Chapter 8: How To Avoid Arguments on Usability:WebWritingAndDesign

Chapter 8: How To Avoid Arguments on Usability

First Way to Start an Argument: Personal Passion

"Everybody likes ... "

This statement is usually bias. Just because you like a certain thing, doesn't mean that you should think everyone else will. Try to avoid this.

Second Way to Start an Argument: Art vs. Commerce

It's hard for the professionals working on the site to agree, because each has a professional passion. By this CEOs, developers, designers, and those in business development all want the site to look different.

There is no Average User

All web users are unique
All web use is depends on the individual

So, don't ask "Do most people like..." instead ask yourself "Does this ... create a good experience for the people that are going to use the site?"

How to End the Argument

Test out the site. Build a version of the page and then watch people try and figure out what the site is and how they can use it. The usability testing moves from trying to figure out what people like to what does and does not work on the site for people.

Chapter 9: Keeping Usability Testing Simple

Don't wait until the last minute to test your site.

Focus Groups are not usability tests

Focus Groups allow people talk about things, while usability testing allows us to watch people use things.

doesn't let us see whether our site works or how to improve it like usability tests do.
focus groups are good to see if the idea behind our site makes sense and if our value proposition is attractive; get's outsiders wants, needs, and likes. (makes more sense to do focus groups in the planning stage before you make a site)

True Things About Usability Testing

If you want your site to be great, you need to do testing.
Testing even just one person is better than not testing anyone at all.
Doing testing in the early stages of the project is better than doing a lot at the end.

it's easier to fix things in the beginning than at the end.

DIY Usability Testing

p.117 graph

How Often Should You Test?

Krug recommends once a month in the morning.

It's easy and doesn't take much time.
Gives us what we need: issues that we can fix.
You'll never have to always decide when to test.
More people are likely to attend if it's always a predictable schedule.

How Many Users Do You Need?

Krug believes 3 people is all that is needed.

DIY tests are qualitative not quantitative.
You don't need to find all of the problems.

We don't have the resources to fix all the problems, so just fix the major ones.

How Do You Choose The Participants?

Recruit loosely and grade on a curve.

It take more money and time to recruit the people who fit into the frame you want to target, so get a mix of your target audience and those who aren't.

Why it's a good thing to use some participants that aren't from our target audience:

You'll make a site that a lot of people can understand, not just your target audience.
We're all beginners. This will help us make the site easier to use.
Everybody appreciates clarity.

How Do You Find Participants?

You can find participants in user groups, at trade shows, Craigslist, social media, customer forums, etc.

Where do you test?

A quiet area with a computer with Internet access, a mouse, keyboard, and microphone. You'll be using screen sharing software and screen recording software to capture everything about the test.

Who should do the testing?

Anyone can be a facilitator, but they should be a people-person. ex. nice, good listener.

Who Should Observe?

As many people as possible. The more people, the more problems and issues they will identify with the participants taking the test.

What do you test? When do you test it?

Test early and throughout the entire development process.
Test competitive sites or sites that have the same characteristics as yours.
Redesigning a site? Test it before you change it so you'll see the problems.'

How Do You Choose The Tasks To Test?

Think of what you're testing and then come up with a list of tasks they'll need to complete.
ex. login process= create an account, log in with an existing username and password, retrieve a forgotten password/username, etc.

What Happens During the Test?

According to a typical 1-hr test:

Welcome: How the test is going to work.
Ask the participants questions.
Have participants tell us what the make of the home page.
Participants try to perform the list of tasks.
Ask participants questions about their tasks or any questions those observing want to ask.
Thank them, pay them, show them the door.

Typical Problems

Users don't understand the site's concept.
The words that users are looking for aren't there. (the creators didn't anticipate correctly)
The site has too much going on.

Debriefing: Deciding What To Fix

Allows team to share their observations and choose which problems they are going to fix and how they are going to do that.
Fix the most serious problems first.

How to make this happen:

Make a collective list of observed problems
Choose the 10 most serious problems
Rate them 1 to 10
Create an ordered list

Keep a list of all of the other problems that aren't as serious (but easy fixes)
Resist the urge to add things
Don't always listen to "new feature requests"
Ignore "kayak" problems (when users don't get something right the first time but figure it out quickly)

Alternative Lifestyles

Remote testing
Unmoderated remote testing

WebWritingAndDesign : NotesOnKrugChaps8to9KMH