Lessons from Homer’s Car in Human-Centered Design

Sam Medrington

This talk will cover the different stages of the Human-Centered Design process, with examples of what can go wrong and examples of how to do it well. Introduction: [start with a clip from The Simpsons] https://www.youtube.com/watch?v=Pw9gaEiQAxY (Episode where Homer meets his long-lost brother Herb, voiced by Danny DeVito. Herb owns a car company and is attempting to design a new car to revive the company’s fortunes. He decides that as an “average American”, Homer is perfectly placed to design the car that the people will want to buy. The results are, predictably, a disaster.) What is Human-Centered Design? “Human-Centered Design is the process in which the needs, wants, and limitations of the end user are given priority.” “It’s about optimizing the user experience, based on how people are able, and want to work, instead of forcing them to adapt and change themselves to work better with the system as designed.” Human-Centered Design gives you: * Lower costs for customer acquisition * Lower support costs * Increased retention * Increased market share, all of which makes for a healthier bottom line * Hidden ROI comes from the result of an overall effort, many benefits of a UX effort are often unnoticed. However: 80% of companies surveyed by Bain & Company believed that the experience they offered was “superior”. However, only 8% of their customers described their experience as superior If Human-Centered Design is so great, why do so many IT projects fail? Overview of Human-Centered Design as specified in the International Standard ISO 9241-210:2010 (six stages). 1. Plan the human-centered design process How this can go wrong: * Not setting clear and stable research objectives (example from FinTech project) * Opaque research methodologies (example from Telco project) * No time for research findings to influence programme delivery (example from Telco project) How to do it well: * Set clear objectives for all user research activities * Be open and transparent in all design activities * Communicate across workstreams 2. Understand and specify the context of use How this can go wrong: * Not being clear on who the users of the product or service will be (examples from Government and Telco) * Not understanding the project team structure (example from Fintech company) * Reports that lack insight (examples from BBC and Telcos) * Cookie-cutter UX Artefacts that don’t offer value (examples from Telco, Fintech) How to do it well * Ensure that target audience is defined and agreed * Be clear on who within the project/organisation needs to use the specification, and get their buy-in * Communicate the context of use in a format that is insightful and provides value 3. Specify the user requirements How this can go wrong: * Incorrect assumptions about user needs (example from BBC) * Focusing on what users say, rather than observing what they do (example from Fintech company) * Alienating users by prioritising the needs of their managers (example from Industrial Manufacturer) How to do it well: * Challenge assumptions around user needs * Take a structured approach to interpreting user feedback * Involve users in the design progress, to increase buy-in 4. Produce design solutions to meet the user requirements How this can go wrong: * Not considering the user in all aspects of design (examples from BBC and Telcos) * Basing the design on legacy systems (example from Shipping company) * Designing navigation systems that are not extensible (example from Finance company) How to do it well * All aspects of design (including Visual design, copy and tone of voice) must be based on an understanding of users and context of use * Follow Interaction Design best practice * “Sprint Zero” for exploratory research and Information Architecture 5. Evaluate the designs against requirements How this can go wrong: * Research can be biased due to conflicts of interest (examples from Telco, User Research Agencies) * Bad research execution/moderation (examples from Telco, Finance company) * Choosing the wrong research methodology (examples from User Research Agencies) How to do it well * Avoid conflicts of interest, by ensuring that incentives are aligned with project goals * Consider strengths and weaknesses of different research methodologies * Be aware of cognitive biases that can impact user research 6. Iterate where appropriate How this can go wrong: * Testing the wrong thing (examples from Telco, Finance company) * Testing in the wrong context (example from Shipping company) * Research activities taking too long (example from Fintech project) How to do it well * Take steps to ensure that research is valid * Conduct in-context research, where appropriate * Ensure that research activities have a cadence that aligns with other Programme activities.

Drupal is a registered trademark of Dries Buytaert.