ebsite discovery is the work of replacing assumptions with decisions before the expensive part of delivery begins. It is not a workshop performed for theatre, and it is not a promise that every uncertainty will disappear. A useful discovery phase identifies the problem, gathers evidence, maps dependencies and produces a scope that the team can explain, estimate and test.
The team reviews business goals, analytics, search performance, customer research, sales questions, support issues, current content, technical systems and stakeholder constraints. Interviews are useful when they surface decisions and contradictions. Opinions should be labelled as opinions until behavioural, operational or market evidence supports them.
- Priority audiences and the context in which they arrive.
- Questions, tasks, objections and trust needs at each stage.
- Existing content to keep, improve, merge, migrate or remove.
- Critical journeys such as enquiry, booking, purchase, application or account access.
- Internal handoffs after a visitor completes the visible website action.
The website may depend on a CMS, CRM, booking platform, identity provider, payment service, product database, analytics setup or inherited hosting. Discovery confirms access, data ownership, API capability, limits, failure handling and security expectations. A line item called “integration” is not understood until both sides know what information moves, when and who owns the source of truth.
The outputs should distinguish must-have launch work from later improvements, name exclusions and assumptions, and connect features to evidence. A sitemap, content model, representative wireflows, technical approach, risk register and measurement plan may all be appropriate. The exact artefacts matter less than whether they support real decisions.
If discovery reveals that the current platform can be improved, the solution should shrink. If a supposedly simple form actually triggers a regulated workflow, the scope should grow before development. Discovery that merely restates the original sales pitch has not reduced risk. Its value is the permission to change course while change is still cheap.
Discoveryissuccessfulwhenthebuildteamcansaywhatitismaking,why,andwhatremainsuncertain.
The output is not a stack of workshop photographs. It is a decision record: audiences and jobs, content and URL inventory, success measures, technical constraints, risky assumptions, responsibilities, a prioritised release and the questions still open.
- Stakeholder interviews reveal goals, constraints and decision authority.
- Analytics, search and support evidence reveal actual behaviour.
- Content and URL audits show migration and ownership work.
- Journey mapping exposes states, handoffs and failure paths.
- Technical investigation tests integrations and legacy assumptions.
- A playback turns findings into scope, priorities and acceptance criteria.