For the person using it, the screen is all there is
Users never see the architecture. They see buttons, words and waiting, and that is the product they judge.
To the user, the interface is the product.
The idea is older than this phrasing. Jef Raskin, who started the Macintosh project at Apple and was Aza Raskin’s father, wrote nearly the same sentence in his 2000 book The Humane Interface, speaking of the customer rather than the user. The son, who ran user experience at Mozilla Labs before co-founding Massive Health, cut it down to the version that circulates now.
Its meaning is simple and easy to lose inside a company. Engineers see the product as its architecture: the database, the sync engine, the clever ranking. Sales staff see a feature list. The person who actually uses it sees buttons, words, delays and error messages, and nothing else. If the ranking is brilliant but the result appears after four seconds of blank page, the user’s product is a blank page.
The iPhone in 2007 is the obvious case. Smartphones already browsed the web and handled email; the BlackBerry and Windows Mobile handsets of the day did both. What changed was the interface, and to buyers that was the whole difference. Google’s search page, nearly empty for a decade, makes the same point from the other side: an enormous system behind it, and for the user a single text box.
The line is weaker where the person using the product is not the person choosing it. Enterprise software bought by a purchasing department is judged on price, contracts, integration and audit requirements. The staff who use it daily may hate the screens and the product still sells. In those markets the buyer’s interface is the proposal and the invoice.
It also breaks when reliability fails. A handsome interface on top of a service that loses data is a bad product, even if the user cannot tell why it went wrong. People meet failures through the interface, but the fix sits below it. Teams that read the line as permission to polish the surface and ignore the plumbing have taken it the wrong way.
For developer tools it holds with a twist. The interface of a payments or storage service is its API, and developers judge that the way consumers judge a screen, by how little they have to think. Stripe, which opened to the public at the end of September, is being sold largely on how few lines of code it takes to accept a card.
The practical use is as a habit before release. Someone should sit with the product as a new user would, knowing nothing of how it works inside, and write down what they see. That list is the product. Everything else is how it gets made.