Most web projects that fail don’t fail in development. What separates them from the ones that succeed is usually something top web design agencies understand and average ones don’t: projects fail in the first two weeks — before a single wireframe is drawn, before anyone has agreed on what the site actually needs to do.
Most web projects that fail don’t fail in development. What separates them from the ones that succeed is usually something top web design agencies understand and average ones don’t: projects fail in the first two weeks — before a single wireframe is drawn, before anyone has agreed on what the site actually needs to do.
Why skipping discovery costs more than doing it
The instinct to get to design quickly is understandable. Clients have waited for this project. They want to see something. They want momentum. And agencies, particularly ones under competitive pressure to win work, sometimes accommodate that impatience rather than push back on it.
The result is a design process built on assumptions. What the site should communicate, how visitors actually behave, what the internal stakeholders really need it to do — all of this gets guessed at rather than established. The design looks fine. Then the review rounds start, and the feedback is contradictory because different stakeholders have different expectations that were never reconciled. Scope creep accumulates. Timelines slip. The project that was supposed to take twelve weeks takes twenty.
Discovery isn’t overhead. It’s risk management. The time spent clarifying objectives, understanding users, and mapping the existing situation before design begins is time that prevents far more expensive problems downstream. Agencies that have genuinely refined their discovery process tend to have significantly higher first-draft approval rates — not because they got lucky, but because by the time design starts, everyone has agreed on what it’s trying to achieve.
What a superficial discovery looks like
Knowing the difference between real discovery and a performance of discovery is useful, because both involve meetings and documents and look similar from the outside.
A superficial discovery is characterised by questions that could have been answered by reading the client’s existing website. What does your company do? Who is your target audience? What do you like in terms of design? These questions generate information, but not the kind that changes design decisions. They’re the equivalent of asking a patient how they’re feeling rather than examining them.
A superficial discovery also tends to involve only one or two people on the client side — usually the marketing lead or the CEO — and assumes that their perspective represents the organisation. It ignores the fact that the people closest to customers (support teams, sales teams, account managers) often know things about what visitors need from a website that leadership doesn’t. It produces a brief that reflects the loudest voice in the room rather than the actual situation.
The output of a superficial discovery is a document that could have been written before the meetings happened. It confirms the stated objectives, lists the pages that need to be designed, and describes the target audience in demographic terms. It doesn’t change how the agency thinks about the project. The design that follows reflects what the agency already knew how to do, not what this specific client actually needs.
What real discovery actually involves
A genuine discovery process starts with a different assumption: that the brief is a hypothesis, not a specification. The client thinks they know what they need. The agency’s job is to test that hypothesis against reality before committing to a design direction.
This involves several things that don’t happen in superficial discovery.
Talking to more than one voice inside the organisation. Sales hears objections. Support knows the recurring issues. Marketing has the data. The CEO has a vision. These are rarely the same thing, and the gaps between them are often where the most important design decisions live. An agency that interviews only one stakeholder is building for that person’s perspective, not for the organisation’s actual situation.
Looking at the existing data before forming opinions. If the client has analytics, session recordings, or any behavioural data about how people currently use their site — that information is worth more than any amount of stated preferences. A page with a high exit rate is telling you something. A section that nobody clicks is telling you something. Understanding the current situation analytically, rather than relying on the client’s narrative about it, produces a fundamentally different brief.
Asking about failure, not just goals. What has the current site failed to do? What have previous redesigns not solved? What feedback do you consistently receive from customers that the site doesn’t address? These questions are less comfortable than “what are your goals” but they surface the real constraints — the things that have persisted through previous attempts and will persist through this one if they’re not addressed directly.
Mapping the decision-making structure. Who needs to approve design decisions? Who can veto them? Are there stakeholders who aren’t in the room but whose opinions will surface during review? Discovery that doesn’t establish the approval chain before work starts is setting up a situation where the wrong person sees the design at the wrong stage and introduces requirements that should have been captured earlier.
Understanding the technical and content reality. What systems does the site need to connect with? What content exists, and what needs to be created? Who will maintain the site after launch, and what are their actual technical capabilities? These questions aren’t as interesting as the strategic ones, but they determine what’s buildable — and discovering a critical integration requirement after design is approved is one of the most expensive mistakes a project can make.
The questions that distinguish a serious agency
The quality of an agency’s discovery can be read in the questions they ask. Not the obvious ones — those tell you nothing — but the ones that reveal they’ve thought carefully about where projects go wrong.
What would make this project a failure? This question inverts the standard goal-setting conversation and surfaces constraints and concerns that wouldn’t emerge from “what are your objectives.” It also tells you what the client is most anxious about, which is usually different from what they lead with.
Who on your team will feel the most impact if this doesn’t work? Not who owns the project — who lives with the consequences. The answer often points to stakeholders who need to be in the room earlier than the agency assumed.
What have you already tried? If this is a redesign, there was a previous site, and presumably some thinking about what wasn’t working. Understanding what’s been attempted — and why it didn’t solve the problem — is one of the fastest ways to understand the real constraints.
If the site launches and performs well, what does that look like in six months? This forces specificity about success metrics. “More leads” is not a success metric. “A 20% increase in qualified demo requests from mid-market companies” is. The specificity of the answer tells you how clearly the client has thought about what they’re asking for.
Who are the visitors you’re currently failing? Not just the target audience in the abstract — the specific segment that has the highest intent and the worst experience on the current site. This question tends to produce more useful design direction than general audience descriptions.
What discovery produces — and what it doesn’t
A well-run discovery phase produces a set of clear, agreed-upon foundations: what the site needs to achieve, for whom, how success will be measured, and what constraints the design needs to work within. It produces alignment across stakeholders before design starts, so that review rounds are about execution rather than reopening fundamental questions.
What it doesn’t produce is a design. Clients who expect to see visual concepts at the end of discovery have misunderstood what the phase is for. Discovery is the work that makes the design possible — the research and alignment that gives the design team something specific to solve rather than a vague aspiration to respond to.
The agencies that do this well tend to treat discovery as a deliverable in its own right, not as a preamble to the real work. They document what they found. They present it back to the client as a shared understanding of the situation. They get sign-off before moving forward.
That sign-off is more valuable than it might seem. It means that when design presents a direction that surprises someone, the response is “this doesn’t match what we agreed” rather than “this isn’t what I imagined.” The first conversation is productive. The second is not.
