Discovery & BA Deliverables

Discovery exists to replace assumptions with a scoped backlog we can estimate. It is not “gather requirements forever” — it ends at a defined exit, not when the client stops talking.

From lead to a scoped backlog

flowchart LR
  L["Lead / Sales"] --> D["Discovery kickoff<br/>with BA"]
  D --> A1["Problem statement"]
  D --> A2["MoSCoW scope"]
  D --> A3["User flows"]
  D --> A4["Lightweight SRS"]
  D --> A5["Risk register"]
  A1 --> X["Scoped backlog<br/>ready for estimation"]
  A2 --> X
  A3 --> X
  A4 --> X
  A5 --> X

The deliverables

Output	What it answers	Owner
Problem statement	Why this project exists, who the users are, what success looks like. One page.	BA
MoSCoW scope	Must / Should / Could / Won’t — draws the first hard line around scope.	BA + client
User flows	The core journeys, screen-to-screen. Diagram-first, not prose.	BA
Lightweight SRS	Functional requirements + non-functional constraints (performance, security, integrations). Not a 60-page spec.	BA
Risk register	Known unknowns, dependencies on the client, external systems, assumptions that could break the estimate.	BA + PM

Exit criteria

Discovery is done when:

• Every Must item maps to at least one backlog item written to the Definition of Ready standard.
• Open questions are logged as risks or assumptions, not left implicit.
• The backlog is granular enough that Estimation can produce a PERT range.

Caution

“We had lots of meetings” is not an exit. If the backlog can’t be estimated, discovery isn’t finished.

The backlog and all discovery artifacts live in Azure DevOps Boards — this page describes the intent, Boards holds the state. Next stop: Estimation & Scoping.