Discovery phase is reshaping how we work. Discovery is the phase clients are most tempted to skip. It is also, by some distance, the phase that determines whether the project succeeds.
Almost every troubled digital project we have inherited from another agency shares a single root cause: discovery was either skipped, rushed, or done as a tick-box exercise to get to the visible work. Wireframes were drawn before the underlying workflows were understood. Features were scoped before the user research had been done. Estimates were given before the integration topology had been mapped.
The damage is not visible until weeks 8, 12 or 16, when the build hits a wall that a proper discovery would have surfaced on day three.
What the discovery phase actually produces
Done well, a discovery phase produces a set of artefacts that make every subsequent decision faster and better:
1. A user journey map – not in the marketing sense, but in the operational sense. Every interaction with the product, from awareness through onboarding through daily use through churn risk, mapped to the moments where value is created or destroyed.
2. A workflow audit. For B2B and operational products, this means walking through actual current processes – usually with a spreadsheet, a whiteboard and a stopwatch. How long does each step take? Where do errors creep in? Where do people work around the system? This audit is where the actual ROI case for the build gets quantified.
3. A data model. Even before any UI is designed, the data model tells you what the product fundamentally is. What entities exist, how they relate, what attributes matter. A clean data model emerging from discovery saves more development time than any other single artefact.
4. An integration topology. What systems must this product talk to, in which direction, with what cadence? Each integration is a project in itself and unsurfaced integrations are the single most common cause of timeline overruns.
5. A risk register. Every assumption that, if wrong, would invalidate the plan. AI feasibility, third-party API limits, regulatory constraints, user adoption thresholds. Discovery exists to surface these before commitment, not after.
6. A scoped, prioritised backlog. Not a wishlist. A ranked list with a clear cut-off between “essential to v1” and “post-launch.” The discipline of forcing this prioritisation in discovery is where most of the cost discipline of a project is set.
The cost of skipping it
clients who push back on discovery usually frame it as a cost question. “We know what we want. Why pay for a phase that doesn’t produce anything visible?”
The honest answer is that a one-week discovery phase costing £4-8k routinely prevents £30-80k of misdirected build work later. We have rebuilt entire applications because the original team skipped discovery and committed to an architecture that didn’t fit the actual workflow. The economics aren’t close.
The other cost is harder to measure but more damaging: stakeholder fatigue. Projects without proper discovery generate revisions and disagreements throughout the build phase, because foundational decisions were never explicitly made. By the time the project ships, everyone is exhausted, the budget is bloated and the product is a compromise. Discovery is where those decisions get made cleanly, in a single concentrated burst, by the right people.
The four conversations discovery is built around
Our discovery sprints are structured around four core conversations, each producing a defined artefact.
Conversation 1: Strategic intent. What is this product for? What does success look like in 12 months? What would make this an unambiguous failure? Who in the organisation needs to be aligned on those answers?
Conversation 2: User context. Who will use this, in what circumstances, with what alternatives? This is often the conversation that requires actual user research – interviews, observation sessions, sometimes diary studies for B2B operational products.
Conversation 3: Operational reality. What workflows will this touch? What systems already exist? Where are the constraints that aren’t obvious from the outside – compliance, contractual, political?
Conversation 4: Technical feasibility. What are the integration points, the data sources, the third-party dependencies? Where might the technology not deliver what we want it to?
Each conversation produces a specific artefact. By the end of discovery, those artefacts compose into a project brief that a senior engineer or designer could pick up and execute against without ambiguity.
How long discovery should take
most projects benefit from a discovery phase between 1 and 4 weeks. The variables are scope, stakeholder count and the maturity of existing thinking.
- A new website rebuild: 1-2 weeks. Most of the strategic clarity already exists; discovery is about converting it into structured artefacts.
- A new web application MVP: 2-3 weeks. User research and workflow mapping take time and skipping them in this scenario is where most MVPs fail.
- An enterprise platform or AI implementation: 3-4 weeks, sometimes longer. The integration topology alone often warrants two weeks of dedicated work.
For projects where discovery surfaces real ambiguity about the product itself, we frequently extend with a rapid prototyping sprint – converting the discovered idea into an interactive artefact that stakeholders can react to, before any production build commits.
The marker of a good discovery output
A good discovery phase produces a brief that is dense, specific and decisive. There are no “to be confirmed” sections. There are no decisions deferred to “discuss in design.” Every screen has a purpose, every entity has a definition, every integration has an owner.
When the first sprint of build work begins, no one is wondering what to build. They are building it.
The strategic point
Skipping discovery is the most expensive false economy in digital project work. The teams that ship great products on time and on budget are not the ones with the best engineers – they are the ones who entered the build with a discovery output that made the engineers’ job clear.
pay for clarity early. Spend the rest on execution.
Considering a new digital project and want to start it the right way? Our Discover phase is the foundation of every successful build we deliver. Get in touch to scope yours.