The First Ninety Seconds of Arriving

You come down an escalator into the arrivals hall at Vancouver and the first thing in front of you is a pair of carved welcome figures, standing where the sightline lands. Behind them, the queue. It is the most honest onboarding flow I know of: one chance, no instructions, every user tired, carrying something, and unable to leave. Airports have been solving this problem for decades and software companies keep solving it from scratch and worse.

What does an arrivals hall get right?

It answers where am I before it asks anything. The sightline from the escalator lands on something that says which country you are in and what kind of place this is, and only after that do you encounter the process. Orientation, then instruction, then transaction, in that order, every time.

It also assumes depletion. Every design decision in that hall assumes the user is tired, carrying luggage, possibly in the wrong time zone and possibly unable to read the local language. Nothing depends on the user being sharp. Compare that with a product tour that assumes someone will read six tooltips in sequence.

And it has no exit. The user cannot leave, which sounds like an advantage until you realise it is why airports invest so heavily in orientation. A trapped, confused user becomes an angry one. A software user just closes the tab, which is cheaper for them and much more expensive for you.

Arrivals hall against typical product onboarding.

Design question	Airport answer	Common product answer
First thing the user sees	Where they are	What a feature does
Assumed user state	Tired, carrying, distracted	Rested, focused, curious
Primary mechanism	Wayfinding signs	Sequential tooltips
Empty state	Signposted, staffed	Apologetic and blank
Success in the first minute	Pointed in the right direction	Completed a tour

Why does most product onboarding invert the order?

Because it is designed by people who already know where they are. When you know the product, orientation seems trivial and the interesting problem is teaching features. So the tour starts with what things do rather than with what this place is, and the new user is being taught the controls of a building they have not yet located themselves in.

The symptom is a tour that gets dismissed. Dismissal rates are usually treated as a design failure of the tour. More often the tour is answering a question the user has not reached yet.

What does wayfinding look like in software?

It looks like structure being visible before it is used. A navigation that shows the whole shape of the product on first load rather than progressively revealing it. Empty states that describe what will be here rather than apologising for being empty. A first screen that makes the map obvious even though nothing has been done yet.

The airport version of this is the overhead sign you can read from thirty metres away, which tells you the three things that exist and roughly where they are. It does not explain immigration policy. It gets you pointed correctly and lets the details arrive when you are close enough to need them.

We ran an onboarding review on an account where the CRM audit found roughly 98 percent of contacts sitting in the first lifecycle stage with nothing behind them. That is the same failure at company scale. Everyone arrived and nobody was pointed anywhere, so they stayed at the door.

How do you design for a depleted user?

Assume no reading. Assume one hand. Assume interruption. Then check whether the first minute still works. Most onboarding survives none of those assumptions, because it was tested by people who were paying full attention and had a reason to persevere.

Then cut the first session to a single outcome. Not a tour of five features, one thing the user came to do, completed. Everything else can wait for the second session, and if there is no second session then the five features were never going to matter anyway.

The first screen is not an introduction to the product. For most of the people who will ever see it, the first screen is the product.

What is the first ninety seconds worth?

Disproportionately much, and for a reason borrowed from how machines read pages as well as how people do. Research on generative engine visibility found that 44.2 percent of citations come from the first 30 percent of a page. Attention front-loads. What is at the top gets used and what is below it frequently does not exist.

The practical version is that the first screen is not an introduction to the product. It is the product, for most of the people who will ever see it. Treating it as a preamble is how good products lose users who never found out they were good.

Related: auditing the funnel before adding traffic, context beats explanation and field notes from the same run of trips.