In this Step, you will:

• Learn what makes up a good usability test discussion guide
• Understand how to write a good usability test task

Before starting your tests, you need to write down the questions and tasks you're going to ask people. This is called creating a discussion guide. Luckily, good discussion guides follow a very predictable pattern.

There is a sample discussion guide included in the material so feel free to jump straight to that, but this section will explain the elements that make up the discussion guide for a usability test.

Boilerplate Intro

Before getting started, you need to introduce yourself and say a few basic things. This is your chance to make the participant comfortable and start to build rapport.

The intro should include:

• A welcome and introduction
• A very quick, high level overview of what you're testing (without giving too much away)
• A reassurance that the participant can't do anything wrong and that they aren't being tested or judged
• A request that the participant think out loud while performing tasks and to otherwise behave as if she were alone
• An approval for recording the session if the session is being recorded
• An opportunity for the user to ask questions

Background Questions

The next step is to write some background questions. Now, you're going to decide which background questions make the most sense to you. This is going to depend on what you're testing for. Don't collect information that isn't necessary. If questions start to get personal - for example, if you're asking about salary - make sure to let the participant know that she doesn't have to answer if she's not comfortable.

Again, you can see sample background questions in the example guide provided, but remember that these are questions that relate to your product. If you're building a B2B product for call center employees, you might want to ask the participant how many calls she typically answers in a day or how often she has to escalate a problem. If you're building a consumer product for meal planning, you might ask how the participant currently shops for groceries or how often she looks for recipes online.

These are questions that will help you understand the participant's performance on the tasks you are about to have her complete.

Tasks

Now you're going to give some people tasks to accomplish on your usability test. These are the core of the usability test, so you're going to need to think hard about what you're actually testing.

Here are some questions to ask yourself before writing your tasks:

• Are you testing the entire product or just specific features?
• Are you testing new features or current ones?
• If you're testing current features, are you testing whether people can use the features at all or are you testing how they use them normally?
• Do the features you're testing require any specific knowledge or data to perform correctly? For example, if you're testing something where the user has to look at financial data, will the user be able to understand looking at sample data, or will she find it easier to understand her own data?
• How complicated are the tasks you're asking the user to perform?
• Will you be running the same tasks on multiple prototypes or products?
• Are you testing whether people can perform this task or whether people will want to perform this task? (Psych! If you're trying to test the latter, stop it. Usability testing only tells you whether someone CAN perform a task.)

Once you've answered these questions, you'll need follow some task guidelines:

• When setting up the task, give the user exactly as much information as she would likely have if she came to the task normally.
• Try not to use leading words when setting up the task. For example, if there is a button the screen called "Create Account" try not to tell the user to "Go ahead and create an account."
• If there is information that the user will need in performing the task, write it down and give it to her ahead of time. For example, if you want her to create an account with a sample name and password, go ahead and give her a piece of paper with that information before she gets started so you don't have to interrupt her when she's performing the task.
• Ask the user to imagine that she is in a specific scenario and then ask her to show you what she'd do, rather than giving her explicit instructions about the task she's supposed to perform.
• Don't forget to occasionally use User Defined Tasks. This is when you ask a participant to think back to the last time she did something specific - searched for something on the internet, filed her taxes, etc. - and then ask her to perform that same task using your product. Obviously, only ask questions that are likely to lead to user defined tasks that can be accomplished with your product.

Here are some examples based on a sample product, Jobs4Pets, that provides a marketplace for finding a job for your pets.

A Bad Task
Create an account on the site.

Another Bad Task
You are going to create an account for this site because you need to create an account in order to post a job for a dog.

A Good Task
I'd like you to imagine that you have a flock of sheep, and you're looking for a border collie to herd those sheep for you. A friend tells you about a site called Jobs4Pets, and you come to this site here. Show me what you'd do to get that herding dog.

Wrap Up
Once the user has performed all of her tasks, you need a quick wrap up. You'll notice that the emphasis here is on thanking the user and telling her that she did great. User research sessions can be surprisingly stressful for people, and taking a moment to acknowledge their help can make a big difference.
A good wrap up will include the following elements:

• Thank the user for spending time with you
• Reassure her that her feedback was very helpful
• Offer her a thank you gift (if you've promised an incentive)
• Ask her if she has friends like her who she would recommend to take part in the research (if you need more participants) - make sure to specify the sort of person you're looking for
• Thank once more and say good bye

Why Does This Matter?

Having a discussion guide helps you make sure that you're prepared when talking to participants and that you've thought through the appropriate questions. It can also help make the data easier to analyze, because everybody gets asked the same thing.
Getting your tasks set up correctly is important because it helps get the user into the mind set she would be in if she came to your product naturally. If you just tell people what to do, you're not giving them room to fail. If you give people a plausible situation in which they might find themselves, you're allowing them to show you what process they would naturally go through.