What the best agency-client relationships actually look like
The best agency-client relationships aren't about being nice. They're about honest communication, clear structure, and shared accountabil...
The Blog
The most expensive mistakes in web projects don't happen during development. They happen before it, in the decisions that get made (or skipped) before anyone opens a code editor.
Discovery is the phase most teams rush through or cut entirely. It's the easiest line item to remove from a proposal because the output isn't a visible product. But what it produces, a shared understanding of what's being built, why, and how, is what determines whether the build goes smoothly or burns through budget fixing problems that should have been resolved in week one.
The cost
Most projects that go sideways don't fail because of bad code. They fail because assumptions went unchallenged.
The content model gets designed around what's on the current site rather than what the business actually needs. Integration requirements surface mid-build when a stakeholder mentions the CRM for the first time. The editorial workflow gets figured out during UAT when the content team finally sees the CMS and says "this isn't how we work."
These aren't edge cases. They're the default outcome when discovery gets compressed into a single kickoff meeting and a shared Google Doc.
The cost isn't just rework. It's the compounding effect, every decision made on a wrong assumption creates downstream dependencies. Fix one and three others need revisiting. By the time the problem is visible, the budget impact is already locked in.
The work
Knowing what discovery should cover is one thing. Running it in a way that actually shapes the build is another. Three principles that make the difference.
Discovery should be its own engagement with its own deliverable. Not a section of the proposal, a distinct piece of work that produces a build specification. This protects both sides: the client gets a detailed scope before committing to the full build, and the agency prices the build on real requirements.
Content editors, marketing leads, integration owners, IT stakeholders. Not just the project sponsor. The people who will use the platform daily need to be in the room during discovery, not introduced during UAT.
A good discovery output includes the decisions that were made and why. When someone questions a content modelling choice in month three, the documentation explains the reasoning, not just the outcome.