Michael Gaigg: Über UI/UX Design

Integrating Prototyping Into Your Projects

Posted by Michael Gaigg

This article was inspired by Integrating Prototyping Into Your Design Process - Using appropriate fidelity for the situation by Fred Beecher which I extend by the following:

Prototyping needs to be iterative throughout the project!

Goal of Prototyping

Prototyping is not only a design tool but a research and communication tool as well.

• It should assist in optimizing the main task (top tasks) and validating its/their efficiency.
• Furthermore this should not add cost to the project but reduce project expenses while increasing ROI.

So the goal is to use different levels of prototype fidelity to incrementally identify and enhance the user's task(s).

Ideally this happens linear (increase visual fidelity as you add functional fidelity) but typically it is bent to either side (see Figure 1) where more emphasize on

• visual fidelity can be beneficial for marketing purposes or
• functional fidelity can assist earlier user feedback trough user testing.

Figure 1: Prototyping in the context of your project.

Integration into your project

Regardless of the project approach you take it will boil down into the fundamental project management phases of Requirements, Design, Implementation (and possibly others). Prototyping should not be solely perceived as a method useful during Design, it is essential during all 3 (or more) phases starting as early as Requirements phase.

I suggest the following approach:

1. Low-fidelity prototyping (paper / digital sketch)
   1. Create paper prototypes or digital sketches
   2. Design navigation architecture (workflow)
   1. Review with client
   2. User testing (optional)
   3. Iterate (until happy)
   4. Revise into 2
2. Medium-fidelity prototyping (simple HTML)
   1. Simple HTML prototyping (maybe even black and white)
   2. Proof basic workflow and important interactions
   1. Review with client
   2. Iterate
   3. Revise into 3
3. High-fidelity prototyping (Enhanced HTML)
   1. Enhance HTML prototype (links and basic functionality)
   2. Settle on design (including corporate design, basic artwork)
   1. Review with client
   2. Iterate
   3. Revise into 4
4. Start 'real' implementation

Implementation Effort

Each prototype (digital sketch, simple HTML, advanced HTML based on simple) should not take more than 40 hours of pure development (not calculating initial meetings and communication and possible variations based on project size) plus 80 hours reviews and iterating with client. Sounds impossible? Think twice. It is so much easier to modify a sketch than programming HTML. The 'real' implementation will be built upon a solid code foundation with a grown-up design already - voila!

Can I skip a prototype?

Yes, obviously you can. But it comes with a cost later on because you miss crucial information from the earlier phase and it is more expensive to implement modifications.

Technical considerations

The argument I hear most often is that 'prototypes' are wasted time/money because they get trashed anyway. This is absolutely not true! Identifying problems early almost always saves money later on, you don't find anything out until you start showing it to people, enhancing the quality of the product will help money flow into your pocket once deployed and most important, prototypes don't necessarily need to and should not be trashed.

Low fidelity prototypes can be more than just ‘paper’, this could be digital wireframes that look like sketches, e.g. Microsoft offers software that tie sketches (SketchFlow) directly into UI design (Expression Blend) and subsequently into development (Visual Studio) - check out the WebsiteSpark Program for almost free licenses.

Don't bend too much!

Danger! Don't bend the curve from Figure 1 too much otherwise you end up with

• a highly functional 'prototype' but without design, i.e. without visual clues whether your client/users will like it (buy it) and without validation that you got your information architecture right OR
• a highly visual 'prototype' that looks sharp, sexy and slick but cannot be used and lack usability ("we just installed the app and now our users complain they [...]" - substitute the appropriate phrase for yourself

Proof-of-concept

Creating medium- to high-fidelity prototypes can be considered proof-of-concept and can be beneficial to or sometimes even required by your project. Looking at Figure 1 that would mean to move their respective dots from Design/Implementation to an earlier phase.

What are your experiences?

Do you use / re-use multiple prototypes within your projects? Do your project structures support prototyping? To which extent?