Most of the sites we end up rebuilding didn’t have a design problem or a technology problem. They had a problem that nobody said at the start what the site was actually for. So you get a site that wants to do everything: sell, educate, collect emails, show references, run a blog, host a forum, and throw in a quiz too. And because it does everything, it does nothing well.

This lesson is about how to avoid that. Before you open Figma or start picking technologies, you need three things: one main goal, one audience, and one metric that tells you whether it’s working.

One main goal, not five

A site can have several features, but only one main goal. That’s the action you want a visitor to take above all else. Typically one of these:

Call or send an inquiry (services, trades, B2B)
Buy (online store, digital product)
Sign up or leave an email (apps, newsletters, waitlists)

Everything else on the site should support that one goal, or it has no business being there. References aren’t there for their own sake; they’re there to build trust so someone works up the nerve to call. The blog isn’t there for fun; it’s there to bring in people who’ll become customers over time.

If you don’t get clear on the goal, design won’t fix it. Quite the opposite. Everyone on the team and on the client’s side starts pushing their favorite thing, and the page starts to bloat. You add a mega menu, three kinds of forms, a slider with ten slides, and a newsletter popup that covers the very button people came to click. A blurry goal always gives itself away in how many unnecessary things the site ends up containing.

Who is it for

The second question: who exactly is supposed to take that action? Not “everyone,” not “people aged 25 to 65.” A specific person in a specific situation.

The difference is enormous. A page for a restaurant owner shopping for a delivery system looks and talks completely differently than a page for a parent looking for an after-school class for their kid. Different language, different objections, a different level of technical detail, a different decision speed. When you write for a specific person, you suddenly know what belongs on the page and what’s just noise.

It helps to write down three things that person is dealing with the moment they land on your site: what they need to know, what they’re afraid of, and what will convince them to act. Those three statements will save you hours of arguing over copy later on.

One success metric

A goal without a metric is a wish. You need one number that tells you, three months from now, whether the site is doing its job.

Goal “call” → number of inquiries / calls submitted per month
Goal “buy” → conversion rate, i.e. what percentage of visitors purchase
Goal “sign up” → number of registrations per week

Watch out for vanity metrics. Visits, likes, or “time on page” look nice, but they don’t pay the bills. Measure what’s directly tied to your main goal. And set up tracking from day one, not as an afterthought, because nobody can fill in data you never collected.

Exercise: your goal in one sentence

Grab a piece of paper or a blank document and finish this sentence until the whole thing makes sense:

This site exists to get (who) to (do what), which we’ll measure by (which metric).

Example: “This site exists to get owners of smaller restaurants in Brno to call for a no-obligation quote on a delivery system, which we’ll measure by the number of form inquiries per month.”

Rules of the exercise:

One sentence, one goal. If you cram two actions into it (“to buy, or at least sign up”), you have two sites and neither of them done properly.
No “and.” The word “and” is the first sign of a blurry brief.
Keep the sentence handy. Come back to it with every decision you make (what goes on the homepage, whether to add a section, what color a button should be). If a given thing doesn’t help that sentence, it doesn’t belong.

Once you have that one sentence, you’ve finished the most important part of the whole project. The rest of the course is just about how to deliver on it as well as you can.