Think about the last time you signed up for something and closed the tab a minute later. You probably didn’t decide the product was bad. You just weren’t sure what to do next.

That’s the job of onboarding. It isn’t a welcome tour or a pile of tooltips. It’s the shortest path from “I just signed up” to “oh, I get it.”

What is onboarding, really?

Onboarding is everything a new user sees and does from their first visit until the product proves its value to them for the first time. That moment is the first win: the first message sent, the first project created, the first result they can see.

A good first win is small, specific and fast. If you can’t name yours in one sentence, your onboarding probably has nothing to aim at yet.

What should a new user see in the first minute?

Three things, in this order:

  1. Where they are. A clear heading and one obvious next action.
  2. What’s coming. A short list of the steps between here and the first win.
  3. That they’ve already started. At least one step shown as done.

Everything else can wait. If a screen asks the user to read, choose and fill in more than one thing, it’s doing too much.

How do guided steps keep people from getting lost?

Turn the path into a short checklist and keep it visible. Something like:

  • ✓ Create your account
  • Add your first project
  • Invite a teammate
  • Publish

Each step is one clear action. When someone finishes one, it checks off and the next one is highlighted. At any moment the user can answer three questions without thinking: what have I done, what’s next, and how much is left?

A few rules I follow:

  • Keep it to three to five steps. Longer lists feel like homework.
  • Make every step a real action in the product, not “read the guide.”
  • Let people skip and come back. A checklist that blocks you stops being help and becomes a wall.
  • Show the checklist where they work, not in a separate tour they’ll dismiss.

Why should step one already be done?

Because people are more likely to finish something they’ve already started.

In a well-known study by Nunes and Drèze, 300 car wash customers got one of two loyalty cards. One was an eight-stamp card. The other was a ten-stamp card with two stamps already filled in, so both groups needed the same number of washes. Over nine months, 34% of the pre-stamped group completed the card, compared with 19% of the other group. The researchers called it the endowed progress effect.

You can use the same idea honestly. If your product asks for an email and password, that’s real progress, so show it: “Account created ✓. 3 steps left.” The user didn’t just sign up. They’re already part of the way through setup.

The word honestly matters. Don’t fake a step the user didn’t take. Count the real ones.

What should you cut from sign-up?

Anything that isn’t needed to reach the first win. Every extra field, from phone number to company size to “how did you hear about us,” is a reason to leave before the product has had a chance to show its value.

Ask for the minimum, then collect the rest later, in context, when it’s clearly useful.

Sometimes the best move is to skip the account entirely at first. When I designed Sampled Planet, adding a song pair needs no account and no password. On Legasky, a private life tracker, there’s no account at all and everything stays in the browser. Neither product asks for trust before it has earned any.

When do tooltips and tours help?

Tooltips help when they appear at the moment a person needs them, attached to the thing they’re about to use. A five-screen tour shown before anyone has touched the product mostly gets skipped, and what people skip, they forget.

A simple test: could the tip be replaced by clearer labels or a better empty state? If yes, fix the screen instead.

Speaking of empty states, they’re the most valuable onboarding screen you’ll design. The first thing a new user sees is usually a blank list. Use it to say what goes here, why it matters and what to do first, with one clear button.

How do you show progress without making it stressful?

Keep it simple and honest. Nielsen Norman Group suggests that showing the number of steps can work better than a percentage, because it helps people estimate the time left even when they don’t know how long each step takes. “Step 2 of 4” is clearer than “50% complete.”

Feedback matters just as much. When someone completes a step, confirm it right away with a checkmark or a short message. When nothing visibly happens, people assume it didn’t work and try again.

What are the most common onboarding mistakes?

  • A long form before any value is shown.
  • An upfront tour about features instead of the first win.
  • A checklist with eight or ten items.
  • No way to skip, pause or return later.
  • Empty screens with no direction.
  • Progress that restarts or disappears if the user leaves and comes back.

Each one has the same root cause: designing from the company’s point of view, “what do we need to tell them,” instead of the user’s, “what do I need to do next?”

How do you know your onboarding works?

Pick the first win, then measure two things: how many new users reach it, and how long it takes. Then look at each step of your checklist and see where people stop. The step with the biggest drop is your next design task.

Watching a few real people use it for the first time is just as useful. Ask them to get set up without your help and note where they pause, reread or go back. Those pauses are your list of fixes.

Quick answers

What is good onboarding in a product? A short, guided path from sign-up to the user’s first win: one clear action at a time, a visible checklist and progress, and nothing that isn’t needed yet.

How many steps should onboarding have? Usually three to five. If the first win needs more, split the work and let people finish the rest later.

Should onboarding include a product tour? Rarely as an upfront tour. Short, contextual tips shown when the user reaches that part of the product work better.

Does onboarding matter for an MVP? Yes. A first version has no brand recognition to lean on, so the first minute decides whether people stay. It’s one of the states an MVP needs before launch.

Never leave people guessing

People will forgive a missing feature. They won’t stick around if they can’t tell what to do next.

If you want new users to get to their first win faster, that’s exactly the kind of product design work I enjoy most.