Over the years I’ve had people contacting me from around the world including the USA, Europe, The Philippines, India and New Zealand.
I thought it might be useful to share some of these questions and my answers with you as well…
♦ How did you end up working in the field of IA and UX ?
I left school at 16 after landed a job at Tearaway Magazine where I did writing (journalism) and graphic design. I guess this gave me a foundation in words and structure.
I then applied to the New Zealand Broadcasting School in Christchurch and got in. From there I got a job at Vidmark in Dunedin where I did television camera work and 3-D graphics. Later on they set-up one of the first web servers in New Zealand and I ended up working with HTML and web graphics.
My later jobs have all been in the web industry and I pretty much just taught myself as I went. It was awkward to fit into any distinct role as I wasn’t a Graphic Designer or Business Analyst – somewhere in between. Actually, I quit a job once when they tried to force me into a role that didn’t fit (or pay what I knew I was worth) because they didn’t really understand what I did even though I knew I was bringing huge value to their clients! (and had billed for the company in the realm of $250k that year too!).
I only realised what I was doing had a name around 2000 when I came across a newsletter (then on paper) by Jarred Spool from UIE.
I got my hands on two books – Information Architecture for the Web by Louis Rosenfeld and About Face: The Essentials of Interaction Design by Alan Cooper. These were my ‘bibles’ for really getting into UX properly and I still highly recommend these books.
And all the books from this site are great too.
♦ What tools of the trade do you consider essential for your work ?
MS Office, Visio and the Web. If you’re using a Mac then MS Office for Mac and OmniGraffle. And a visual diary with a good black pen.
♦ Do you have preferred methods for usability testing ?
I’m “old school” and prefer simple observation. I usually test with 5 users per persona type.
Many people now use computers, eye-tracking etc – but I think that’s overkill. Most usability issues become pretty obvious after just a few tests.
♦ Are there any particular IA / UX practitioners whose work you follow ?
♦ What do you consider are the key competencies an effective Info Architect must have ?
The ability to analyse the complex interplay of people, content and data – and turn this into a simple/effective user experience. I’m a strong advocate of user research (personas) and using these as the basis for all your design decisions. The personas should also include profiles of the business owners – not just end-users.
Increasingly IA’s are coming from a Librarian background and far more qualified than I am at understanding classification methods etc – but I have observed that people with this background do tend to over-complicate things. For example, a librarian I worked with wanted 300 categories for an intranet and I had in mind 12. I had to work hard to convince them to get it down to 30 (a compromise). You see, people just get overwhelmed by information – it’s our job to make it succinct and feel intuitive and create many pathways to the same content.
We also need to understand that there’s different types of people with differing abilities. Many think/learn visually, others prefer words, some look for technical terms, some need ‘layman’s’ terms etc. Anyway, get this book if you don’t have it already!
♦ Do you belong to any specific professional associations / networks ?
Not at present. I have been a member of the UPA and IAI in the past – but they’re very US-centric. There’s a usability meet-up group in Wellington I sometimes attend. Auckland has a very active group who meet monthly.
♦ Can you please explain exactly how method acting it used in your technique? Are participants led through an acting intro/warm-up first? Can you please explain how your scorecard relates to the “acting” component?
It’s not a technique I use much these days but it was effective when I managed a team of UX Designers who already had the will and mindset to give it a go.
So, I didn’t do warm-ups or anything like that and would leave the approach up to each designer. But I would suggest that, to understand their audience, they needed to observe and experience aspects of their audience’s lives.
So I would encourage the designers to:
- Observe the target audience doing their job, shopping, doing stuff – either in a structured way (task observation and an interview), or, if in a public space, just watch people. For example, I observe how people use mobile phones on the bus, or how they use shopping lists at the supermarket.
- Actually try what the audience was experiencing (e.g. registering for a website, ordering a coffee – whatever is relevant to the project).
The key thing is to mimic the audience and their real lives – the example I used in my article was a mother holding a baby, on the phone and using a computer. When designing a user interface I wanted the designers to actually try that and see how it felt. It influences their design decisions.
The scorecard was an adaptation of the tradition ‘heuristic usability review’ – I observed that while a review might tick all the right boxes, the user could still end up feeling frustrated. So I added scenarios/scripts for the testers as part of a review to draw attention to task satisfaction. A better method is ‘observational usability testing’, but there’s not always the budget to do this.
I used the method acting approach for several years myself (in my own head) before sharing it with my team and others. Some other team members did struggle with this approach – I think it takes the right sort of person – someone who can emphasize or can act in role while actually believing they are that person.
I think I find it easy as I’ve had a wide and varied life experience and automatically tuned into the emotions and body language of others (the survival instinct?).
So, suggest you just try and it and invent your own techniques – the trick is to base it on actually observations of your target audience and take it from there.
♦ As a UX practitioner, have you ever been asked to test a concept of a site or application? The expectations from our end is to provide recommendations to make the user experience compelling. However, in my opinion, the concept of the projects are “half-baked” to put it kindly.
Testing at the early stage can be tricky, but yes, it’s all part of the User Experience process and the earlier you get involved the better.
I think all you need to do is take a look at the half-baked designs (assuming they have some visuals?) and use your instincts to see if the design is heading in the right direction. As always, think of the users. Ask who the site/application is targeted at, what they expect users to achieve there and then imagine you are the user.
Unless the site is intended to be wildly creative the site structure should probably follow basic good design rules – e.g. navigation at top, consistency from screen to screen, buttons labelled appropriately etc.
As part of your feedback you might want be proactive and create wireframes to show how the design might be improved – so point out the problems, but also possible solutions.
If they project is still just a loose concept and has no visuals then once again tie to back to the target users, their goals, motivations etc – will the concept meet their needs or desires?
If they’ve fundamentally got the whole user experience wrong from the start then you need to think strategically how you are going to approach the people who had the concept in the first place. Chances are it’s too late to go back and start the process again, so you need to once again, be proactive – point out what’s broken and what the impact is likely to be on the user and the business. Where possibly search the web to try and find evidence or reports to back your conclusions.
Try and suggest solutions, not problems.
♦ I’d like to know the types of user research (interviews, ‘persona hypothesis’, etc) as well as the methods involved in conducting these kinds of research.
I recommend you take a look at this slideshow which outlines my user research process:
Basically I like to keep it simple and use three methods:
1. PERSONA HYPOTHESIS
This is a meeting with your client to explore what the collective group (think they) know about their customers. I start by getting the people in the room to write down on post-it notes every customer-type they can think of. I then cluster the notes into similar groups (e.g. by job title, age-group, behaviour patterns – how you group them depends on the product or service you are working on). I then label each cluster (e.g. one cluster might bee ‘Potential Customers’) and create one persona for each cluster. If you have too many clusters then prioritise – you might only need to document the primary persona.
A really useful toolkit I’ve used and highly recommend is
The type of people to invite to these meetings are those who’ve had ACTUAL contact with customers. Project sponsors usually haven’t met the customers or will assume their own life experience. So for Benetton, for example, I’d request the people who come to the meeting are frontline shop staff, call-centre staff etc.
For this meeting I’d take in a template I want to be filled out (on a laptop or paper), and try and complete as much of this as possible within the meeting. And I try and make sure the meeting has no more than six people – too many people makes it hard to coordinate.
2. USER INTERVIEWS
Using the persona hypothesis to identify the user groups I then know who I need to interview. For each persona I like to interview around 5 people. They need to match the persona as closely as possible (e.g. potential customer, female, age 20-30).
Locating people to interview can be a hassle and time consuming. First you need to find them, arrange the appointment and then confirm the day before to make sure they turn up (of possible you should go to them though). I usually give this task to my client, but failing that you can try a market research company to find them on your behalf. You will need to provide them with a “screening questionnaire” – to make
sure they match the type of person you are looking for.
At the interview I usually spend 30min to 1hr speaking with each person. I ask them about their job, experience, interests etc. See the slideshow for tips on what to look for.
In the early days I had a long questionnaire prepared and went through question by question. But this method is too formal and stilted. Once you get more confident then try and treat it like a conversation – this makes the person being interviewed feel more relaxed and they tend to tell you more interesting things. A great question to begin with is “so, tell me about what you do during a typical day”.
3. FIELD RESEARCH
This means spending a whole day with a user to observe their daily pattern, interactions with other people and gives you a more in-depth view of their life. I rarely get the opportunity to do this due to budget constraints.
What I use to record information: Usually I simply take notes on paper, but sometimes I’ll record interviews using a digital recorder, plus a digital camera for photos. I’ve never shot video of a user, but this is a possibility. The less intrusive you are the better since people tend to feel awkward if they are being recorded or filmed.
♦ How much time do you typically spend on user research?
There’s no hard and fast rule, but to create a fairly in-depth persona hypothesis could take up to a day, but sometimes as little as an hour.
To create ‘quick personas’ (example here: http://www.zefamedia.com/wp-
I’d usually create these on the spot during a workshop with the client and then refine them after the meeting. To create a set of say, 5 personas, would take me about 8hrs. The breakdown as follows:
- Planning: 1hr
- Workshop with client: 1 to 2hrs
- Create draft persona set: 2 to 4hrs
- Feedback and changes: 1 to 2hrs
A more-detailed persona based on actual user research would take about 3 days – roughly the time spent as follows:
- Day 1: Persona hypothesis – I usually start with this anyway as it helps you decide who to interview.
- Day 2: Field research – you get to meet and interview sample users who match the hypothesis. The purpose of this is to check if your hypothesis was correct and add more detail.
- Day 3: Write-up the persona profile, including scenarios, photos, diagrams if needed.
So, if you have decided to profile 5 personas in-depth, you’d need 3 days x 5 personas = 15 days.
♦ Do you have any stories and experiences you can share about discount usability methods?
Almost everything I’ve ever done uses this method!
This is mainly due to the nature of the New Zealand market I think. There’s not much money around for user testing and the deadlines are always ridiculous. Do you have the same where you are?
So I make my prototypes/wireframes in Visio with the page size set to A3. This means when I print the paper prototype out it’s about the same size as a 20-inch screen, so it’s readable.
Sometimes it’s also good to test prototypes on-screen – especially for Web Applications which are more transactional/interactive. So how I’ve done this is to convert my Visio prototype to a PDF. In Acrobat I then
assign hyperlinks to the navigation so the user can click from screen to screen. It’s a bit fiddly, but it works great.
I recommend you get this book http://www.paperprototyping.
♦ With regards to conducting the test, how many subjects do you use on average?
I used to do it with 8 people, but in my experience 5 is about right.
So I always ask for 5 test subjects per persona or audience segment. A few times though I’ve tested with as few as 3 people.
♦ Do you have any tips in conducting the test, post-test data gathering and analysis?
I keep it simple and print out the screens and simply write notes on the screens (with a blank panel for notes on the right) as the user conducts the test.
I’ve also looked into recording tests. Morae is super expensive and overly complex. If you are using a Mac I recommend you take a look at http://www.silverbackapp.com/ – I haven’t used it, but my friend Lulu has and she says it’s great.
For analysis I create a table with the following headings:
- Task Number
- Task Question
- Observations (my notes)
- Example Comments (from the users)
- Number of Users Affected (e.g. 4/5)
- Issue Severity
- Recommended Fix
In the report summary I then group the issues into categories – High, Med and Low Priority – with bullet point summaries of each issue.
If the Recommended Fix is particularly complex then you might want to show the design solution by creating a wireframe diagram to explain it.
A tip on getting buy-in for “usability” – “usability” in my view is a misused term as “usability” is the successful outcome of a good design process. So instead you need to learn to justify the need of “good design”, ROI, “good for business”, “customer loyalty”, “customer satisfaction”, “word of mouth” – in-fact any language your client will understand as adding value to their product.
With any new client (or even a manager within your own company) a good practise to actually profile them – what’s their motivations and goals? Are they design-focussed or technology focussed? Then work on a strategy to get buy-in for what you offer. It’s not easy though – every UX consultant experiences what you do!
♦ Can you share with me your insights on card sorting?
I’m actually not a huge fan of card sorting as I’ve found it assumes that content needs to live within a fixed hierarchy.
You may have seen my blog on the subject.
I prefer providing multiple pathways to content in a way which makes sense to a variety of users (e.g. technical and non-technical terminology).
So how I do it is to just get users to tell me where they’d look for stuff, and then I try and accommodate everyone’s needs.
So for example, an iPod might be found under any one of these pathways:
(by device type) Music Players > iPod
(by brand) Apple > iPod
(by format) MP3 > Music Players > iPod
However, card-sorting does have it’s uses if you are using a website with fixed or static navigation (or space for navigation is limited).
An Open Card Sort is often preferable – this helps you to gather ideas and alternatives – you could then pick the best labels and follow this up with a Closed Card Sort.
♦ In the middle of a study,we were asked to modify the cards. Is it wise to ask the participants of the study who have completed their card sorting to do a new one based on the improved content?
I don’t see any issue with “asking the participants of the study who have completed their card sorting to do a new one based on the improved content”.
For me I rely less on facts and figures and rely more on intuition and observation – you’ll just know as you progress through the card-sorting exercise if the labels are better than before.
As for analysing data – this might be frowned upon by other UX experts – but I don’t put much effort into this. By the time the card sort exercise is over I already have a good idea of what’s working and what isn’t. Any data I use to help inform my overall I.A. – it’s not the sole basis for my decisions.
You might like to try an online card-sorting tool? A popular one here in New Zealand is Optimal Sort.
Got a question of your own?
Send me a comment below!