Why two weeks

Clients ask why discovery is not one week or six. One week produces opinions without evidence — integration paths guessed, data quality assumed, governance discovered later at cost. Six weeks produces reports that age before engineering starts and teams that lose executive attention. Ten working days, with the right three people and prepared stakeholders, is the balance we have standardised.

What two weeks must answer

• Which use case ships first to production, and what is explicitly out of scope?
• What data and integrations does that use case require, and what is the readiness gap?
• Where do models and agents run to satisfy residency and security?
• What governance controls are non-negotiable before go-live?
• What does build and run cost, under fixed scope?
• Can we credibly commit to production by week twelve of the build?

If any answer is we do not know yet, discovery is not finished. We extend only when access or stakeholder availability blocks us — and we name that blocker in writing, not absorb silent delay.

The cadence inside ten days

Days one and two establish the map: systems, owners, constraints, success criteria. Days three through five go deep on data, integrations and process walkthroughs — often the longest days, requiring business and IT in the room together. Days six through eight turn findings into a scored portfolio, architecture sketch and remediation list. Days nine and ten build the cost model, delivery plan and readout deck your sponsor presents upward.

Preparation on your side accelerates us: stakeholder list, read-only access to key systems, sample documents or cases, existing architecture diagrams even if stale. We send a preparation checklist when discovery is booked.

Fixed fee, credited to build

Discovery is a fixed fee regardless of outcome — proceed or not. If you engage us for the build within sixty days, the fee credits against that contract. That aligns incentives: we want you to proceed with a plan that works, not with false optimism. Saying not yet is a valid discovery outcome.

Two weeks versus ongoing advisory

Some firms sell continuous strategy retainers. We sell a decision point. Advisory without a path to code is not our business. If the discovery conclusion is foundations before agents, the proposal says so with dates and price. If the conclusion is proceed, engineering starts with scope already argued and settled.

When two weeks is not enough

Rare complex environments — many business units, contested data ownership, parallel countries — may need a bounded extension or a second discovery module on a specific domain. We propose that on day five with explicit added fee and deliverables, not drift.

Relationship to the twelve-week build

Discovery is weeks one and two of the overall journey to first production. The build proposal covers weeks three through twelve for Engineer and Deliver. Run is separate. Total time from kickoff to production is fourteen weeks when discovery flows directly into build — a pace executives can plan around.