In UX design, what distinguishes a sketch, a wireframe and a prototype, and which question is each best at answering?
answer
- explore, structure, behave
- many rough ideas, fast
- layout and priority, no styling
- simulates behaviour to try a task
- fidelity is a separate axis
basics
~20 sA sketch explores many rough ideas quickly; a wireframe fixes structure, content priority and layout without visual styling; a prototype simulates behaviour so people can attempt a task. Each answers a different question, at whatever fidelity that question needs.
solid answer
~40 sThey differ in **what they commit to**. A **sketch** is quick and disposable, for exploring many alternatives — should a car-rental results page lead with vehicle class or with price? A **wireframe** commits to **structure**: which content appears, in what priority and roughly where, usually kept in greyscale so discussion stays on layout and hierarchy rather than colour. A **prototype** commits to **behaviour**: it simulates the product so someone can attempt a task — find a pick-up location, compare two cars — while the team watches what happens. A **mockup**, often confused with these, is a static picture of the intended visual design. Fidelity is a separate axis: a prototype can be paper or close to production. The practical skill is choosing the cheapest artefact that answers the question the team has right now.
go deeper
Recall what each artefact commits to: a sketch to an idea, a wireframe to structure, a prototype to behaviour, and a mockup to visual design.
Explain why fidelity is a separate axis with several dimensions, and how to choose the cheapest artefact that answers the current question.
Show how you steer a team away from polishing the wrong artefact, and when to skip steps because the question is already answered.
Set expectations across product and engineering about what each artefact is for, so early exploration is not mistaken for committed scope.
## Three artefacts, three commitments Sketches, wireframes and prototypes are often treated as steps on one ladder of increasing polish. It is more useful to see them as artefacts that **commit to different things**, and therefore answer different questions. Choosing the wrong one wastes effort: polishing an idea that should have been one of ten sketches, or debating colours on a layout whose structure is still wrong. ## Sketches A **sketch** is a fast, rough drawing of an idea. - It is cheap enough to make many — commonly several alternatives for one problem in minutes. - It is **divergent**: its job is to widen the options before narrowing them. - It commits to almost nothing, so people feel free to discard it. - On a car-rental site, a page of sketches might compare results led by vehicle class, by price, by a map of pick-up points, or by a single recommended car. ## Wireframes A **wireframe** is a skeletal layout of a screen. - It commits to **structure**: which content and actions appear, their priority, and their approximate placement. - It deliberately leaves out visual design — colour, imagery, typography — so reviews stay on hierarchy and flow. - It is usually static, one screen or a set of screens. - A results-page wireframe settles, for instance, that price, seats, luggage capacity and cancellation terms appear on every card, with price most prominent. ## Prototypes A **prototype** is a simulation of the product that someone can **use**. - It commits to **behaviour**: what happens when a person acts. - It exists to be tried, usually against a task, so the team can observe rather than ask. - It can be paper, linked screens, or code — fidelity is independent of the fact that it is a prototype. - A pick-up search prototype lets a person type a city, see locations and choose one, even if the data is faked. ## Side by side | Artefact | Commits to | Best question it answers | Typical effort | Weak at | |---|---|---|---|---| | Sketch | An idea | Which directions are worth pursuing? | Minutes | Detail, behaviour | | Wireframe | Structure and priority | Is the right content here, in the right order? | Hours | Look and feel, interaction | | Mockup | Visual design | Does it look right and on-brand? | Hours to days | Behaviour | | Prototype | Behaviour | Can people do the task, and where do they struggle? | Hours to weeks | Depends on its fidelity | ## Fidelity is a separate axis **Fidelity** is how closely an artefact resembles the finished product, and it has several independent dimensions: - **visual** — from line drawings to final styling; - **content** — from placeholder text to real copy, prices and data; - **interaction** — from static images to responsive, stateful behaviour; - **breadth and depth** — how many features are covered, and how fully each works. A wireframe is usually low on visual fidelity but can carry real content. A prototype can be low fidelity everywhere — paper screens swapped by hand — or high on interaction and low on visuals. ## Where each fits on a real project On a car-rental site, a team might sketch a dozen ways to present search results in an afternoon, wireframe the two most promising to settle what each result card must show, link the wireframes into a rough prototype to see whether people can get from search to a chosen car, and only then commission mockups of the chosen direction. Steps can be skipped when the question is already answered — a small change to an existing, well-understood page may go straight to a mockup — and repeated when a finding reopens an earlier question. ## Choosing the artefact 1. **Name the question** the team needs answered this week. 2. **Pick the artefact whose commitment matches it**: directions → sketches; structure → wireframes; behaviour → a prototype; look → a mockup. 3. **Set each fidelity dimension** only as high as the question needs. 4. **Throw away freely**: the artefact's value is the answer, not the file. 5. **Move on** once the question is answered, rather than polishing the artefact that answered it. The same logic applies to a website, a native mobile app or a kiosk: the artefacts are about the kind of commitment being tested, not the platform or the tool used to make them.
- Is a wireframe just a low-fidelity prototype?Not quite. A wireframe commits to structure and is usually static; a prototype commits to behaviour and is meant to be used. Linking wireframes together turns them into a low-fidelity prototype, but a single wireframe answers “is the right content here in the right order?”, not “can people complete the task?”.
- Why keep wireframes in greyscale?Colour, imagery and type draw attention and invite opinions about taste, which pulls a review away from hierarchy and flow. Greyscale signals that the visual design is not decided yet, so people discuss what content appears and in what priority. It is a convention with that reason, not a rule.
saying these in an interview costs you the question
- A prototype must be high fidelity; anything rough is only a sketch.
- Wireframes should include final colours so stakeholders can approve them.
- Sketching is a waste of time once a design editor is available.
- A mockup and a prototype are the same artefact.
- Each project must pass through sketch, wireframe, mockup and prototype in order.