Home/Process

From rough idea to real product.

Six phases, each with a decision at the end of it. You always know what you'll have, when you'll have it, and what it lets you decide next.

Six phases, one roadmap

1
Discover
1–2 weeks
2
Validate
1–2 weeks
3
Design
2–4 weeks
4
Build
6–12 weeks
5
Launch
1–2 weeks
6
Scale
Ongoing
01

Discover

1–2 weeks

We learn the business before we design anything: how the work happens today, who it fails, and what would make this project not worth doing.

What happens
—Stakeholder and user interviews
—Workflow and systems mapping
—Market and competitor read
—Opportunity sizing
You end up with
Problem statement everyone signs
Opportunity map
Go / no-go recommendation
Team: Strategy lead, designer
02

Validate

1–2 weeks

We define the smallest product that can be judged honestly, and check the assumptions that would sink it.

What happens
—MVP boundary and scope
—Concept tests with real users
—Technical feasibility spikes
—Success metric definition
You end up with
MVP scope document
Clickable prototype
Metric and tracking plan
Team: Strategy lead, designer, tech lead
03

Design

2–4 weeks

Experience and architecture designed together, so the interface and the system don't surprise each other later.

What happens
—Core flows and screen design
—Design system foundations
—Data model and API shape
—Edge and empty states
You end up with
Production-ready designs
Component library
Architecture decision record
Team: Designer, tech lead
04

Build

6–12 weeks

Engineering in visible increments — something demoable every week, with quality gates that don't slip.

What happens
—Two-week increments, weekly demo
—Automated tests and CI
—AI features with human review
—Instrumentation as we go
You end up with
Working product in staging
Test suite and pipeline
Analytics live from day one
Team: Squad: 2–4 engineers, designer, PM
05

Launch

1–2 weeks

Production infrastructure, monitoring and a rollout plan — plus the growth work already running before the doors open.

What happens
—Infrastructure and security review
—Monitoring, alerts, on-call
—Staged rollout and rollback plan
—Launch campaign and landing pages
You end up with
Production environment
Runbook and dashboards
Launch plan executed
Team: Tech lead, engineer, growth lead
06

Scale

Ongoing

Now the real work: read what users actually do, improve the weakest link, automate the parts that repeat.

What happens
—Experiment cadence on product and channels
—Retention and lifecycle programmes
—Automation of manual operations
—Quarterly roadmap reset
You end up with
Monthly reporting you can audit
Prioritised experiment backlog
Automations in production
Team: Growth lead, engineer, analyst
How it feels to work with us

Few rituals, held properly.

Enough process to stay honest about progress, not enough to become the job.

One board, one truth

You see the same board we do — no separate client view, no status theatre.

Weekly demo

Working software every week, in your hands. Progress is shown, not described.

Decision log

Every significant call is written down with the reasoning, so new people inherit context.

Named humans

The people in the kickoff are the people doing the work. No bait-and-switch staffing.

Straight answers

The questions everyone asks.

How much does an MVP cost?

Most first releases land between a focused two-week sprint and a three-month build. We give a fixed range after Discover, before you commit to Build.

Can you start mid-project?

Yes — often with a short audit first, so we inherit the codebase honestly instead of promising a rewrite.

Who owns the code?

You do, from day one. Repositories, infrastructure and analytics live in your accounts.

Do we have to use AI?

No. We propose it only where it removes a measurable cost, and we'll tell you when a rule or a better default is the cheaper answer.

What if we need to stop?

Engagements break at phase boundaries. You keep everything produced up to that point, documented.

Do you do marketing without building?

Yes, and building without marketing too — though the results are usually better when the two share a roadmap.

Phase one starts with a conversation.

Bring the problem. We'll tell you which phase you're actually in — sometimes it isn't the one you think.

Start a Project →