Scoping a digital project is the most under-invested part of most agencies' workflows. People rush it because it isn't billable, and then they spend the next six months arguing about what was actually in scope. Here's the one-week discovery process we use to avoid that.

Why one week

Less than a week, and you don't have time to truly understand the problem. More than a week, and you're either over-engineering the discovery or you've identified that the project is too complex for fixed-scope and should be a retainer engagement.

One week is also the right length to commit to as a discrete piece of work — clients can carve out the time, and our team can focus on one engagement at a time.

Day 1: Kickoff and stakeholder interviews

We start with a half-day kickoff workshop. The client brings the decision-makers — usually a founder or product lead, an engineering lead if relevant, and whoever owns marketing. We bring the design lead and the engineering lead who'll likely run the project.

The workshop has three sections:

  • Vision and constraints. What does success look like in six months? What's non-negotiable? What's flexible?
  • Users. Who is this for? What do they currently do? What's their alternative if our project doesn't exist?
  • Risks and unknowns. What scares you about this project? Where are the integrations or migrations you're nervous about?

The afternoon is for one-on-one stakeholder interviews. We talk to anyone who'll have to live with the result.

Day 2: Technical and design audit

If the project involves replacing or building on existing systems, we audit them. For a marketing site rebuild, that means crawling the existing site, identifying every page type, and noting performance, SEO, and accessibility issues. For a software project, it means reviewing the existing codebase, the data model, and the deployment story.

We're not yet trying to solve anything. We're inventorying reality.

Day 3: Architecture and design exploration

Now we start sketching. The design lead is exploring layout directions, key page types, and information architecture. The engineering lead is sketching the technical architecture — what's the data model, what services are involved, where are the failure modes.

These are deliberately low-fidelity sketches. We're testing approaches, not committing to any. The day usually ends with two or three viable directions and a clearer sense of which one fits.

Day 4: Scope, milestones, and budget

This is the hardest day. Based on the previous three, we draft a scope: the exact pages, screens, features, and integrations included. We draft milestones: what's delivered at each checkpoint. And we draft a budget: how many designer-weeks, how many engineer-weeks, what the timeline looks like.

The crucial discipline is being honest about what's not included. The scope document lists exclusions almost as prominently as inclusions. "Multi-language support is not included." "Migration of historical CMS content is not included unless we use the optional migration package."

If we can't write a scope with confidence, that's the signal that this project should be a retainer or that we need more discovery before quoting.

Day 5: Review, revise, and deliver

The last day is a working session with the client. We walk through the scope document, the milestones, the budget, and the assumptions. They push back. We revise. By the end of the day, we either have a proposal everyone agrees with — or a clear sense of what would need to change.

The deliverable is a single document:

  • Problem statement and project goals.
  • Recommended approach and key decisions.
  • Detailed scope with inclusions and exclusions.
  • Milestones and timeline.
  • Budget and payment schedule.
  • Assumptions and dependencies.
  • What happens after launch (support, retainer, handoff).

Why this works

The week of discovery is almost always cheaper than the cost of getting a project wrong. We've never had a project go badly off-scope when we did this discovery well. We've had several projects go badly off-scope when we skipped it.

The client is paying us for clarity, not just for code and pixels. The scoping week is where most of the clarity gets manufactured.

When we skip it

The exception is small projects — under £15K — where a week of discovery would consume too much of the budget. For those, we run a one-hour scoping call and write a one-page scope. The cost of a misstep on a small project is small enough that the lighter process is the right trade-off.

For everything bigger, the scoping week pays for itself many times over.