The actual list. Including what we don't.
Most agency sites describe capabilities in the abstract and leave you to guess whether your project is one of them. This page is the concrete version - every kind of thing we ship, roughly what it costs in weeks, what comes as standard regardless, and the work we turn down.
12
Kinds of build we take on, listed in full below
4
Services they're bought under, nothing hidden between
1wk
Shortest engagement - a landing page or an audit
100%
Of the code and infrastructure ends up yours
Twelve things we build, in plain terms.
Timelines are the honest middle of what we've actually run, not the best case. Each one links to the service page it's sold under.
Marketing sites
The site that has to carry the brand and still convert. Hand-coded, fast, and structured around the one action a visitor should take.
- Positioning and message hierarchy before any design
- A design system, so page nine looks like the homepage
- CMS on the pages that actually change
- Core Web Vitals tuned on real hardware
Landing pages and funnels
A single page built to be measured. One offer, one argument, one action - and the instrumentation to tell you whether it worked.
- Copy structure written around the objection, not the feature
- Forms wired to your CRM or inbox with an SLA clock
- Event tracking on every step of the funnel
- A/B variants where there's enough traffic to read them
SaaS platforms
Multi-tenant products with the unglamorous parts done properly - auth, roles, billing and a data model that survives its second year.
- Organisations, invitations and row-level data isolation
- Stripe subscriptions with entitlements derived from billing state
- Onboarding and activation designed as a flow, not a screen pile
- CI, preview environments and error tracking from the first deploy
MVPs and prototypes
The smallest credible product, built to production standards. Small scope, full quality - so what you learn is about the idea, not the duct tape.
- A scope workshop that ends in a written not-now list
- Clickable prototype in week one, before production code
- Accounts and payments if the loop needs them
- Analytics on the events that define activation
Internal tools and admin
The system your own team lives in. Usually the highest-return thing we build, and almost always the thing nobody has had time to do.
- Operations replacing a spreadsheet that three people edit at once
- Bulk actions, imports, exports and audit trails
- Impersonation, refunds and debugging without a database client
- Permissions enforced server-side, per role
Dashboards and reporting
Numbers your team trusts enough to act on, recalculating themselves instead of waiting on somebody's Tuesday export.
- Customer-facing analytics and internal reporting from one model
- Aggregations pre-computed where the query would be slow
- Scheduled email digests and CSV export
- Charts that stay readable at the sizes people actually use
CRM and sales automation
Your pipeline modelled as it really runs, with the manual admin inside it removed. Not a generic tool your team quietly works around.
- Every lead source - form, email, WhatsApp, calls - in one queue
- Follow-up sequences, reminders and stage-change notifications
- Forecasting that doesn't need reconciling by hand
- Migration dry-run before anything goes live
Integrations and APIs
The plumbing between systems that already exist. Usually invisible, usually the reason somebody's afternoon disappears every week.
- REST and webhook integrations with retries and idempotency
- Stripe, WhatsApp Business, Google Workspace, accounting tools
- Scheduled syncs with reconciliation you can actually read
- A public API for your product, documented, if you need one
E-commerce and checkout
Storefronts where the checkout is treated as the product. Headless where it earns its place, hosted where it doesn't.
- Headless Shopify or a custom Stripe checkout, whichever fits
- Product, variant and inventory modelling that matches your catalogue
- Checkout tuned for speed - the step where every 100ms shows
- Order, fulfilment and refund flows for your ops team
Design systems
Type, colour, spacing and component states defined once and shipped as code, so the tenth screen doesn't need a designer to look right.
- Tokens shared between Figma and the codebase
- A component library with every state built, not just the happy one
- Accessibility baked into the components rather than audited after
- Documentation your team can hand to a new developer
Rescue and rebuilds
Something exists, it's paying customers, and everyone is scared of it. We start with a paid audit and tell you honestly what's worth keeping.
- Architecture, data model, auth and test coverage reviewed
- A written verdict: keep, refactor or replace, and in what order
- Incremental migration so the product stays live throughout
- Sometimes the answer is that a rewrite costs less - we'll say so
Performance and SEO work
A focused pass on a site you already have. Measured before, measured after, with the numbers put in writing either way.
- LCP, CLS and INP profiled on real devices and connections
- Semantic markup, heading structure and JSON-LD schema repaired
- Images, fonts and third-party scripts brought under control
- A written report of what changed and what it moved
Something not on this list? Ask anyway - a straight yes or no takes us a message, not a meeting.
The floor, not the extras.
These eight arrive with a one-week landing page and a sixteen-week platform alike. They're not a premium tier and they never appear as a line item - a build without them isn't finished, it's just handed over early.
Ask what that means for your buildA performance budget
Measured on real hardware before launch, not on a fast laptop over office wifi. If it misses, it doesn't ship.
Semantic, indexable markup
A real heading hierarchy, metadata per route, Open Graph cards, sitemap, robots and JSON-LD. Search engines shouldn't have to guess.
Responsive down to 320px
Designed at mobile width first and checked at every breakpoint between, because that's where most of your traffic already is.
Accessibility as you go
Contrast, focus states, keyboard paths and labelling handled in the components, rather than audited into existence afterwards.
Analytics and error tracking
Wired before launch, not after the first incident. You get a funnel and an alert, not a dashboard nobody configured.
CI and preview environments
Every pull request builds and deploys to its own URL. Reviewing a change means opening a link, not booking a demo.
Documentation and handover
A README that gets you running in one command, architecture notes on the odd decisions, and a walkthrough with your team.
Ownership, from commit one
The repo sits in your GitHub organisation and everything deploys into accounts registered to you. We can be removed any day you like.
A boring stack, on purpose.
Chosen so any competent developer can pick the codebase up in an afternoon - including the one you hire after us. We'll work in yours instead where you already have one.
Interface
Data and back-end
Platform services
Content
Infrastructure
Measurement
Four things we'll turn down.
A studio that says yes to everything is telling you something. These come up often enough that it's worth answering here rather than three calls in, when you've spent a fortnight on it.
Native iOS and Android apps
We build responsive web and installable PWAs. If you genuinely need the App Store and platform APIs, you want a mobile studio, and we'll say so rather than learn on your budget.
Patching someone's WordPress theme
Plugin conflicts and theme surgery aren't work we're good value at. A rebuild often costs less than a year of patching - but if it doesn't, a local freelancer is the honest answer.
Bespoke where off-the-shelf fits
If HubSpot, Shopify or Webflow would genuinely do the job, we'll tell you on the first call. Paying for custom you don't need is the most common way this money gets wasted.
Token launches and growth-hack funnels
Crypto token projects, scraped-list outreach and anything whose business model depends on people not reading carefully. Not a moral lecture - just not what we build.
Twelve builds, four services.
Whatever the build, the engagement is one of these four - and the process behind it is the same five stages at different scales.
Tell us which one you need.You'll get a number back.
A written scope and a fixed price, normally inside 48 hours. If what you need isn't something we should be building, you'll hear that instead - and usually who to talk to.
- No retainer to sign first
- Fixed scope, fixed price
- You own the code