UX Scotland 2018 ran from 13–15 June at Dynamic Earth, Edinburgh. I attended one of the three days to explore and learn from what other UX professionals have been doing. I wanted to summarise the juiciest talks from the day that are worth sharing, which cover the following topics:
- Designing connected products
- Asking better questions
- The lost art of task modelling
Designing connected products
The keynote speaker that day was the well-known UX and product design consultant Claire Rowland (lead author of the insightful book Designing Connected Products). It was an inspiring talk that focused on designing systems of connected products beyond the realm of UI.
Claire consolidated seven important questions to ask while designing a system for connected products:
1) Is it a product or a service?
a. Many products are a mix
2) How does your product work?
a. Consider the times when the product goes offline for whatever reasons
3) How do devices connect?
a. Explore what happens in the spaces when they’re not connected
4) Which code runs where?
a. When parts of the system lose connectivity or power, what stops working?
b. What are the risks of underestimating support/service costs?
5) How do we design not just for individual UIs but for distributed UX?
a. How should the user understand which bit does what?
b. What’s appropriate given different form factors and conflicting conventions?
c. However different the UI is from its connected physical product, the identical functions must at least have the same name
6) How often do devices connect?
a. Don’t think of devices and UI, design for the spaces in-between
b. Any product that users rely on, should maintain basic functions in the temporary absence of connectivity
7) How do we give users visibility of system functions and privacy?
a. Algorithms reflect our biases and users are often unaware of why they don’t see intended results by interacting with a system
b. Bring visibility of what is collected and shared
Here are some examples of connected products and services that have caused annoyance to users:
Smart food vending machine
A smart food vending machine in Japan dispenses food and drinks to the public, based on the gender and age of the user in front of the machine. This means women in their 20s get a slightly sweeter beverage. What’s the value in the smart machine over-riding human choice, independence and judgement in this scenario?
Self-propelled intelligent buggy
Smartbe, a very expensive self-propelled intelligent buggy that lets a parent jog behind it and synchronises its motion with the parent’s motion is being built with an app to control various features.
Some things it doesn’t answer are how it would perform when running downhill, navigating uneven surfaces, potholes, edges, obstacles on pavements such as lamp posts or other pedestrians? Some connected products have the potential to fail in worse ways than the convenience they offer or the products they replace.
Fitness tracking app
Strava accidentally pinpointed the location and layout of secret military bases where fitness-conscious soldiers had been running around with their fitness trackers or phones. Whoops!
The heatmap shows a military base in Helmand Province, Afghanistan with the route taken by joggers highlighted by Stava. The incident involved information that users themselves made public, but such data is also vulnerable to hacking:
Smart spoon for hand tremors
Gyenno, a £210 smart spoon offsets hand tremors (by 85%) for patients with Parkinson’s disease, Multiple Sclerosis and stroke. But, this has design problems such as no battery level indicator, the spoon being too small, handle being too large and the connected fork being irreplaceable if it breaks.
I did some research and found a rather compelling product in comparison (not digital but smart indeed). Here’s Ornamin, a £11 spoon guard on the market that does the same job efficiently. As the spoon guard material is flexible, it simply fits all spoons and does its job:
Claire gave the following tips for making an Internet of Things (IoT) product:
- Put together a diverse team. It takes a diverse team to make an IoT product or service, more diverse than for pure software.
- No one fully understands everyone else’s jobs, but try to understand their biggest concerns (e.g. saving power, trying to ensure the hardware can support future features that might not be defined yet).
- Foster communication, so everyone shares what they’re working on and their issues.
- Involve everyone in the team throughout the process. What looks like a design decision could impact API design or even firmware.
Asking better questions
UX designer Chris How definitely made this topic interesting and entertaining enough to not forget. A significant part of a UX designer’s life involves asking questions… more importantly, the right questions. The sort of questions that would give unbiased answers to unlock design ideas.
Chris How’s guide to asking better questions has 3 main points:
- What makes a good question?
- How do I get better answers from my questions?
- How do I get better at asking questions?
What makes a good question?
There are eight factors to consider in what makes a good question.
1) End in a question mark
2) Has a purpose
a. Start with a problem statement
c. List of what I’m trying to understand
3) Insightful and actionable
4) Opens up a conversation. Don’t use:
a. Closed questions, e.g. yes or no questions like “do you want a cup of tea?”
b. Closed-ended questions, e.g. “do you prefer tea or coffee?”
5) Should be neutral and free of bias
6) Should be interesting
7) Is short
a. Avoid ‘and’, commas and too many words in a question
8) Can be answered
a. Avoid jargon
How do I get better answers from my questions?
Chris gave five top tips on how to get better answers from your questions:
1) Listen and observe more than you talk
a. Use the power of silence
b. The more you talk, the less insight you’re getting
c. Talk for less than 1/5th of the session
d. Get comfortable with silence
e. Let interviewees fill in the gaps
2) Be aware of moderation bias
a. Avoid prompting, leading and confirming
b. Don’t praise answers or nod in agreement
3) Ask fewer questions but go deeper
a. Use the 5 Whys
4) Know your killer question and when to ask it
a. Frame it as ‘If I could only ask one question, what would it be?’
5) Empower the interviewee
a. Unlock insight through interest and empathy
b. Put the interviewee in the driving seat
c. Do a dry run of the sample questionnaire
Things to avoid when asking questions
Chris also provided a handy bingo of what to avoid when asking questions!
The lost art of task modelling
UX expert Jesmond Allen gave an interesting inside peek into task modelling she’s done for some big clients such as the NHS, amongst others. There is a fair bit of psychology applied in task modelling.
What is task modelling?
- Task modelling is a practical design tool
- Uses a diagram to break down a task into small sub-tasks
- Incomplete task creates stress in users
- Too much focus causes mistakes
How to do task models
- Understand your users through primary research with end users
- Understand user’s tasks and the order they do it
- Identify common steps
- Represent findings through diagrams
Applying the Zeigarnik Effect and mental sets in task modelling
Our goal is to create designs that enable users to complete their tasks. The Zeigarnik Effect is a psychological phenomenon describing a tendency to remember interrupted or incomplete tasks or events more easily than tasks that have been completed. Incomplete tasks trigger stress.
Hadassah Lipszyc, a CBT psychologist from The Blue Tree Clinic shares how the Zeigarnik Effect plays an important role in mental health:
“Incomplete tasks and procrastinating often lead to frequent and unhelpful thought patterns. These thoughts can impact on sleep, trigger anxiety symptoms and further impact on a person’s mental and emotional resources.
“When used personally or in a therapeutic setting such as cognitive behaviour therapy, the Zeigarnik Effect can promote mental wellbeing by motivating someone to complete tasks, develop healthier habits, set goals and resolve issues that are being postponed.
“Completing tasks successfully can provide a sense of accomplishment whilst increasing one’s self-esteem and confidence. Additionally, a person who can find closure for stressful events or tasks will likely experience a long-term positive impact on their psychological well-being.”
What are mental sets?
What have we learned?
- Humans fixate on their task
- We complete tasks in a specific order
- Until we consider it finished
- Then we move on to something else
How task models help
Task models help find problems and thereby create better solutions. Have you ever forgotten to collect your bank card from a cash machine after withdrawing some cash (perhaps in a foreign country?)
This is explaining it step-by-step:
- You went to the cash machine
- Inserted your bank card
- Requested how much cash you wanted
- The machine spat out cash for you
- You got busy collecting your cash
- The machine spat out your bank card while you were busy collecting cash
- You took the cash and left (forgot to collect your bank card)
Here’s an improved process where the machine helps you collect your bank card first (no problem if you’re living in the UK):
- You go to the cash machine
- You insert your bank card
- Request how much cash you want
- The machine first spits out your bank card and won’t spit out cash until you do this
- You collect your bank card
- The machine spits out cash for you
- You collect your cash and leave (this time, with your bank card)
The 3 laws to remember for Task Modelling
The realm of UX design is ever-changing and expanding. Our experienced team is constantly adapting to the industry, as digital takes more of a precedence in our everyday lives. If you would like to discuss how UX can help your business, get in touch with the team today.