The Challenge

A crucial challenge within user centered design (UCD*) is to involve users to, for instance, obtain information about their needs and to evaluate concepts. UCD methods are intended to help designers in this challenge. However, product development practice is extremely hectic and messy, often leaving practitioners with very little time to explore beyond the methods they already know by heart. As a consequence designers often stick to what they know, leaving many potentially beneficial methods unused and those hindering the development of the field and eventually the development of better products for users. Practitioners need to be able to quickly find the appropriate method, assess its qualities and learn how to apply it. But with over 200 methods and techniques available, how do you choose what’s best?

Even while many methods are spread over a wide variety of sources, such as on-line collections, I didn’t felt they were empowering designers and researchers in finding methods that would be a better fit with the challenge they are facing. Talking to designers and researchers in my network, there seemed to be a need for a better solution. An online resource would be the interactive and easy-to-update solution I was looking for. In 2011, I suggested the idea to members of the Design for Usability research project (DfU); a collaboration of the Dutch Delft University of Technology, University of Twente and University of Technology Eindhoven together with the companies PhilipsOcéUnileverT-Xchange and Indes. It didn’t take long for this project to be part of a bigger vision to support industry in the creation of usable products.

The process of creating UCDtoolbox

The process of selecting the right method for user-centered design seemed quite complex because of the amount of methods to choose from, many criteria that can be used for selection, and – added to that – the often extreme constraints in terms of resources (time, staff, budget) that product development practitioners face. After a couple of interviews on the challenge and product development processes in general, we took the practitioner’s knowledge about the project as a starting point for method selection:

  1. The type of product that is being worked on, for instance a physical product, an interface or an environment;
  2. The goal for applying the method such as analysing users and their context, synthesizing solutions, simulating a design or solution and evaluating a design;
  3. Limiting factors and available resources such as available time span and staff; and
  4. Optional criteria such as desired study location and participant details

We tagged some methods based on these criteria and see if there was a common ground between practitioners: for what types of products and for which goals could one use Cultural Probes? First indications after a test with the DfU partners were that we could be on to something here so we went a step further.

A more mature version of the selection procedure was embedded in an interface with 14 methods for testing purposes. The selection procedure and interface was evaluated during the Design for Usability symposium of 2011 and with an international group of academic researchers of the TwinTide group.

The main screen of the UCD Method Exploration Tool

A first working prototype of the interface with the selection procedure was built with Microsoft PivotViewer** and tested with 20 participants at the Chi-Sparks 2011 conference. The selection procedure was again refined and two slightly alternative interface designs that now also included method information based on previous user feedback. Each detailed method page contains step-by-step instructions to support execution. And because its common practice to slightly change the method to better match specific goals or resources we also added tweaks to the descriptions.

Screenshot of the prototype in PivotViewer

This package was once more evaluated – this time with practitioners, researchers and students. Based on their feedback, a final iteration was made in the development of both the selection procedure, the interface design and presentation of the content. This was presented during the Design for Usability symposium of 2012 and shown at the The Web and Beyond conference 2012.

Master Thesis and Appendices

Part of the process and results of this project are described in a Master thesis, accompanied by its appendices. In addition to that, the book “Design for Usability – Methods & Tools: A practitioners guide” contains a chapter about the project.

* UCD; this umbrella term for User Centered Design and includes terms like User Experience (UX), Customer Experience (CX), Interaction Design and Experience Design (XD).

** PivotViewer was found as an interesting candidate to develop the method selection tool in; it matched with about 90% of the initial requirements. However, we also discovered a number of issues in both the data handling and the interface so we decided not to use it for the development version.

The Private Beta

The final design was technically developed from April 2012 and was launched as a private beta in January 2013. Unfortunately, funding ended not long after the lauch of the private beta and the UCDtoolbox was not improved from that moment on.

The current public beta

By late 2016, Tristan started redesigning the toolbox and laying the groundwork that would make further development of the platform easier. The according public beta was released in January 2017 and key-improvements included:

  • Full responsiveness – works any where, any device
  • Selected filters appear in the url for easy sharing a method selection
  • Full-page method descriptions that allow for easier improvement
  • Embedded the earlier stand-alone separate glossary for all methods
  • Easier navigation and removed the formerly required registration
  • Many more minor changes

In the design process, it was not possible to (yet) transfer the following features of the private beta:

  • Increasing size of the method card during filtering due to responsive behaviour
  • Make cards invisible during filtering