All methods

Looks Like / Works Like Prototype

Typical duration:
30m

A dual-prototype approach where one prototype shows what the solution looks like (form, aesthetics) and a separate one demonstrates how it works (function, mechanics), enabling focused feedback on each dimension.

Purpose

The Looks Like / Works Like approach splits a concept into two separate prototypes: one that demonstrates visual design and form, and another that demonstrates functionality and logic. This separation prevents conflation of aesthetic and functional feedback, allowing clearer, more actionable insights on each dimension.

At a glance

  • Group size: 2–8
  • Duration: 30 minutes
  • Difficulty: Moderate
  • Facilitation style: Parallel prototype building
  • Setting: Workshop table with both visual and functional prototyping materials

How to run

  1. Identify the two dimensions (5 min): Clarify what "looks like" means for this concept (visual style, form factor, branding) and what "works like" means (interactions, logic, data flow).
  2. Split and build (15 min): One sub-team builds the Looks Like prototype (visual mockup, rendered image, physical model). Another builds the Works Like prototype (clickable wireframe, mechanical demo, role-play script).
  3. Test separately (5 min each, 10 min total): Show the Looks Like prototype first: "Does this feel right? Does it fit the context?" Then show the Works Like prototype: "Does this make sense? Is anything confusing?"
  4. Synthesise (5 min): Compare feedback. Are there conflicts between what users want it to look like and how they want it to work?

Materials needed

  • Visual prototyping: paper, markers, foam, 3D printing, or Figma
  • Functional prototyping: clickable wireframes, mechanical mockups, or code
  • Separate feedback forms for each dimension

Pitfalls & common mistakes

  • Testing both together: The whole point is separation. A beautiful prototype that also works will get conflated feedback.
  • Neglecting one dimension: Teams tend to over-invest in the one they are more comfortable with.
  • "Looks Like" = final design: This is a test, not a delivery. Keep fidelity appropriate.
  • Never reuniting the two: Eventually the dimensions must merge. Plan for integration.

Inclusion & accessibility

  • Describe the Looks Like prototype verbally for visually impaired participants.
  • Ensure the Works Like prototype is operable by participants with different abilities.
  • Accept feedback in multiple formats.

Variations

  • Looks Like / Works Like / Feels Like: Add a third prototype that tests the emotional experience.
  • Competitive Looks Like: Create Looks Like prototypes in two different visual styles and compare.
  • Looks Like / Works Like sprint: Build both in a single 2-hour sprint session.

When NOT to use this

  • When the concept is too simple to separate into two meaningful dimensions.
  • When time only allows one prototype — build a combined one at an appropriate fidelity level.
  • When the two dimensions are so tightly coupled that separating them would be misleading.

Source & further reading

  • Houde & Hill — "What do Prototypes Prototype?" (Handbook of HCI, 1997)
  • d.school, Stanford — prototyping frameworks
  • Buxton, Bill — *Sketching User Experiences* (Morgan Kaufmann, 2007)

Turn the building blocks into an agenda

Create the day, drag the blocks into the order you have in mind, and let the times work themselves out. Or describe what you are planning to your AI and let it write the first draft.

Try it free for 14 days
Looks Like / Works Like Prototype · GoodWorkshop