for workflows that off-the-shelf products cannot cover, built around decisions, roles, and measurable throughput.

05

API integration and PWAs

Explicit contracts, idempotent writes, third-party integrations, and offline-capable web experiences where the brief calls for them.

06

Operational handover

Runbooks, dashboards, architecture records, and a team that can ship the next release without reverse-engineering the last one.

Use cases

Where this service helps most.

The same service looks different in each company. These are the situations where it matters most.

01

A revenue platform outgrew its CMS

A marketing site, customer portal, and commerce flow live on three stacks with three versions of the truth. We consolidate the experience and avoid a risky big-bang rewrite.

One content and identity model, released market by market

02

A product team needs a web front door

The API is strong, but onboarding, billing, and account management still happen through support tickets and spreadsheets.

Self-service web experience with billing and role-aware access

03

Operations runs on spreadsheets

The real workflow lives in email, exports, and a spreadsheet only one person understands. We model the process and replace the manual reconciliation around it.

A web platform with audit, permissions, and a measurable close

Process

Four stages. Working software at every review.

Discovery leads to an architecture decision, a build with visible increments, and an operating model that stays after we leave.

What stays constant

One accountable lead, a working demo at each review, and documentation from the first release. Scope and risk set the plan, and we will say when a smaller project is the better answer.

  1. 01

    Discover

    Interview the people who use and operate the platform, audit the current stack, and agree the numbers that define success.

    Output: Discovery brief, system audit, success metrics

  2. 02

    Architect

    Define the target architecture, content and identity models, a clickable prototype, and a delivery plan that puts the risky work first.

    Output: Architecture decisions, prototype, release plan

  3. 03

    Build

    Working increments, staging on every commit, and a demo you can click. Performance and accessibility budgets run in CI.

    Output: Production releases, test coverage, living documentation

  4. 04

    Operate

    SLOs, monitoring, cost reporting, and a roadmap that responds to production.

    Output: Runbooks, SLO reviews, improvement roadmap

Stack & standards

Chosen for durability.

Fewer moving parts, clear ownership, and a handover your engineers can defend. The stack follows the constraint.

Angular

Accessible, long-lived application interfaces

TypeScript

Shared contracts across client and server

Node.js

APIs, integration, and server rendering

PostgreSQL

Durable relational data and reporting

Sanity

Editorial content without release friction

Vercel

Preview deployments and edge delivery

How we choose

01

Custom build versus template

A template is cheaper until the workflow stops fitting. We recommend custom when the platform is revenue-critical, the content model is unusual, or the integration surface is the product.

02

Monolith versus composable

Composable adds complexity. We split systems where ownership or release cadence demands it, and keep the rest simple.

03

Rebuild versus extend

A rewrite is a business decision. We map what can be stabilised, what must be replaced, and the smallest sequence that reduces risk.

Measured outcomes

“They rebuilt the core without downtime. The close now runs from one reconciled ledger.”
Elena RossiChief Technology Officer, Meridian

Selected result

One reconciled close

Read the Ledger case study
All work

FAQ

Questions we hear often.

The questions buyers ask before a web development project. If yours is not here, ask it directly.

Ask an engineer
01 How much does custom web development cost?

Most fixed-scope web platform projects start in the mid five figures and scale with integration, content, and compliance complexity. A focused marketing site is smaller than a portal or commerce platform. We scope against outcomes and give you a written range before design begins.

02 Do you work with our in-house engineering team?

Yes. Most of our best work happens alongside a client team. We can own the architecture and delivery, embed specialists, or run a joint squad with your engineers and rituals.

03 Can you migrate an existing site without downtime?

Usually. We use a phased migration, redirect mapping, parallel environments, and a rollback plan. Content and SEO are audited before cutover, and we keep the old path available until the new one is proven.

04 What stack do you recommend?

The stack follows the constraint. Angular, TypeScript, Node, PostgreSQL, and a headless CMS cover most enterprise web platforms. If your team already has a strong stack, we work with it and skip the rewrite.

05 Do you handle SEO and accessibility?

Yes. Semantic markup, Core Web Vitals, structured data, and WCAG 2.2 AA checks are part of the build. We review them throughout the project.

Next step

What should your web platform stop making people work around?

Tell us the workflow or the failing release. A senior engineer will reply with a concrete plan and a clear next step.

Scope a web platform Direct reply from a senior engineer