Outcome agreed first
We define the commercial result and how it will be measured before scoping features.
How we work
A process built to remove surprises
Six stages, each with defined outputs and a genuine decision point. At every stage you can stop, change direction, or continue — and you own everything produced up to that point either way.
What we commit to on every engagement
Bad news travels fast
If a sprint slips you hear it that week, in the delivery report, not at the milestone.
No layers between us
You have direct access to every engineer on your squad, in your channel.
Handover from day one
Documentation is written as we go, not compiled in a panic at the end.
The six stages
What happens, and when
Durations are typical for a mid-sized engagement. A two-week rebuild and a twelve-month programme follow the same shape at different scales.
-
01 1–2 weeks
Discover
We sit with the people whose work the software will change, map current and target workflows including the exceptions, and surface the constraints that never make it into a brief.
- Stakeholder interviews
- Workflow & data mapping
- Risk register
- Prioritised backlog
-
02 1–2 weeks
Define
Discovery becomes a plan: architecture, scope boundaries, a costed roadmap with a stated confidence range, and an explicit list of what is not included.
- Solution architecture
- Scope & estimate
- Success metrics
- Delivery plan
-
03 2–5 weeks
Design
Information architecture, interaction design and a tokenised component system — tested with real users and annotated for accessibility before engineering starts.
- Journeys & wireframes
- Interactive prototype
- Design system
- Usability findings
-
04 Ongoing sprints
Build
Two-week increments, each ending in something demonstrable on a real environment. Trunk-based development, mandatory review, automated quality gates.
- Working increments
- Automated tests
- CI/CD pipeline
- Sprint demos
-
05 1–2 weeks
Launch
Load testing, security review, rehearsed cutover and a rollback plan. Releases are staged and observable, so problems are caught by monitoring rather than by customers.
- Load & security testing
- Cutover runbook
- Monitoring & alerts
- Go-live support
-
06 Continuous
Evolve
Post-launch, we measure against the outcomes agreed in discovery and iterate. Support retainers cover monitoring, patching and a monthly enhancement allowance.
- Outcome reporting
- Iteration backlog
- Security patching
- Capability handover
Discovery in detail
Two weeks that de-risk everything after
Fixed-price software fails when scope is guessed. Discovery converts guesswork into evidence, and it is the reason our estimates land within about 15% instead of drifting.
- Days 1–3 · Understand the business. We sit with the people who do the work today, not only their managers. Every engagement surfaces at least one constraint that appeared in no brief.
- Days 4–6 · Map current and target state. Workflows including the exceptions — the manual override, the customer with different terms, the report that must reconcile to the penny.
- Days 7–9 · Prototype the risky part. Whatever is most uncertain gets built roughly, turning the largest unknown into a known quantity.
- Day 10 · Architecture and estimate. A documented architecture, prioritised backlog, costed plan with a stated confidence range, and an explicit list of what is excluded.
Discovery costs €9,500 and is credited in full against a delivery contract starting within 90
days. You own every artefact regardless of whether you continue with us.
Discovery deliverables
- Stakeholder research findings
- Current & target workflow maps
- Clickable prototype of the core journey
- Solution architecture & decision records
- Risk register with mitigations
- Prioritised, costed backlog
- Written recommendation — including "do less"
What discovery changes
- ~15%typical estimate accuracy after discovery
- 1 in 8discoveries end with us recommending less work
- −60%fewer change requests during delivery
During delivery
What a fortnight actually looks like
No status theatre. Here is the real rhythm of an engagement once the build starts.
Monday · Planning
The squad and your product owner agree the increment scope. Anything carried over is explained, not quietly rolled.
Daily · Async standup
Written updates in your shared channel by 10:00 local. Blockers are flagged with a named owner.
Continuous · Merge to main
Trunk-based development with mandatory review. Every merge deploys to a live environment you can use.
Weekly · Quality report
Performance budget, accessibility gate, test coverage and dependency health, sent whether or not you ask.
Fortnightly · Demo
Working software on a real environment, not slides. Feedback goes straight into the next planning session.
Monthly · Outcome review
Progress against the commercial metrics agreed in discovery, with a frank view on whether we are on track.
Process questions
Common things clients ask
Every increment ends on a live environment you can use, plus a demo. You have access to the board, the repository and the team channel throughout — there is no reporting layer between you and the people building.
On squad engagements you simply re-prioritise the backlog; there is no change-request process. On fixed-scope work, changes are quoted transparently against the agreed baseline.
Mandatory code review, automated tests, performance budgets and accessibility gates enforced in CI, and DORA metrics tracked per team. These are pipeline gates, not aspirations.
You hear about it from us first. We hold weekly delivery health reviews internally, and anything amber is raised with you in that week rather than at the milestone.
Available for new engagements
Start with discovery
Two weeks, a fixed fee, and a plan you own outright. It is the cheapest way to find out whether the project you have in mind is the project you actually need.
- Reply within one business day
- No obligation, no sales sequence
- NDA on request before we talk