Ask people why a website failed and you will hear about the wrong colours, a slow page, a clumsy menu. Look closer and the real cause is almost always older than any of that. Somewhere at the start, a question went unanswered, and every decision afterwards was built on a guess.
The good news is that the questions are few, and none of them needs special tools. They need an hour, a quiet room and the willingness to say things out loud that everyone assumed were obvious.
What is this page for?
Not "what is the website for", which always gets a grand answer. Pick the single page that matters most and ask what a visitor should be able to do or understand after thirty seconds there. If the answer is a list of five things, none of them will happen.
A useful trick is to finish this sentence: "If someone only remembers one thing, it should be ___." Then check whether the current page would actually make them remember it. Most can't pass this test, and that is fine. It is exactly why you ask.
Who is actually visiting?
There is a big difference between the people you imagine and the people who arrive. Open your analytics, or ask the three customers you talk to most. Where do they come from? What device are they holding? What were they trying to solve right before they typed your name?
A site built for a vague "target audience" reads like a brochure. A site built for a specific person on a specific Tuesday afternoon reads like help.
What do people already believe?
Every visitor arrives with a belief about you, formed by a referral, a search result, a competitor, or a bad experience with a similar company. Your job is rarely to inform from scratch. It is to confirm what is true, correct what is not, and do it before they lose patience.
Write down the three most common misunderstandings about what you do. Then look at your homepage and ask whether it addresses even one.
What are you willing to leave out?
This is where most projects quietly go wrong. Adding is easy: every stakeholder has one more thing that must be on the page. Subtracting is hard, because it requires choosing, and choosing means someone is disappointed.
A good rule is that every element has to earn its place by serving the goal you wrote down at the start. If you can't say which goal a section serves, it is decoration, and decoration costs attention that your real message needs.
How will you know it worked?
Before launch, decide what success looks like in numbers you can check: enquiries per month, time to find a price, how many people reach the contact form. Pick two or three, no more.
Without this, the review after launch becomes a debate about taste. With it, the conversation is about evidence, and taste goes back to being a tool rather than a weapon.
The cheap tests that save the expensive mistakes
You do not need a research budget to learn most of this.
- The five-second test. Show the page to someone unfamiliar for five seconds, then ask what the company does. Wrong answers are data.
- The sticky-note sort. Write every piece of content on a note and sort them by importance without looking at the current site. The pile that ends up at the top is your homepage.
- The competitor swap. Cover the logo and see whether the page could belong to a rival. If it could, you have not said anything distinctive yet.
- The one-sentence brief. If you cannot explain the project in a sentence to a friend, nobody building it can either.
Why this feels slower, and isn't
Clarity work is invisible. There is nothing to show in a meeting, no pretty screen to share. It is tempting to skip it and jump to something tangible.
But the cost of an unanswered question grows with every decision made on top of it. A change in week one is a conversation. In week eight it is a rebuild. The hour you spend at the start is the cheapest hour of the whole project, and the only one that gets more expensive the longer you wait.
If you remember nothing else, remember this: most design problems are decision problems in disguise. Solve the decision first, and the design usually gets easier than you expected.