5. User experience

§Ask a founder whether their startup does user experience and they will point at the interface. That is the first misunderstanding, and it is expensive, because it puts the effort at the end of the process where it can change the least.

§User experience is the whole of what a person goes through to get a job done with your product: how they found it, what they expected, what they had to learn, where they got stuck, whether they came back. The interface is the last and thinnest layer of that. Don Norman, who coined the term in the 1990s, has spent years complaining that it got narrowed to screens.

§ 5.1The five planes#

§The most useful way to see this is still Jesse James Garrett's five planes, from The Elements of User Experience. Read from the bottom up, they are the order in which experience gets decided.

§Strategy. What the product is for and who it is for. The user's need and the company's objective. This is Product Strategy's territory, and it is the base of the experience: a product for the wrong person has a bad experience no matter what the screens do.

§Scope. What the product does and does not do. Which features exist. This is where most of the experience is actually set, because scope is what determines whether the user's job is finished or half finished.

§Structure. How the pieces fit together: what leads to what, what the user sees first, how they get from any point to any other. Information architecture and interaction design live here.

§Skeleton. The arrangement of elements: where the button is, what the navigation contains, which fields the form has.

§Surface. What it looks like: type, colour, spacing, motion. The interface.

§The practical consequence for a startup is that the product owner is doing UX every time they decide what the first version includes and how its parts connect. Those decisions are the experience. A designer joining later can make the surface beautiful and the skeleton clear, but cannot rescue a scope that leaves the job half done.

§ 5.2Seven questions to ask of anything you ship#

§Peter Morville's honeycomb is the other tool worth keeping. It lists seven qualities a product needs for its experience to be good, and it works as a checklist before anything ships.

§Is it useful: does it do a job someone actually has? Is it usable: can they do that job without help? Is it findable: can they get to what they need? Is it credible: do they believe it will work and that their data is safe? Is it desirable: do they want to use it, or merely tolerate it? Is it accessible: can people with different abilities use it? And is it valuable: does it deliver enough to the user and the business that both keep showing up?

§The order in which those fail at a startup is predictable. Useful comes first, and it is the one only discovery can settle. Usable and findable are the ones an owner without a designer gets wrong most often, because they know the product too well to see where a stranger gets lost. Credible is the one AI products fail most today, since a model that is confidently wrong once costs more trust than ten right answers earn back.

§ 5.3When there is no designer#

§Most early-stage teams have no UX designer, and the owner holds the discipline by default. That is workable, on three conditions.

§First, decide scope with the user's whole job in view, not the feature. "Can a customer be recognized when they arrive" is a job; "loyalty card" is a feature that may or may not do it. Second, draw the structure before anyone builds: on paper, as boxes and arrows, so that the path from landing to done is visible and the dead ends show. Third, watch a stranger use it. Not a friend, not a colleague: someone who owes you nothing, given a situation rather than instructions, while you stay silent. Product Sprint covers how. Here the point is that the watching is not optional, and that it is the owner's job to do it, not a designer's job to do later.

§ 5.4Where this hands off#

§UX is decided at strategy, scope and structure. Strategy is Product Strategy's whole subject. Scope and structure are what a written position from a sprint gives you: which questions the prototype answered, which flows broke in a stranger's hands. The next chapter is about the plane that gets all the attention and deserves the least until the others are settled: the surface.