How to Prototype for User Testing

“If I have a thousand ideas and only one
turns out to be good, I am satisfied.”
– Alfred Bernhard Nobel

UXD prototyping is a robust topic and difficult to adequately cover in this type of overview guide. Therefore, I do so with the caveat that I am only providing the tip of the iceberg to get you started, with some references to where you can learn more.

Prototyping is not a research method, it is a research tool. An important thing to understand about UXD prototypes is that the term “prototype” itself can mean something slightly different to the UXD community than to the software development community at large. For example, one of the more common engineering usages of the term refers to an operational prototype, sometimes called a proof of concept. This is a fully or partially functional version of an application. UXD prototypes are, however, not usually operational. Most often they are simulations focused on how a user might interact with a system. In the case of the “paper prototype”, for example, there is no functionality whatsoever, just paper print-outs of software screens.

There are many different names for types of prototypes in software engineering, most of which describe the same (or very similar) concepts. Here are some of the most common terms:

  • Operational – Refers to a fully or partially functional prototype that may or may not be further developed into a production system. User testing prototypes can be operational, such as during late stage validation testing, though a program beta is more commonly used at this stage.
  • Evolutionary –As the name implies, an evolutionary prototype is developed iteratively with the idea that it will eventually become a production system. User testing prototypes are not “evolutionary” in the strictest sense if they do not become production systems. However, they can be iterative and evolve through different design and testing cycles.
  • Exploratory– Refers to a simulation that is intended to clarify requirements, user goals, interaction models, and/or alternate design concepts. An exploratory prototype is usually a “throw-away” prototype which means it will not be developed into a final system. User testing prototypes would usually be considered exploratory.

Semantics aside, prototypes are probably the single most powerful tool for the researcher to understand user behavior in the context of the product being developed. Some commonly used prototypes in user testing are;

  • Wireframes – A wireframe is a static structural description of an interface without graphic design elements. Usually created in black, white and grey, a wireframe outlines where the content and functionality is on a screen. Annotated wireframes are wireframes with additional notes that further describe the screens’ content or interactivity.
  • Design Mockups – Similar to wireframes, a design mockup is a static representation of an interface screen. However, design mockups are full color descriptions that include the intended graphical look and feel of the design.
  • HTML Mockups –Used in web site design, HTML mockups refer to interface screens that are created in the Hypertext Markup Language and so can be viewed in a browser. Most often, HTML mockups simulate basic functionality such as navigation and workflows. HTML mockups are usually developed with wireframes or a simplified version of the intended look and feel.
  • Paper Prototype – A paper prototype is literally a paper print-out of the designed interface screen. A paper prototype could be of wireframes or design mockups. In addition, it could be one page to get feedback on a single screen, or a series of screens intended to represent a user task or workflow.
  • PDF Prototype – A PDF Prototype consists of a series of designed interface screens converted into the Adobe Acrobat (.pdf ) file type. Like HTML prototypes, PDF prototypes can simulate basic functionality such as navigation and workflows. However, they take less time to create than HTML prototypes and can be created by someone without web development skills.
  • Flash Prototype – A prototype developed using Adobe’s Flash technology. Flash prototypes are usually highly interactive, simulating not only buttons and workflows, but the system’s interaction design as well. In addition to being a quick prototype development method, Flash prototypes can be run via the web or as a desktop application making it very portable tool.

Why Use Prototyping

  • It saves money by allowing you to test and correct design flaws before a system is developed.
  • It allows for more freedom to explore risky, envelope-pushing ideas without the cost and complexity of developing it.
  • Since prototypes are simulations of actual functionality they theoretically bug-free. Test results are less likely to be altered or impeded by implementation issues.

When is a Prototype Most Useful?

  • When you are trying to articulate a new design or concept
  • When you want to test things in isolation (i.e. graphic design separate from information layout separate from interaction design)
  • To gather user feedback when requirements are still in a state of flux and/or can’t be resolved
  • To evaluate multiple approaches to the same user task or goal to see which users prefer

Limitations of Prototyping

  • A prototype will never be as accurate as testing on a live system; there is always some deviation between the real world and the simulation.
  • Depending on where you are in the iterative research process, there is a point of diminishing returns where the amount of effort to create the prototype is better put into building a beta.

Creating a UXD Prototype

Because the actual prototype creation process is highly specific to what you are creating and what tools you are using (paper napkins, layout tool, whiteboard, etc) I’ve included a few considerations for defining a prototype instead of detailing the mechanics of creating one.

  1. Consider your Testing Goals. – Are you looking to understand how users perform a specific set of tasks? Do you need to watch users interact as naturally as possible with the system? Or are you trying to get users’ responses to various experimental ideas and concepts? The answers to these questions will help you make some key decisions about the structure of your prototype.
  2. Decide on Degree of Fidelity. What level of fidelity will the prototype achieve? Here, the term fidelity refers to the degree to which a prototype accurately reproduces the intended final system. A low fidelity prototype might be a PowerPoint deck of wireframes. A high fidelity prototype could be an interactive simulation with active buttons and representative content.A good rule of thumb is to develop the lowest fidelity prototype possible to achieve the goals of your study. This ensures a lack of commitment to the ideas presented and allows more time and money for the recommended iterative process. If a significant amount of time is taken to create an initial prototype with all the bell and whistles, designers are less willing to see when the concept is not working, less likely to change their designs. In addition, the amount of time that goes into building one high fidelity prototype would have been better used building multiple lower fidelity versions that allowed for more testing in-between each revision.
  3. Scripted Tasks or Natural Exploration? – Another consideration when defining your testing prototype is what content and functionality should be included. Will a preset walkthrough of key screens be sufficient, or do the goals of your study require that users are able to find their own way around the system? On average, I tend to think that enough insight can be gleaned from a series of walkthrough tests and other research methods to warrant using these, leaving the open-ended user-directed tests to be conducted on a product beta or via A/B testing[1] on a live system. With a sufficiently complex system you can quickly hit a point of diminishing returns regarding the amount of effort it takes to simulate functionality vs. actually building it.

Additional Resources

  • Microsoft Visio (office.microsoft.com/en-us/visio) – The “old school” standby for wireframes and workflows.
  • Adobe InDesign (www.adobe.com/products/indesign) – Arguably the industry standard tool. Layout tool with key functionality conducive to prototype development.
  • Balsamiq – (www.balsamiq.com) – Good low-cost alternative when you only need the basics.
  • Further Reading – Rettig, Marc. “Prototyping for Tiny Fingers.” Communications of the ACM. Vol. 37, No.4. April, 1994.


[1] A/B Split Testing refers to a testing method where in a live system an alternate, experimental design is shown to a percentage of users while the baseline control design is shown to the rest. Analysis comes from observing any notable differences in user behavior between the experimental design and the control.