I recently designed a website and I would like to test its' usability against end users.

I am not sure about how I would set tasks, as I have read online not to give any obvious clues.

One task I would like users to undertake, is to vote on a poll. Now this poll is on my homepage - and simply telling the user to vote on the poll is practically countering against the whole purpose of usability testing. This is how I have defined this task (see below), however I do not know if the task is clear and if I have given away too much information.

The task:

You are entitled to a voice and you are able to express your thoughts and views to some of the questions published on this website. How would you go about answering some of these questions?

Firstly, how can I rephrase and improve the above task? Secondly, what are examples of usability issues?

Would appreciate it if someone could modify my task or give me some general advice.

5 Answers 5


Wording for tests depends what you're testing for - If you're trying to find out if they can find where you've put the voting tool and how to use it then it's probably OK to ask them to look at the page and tell you how they would go about voting for 'X'.

If you're trying to find out if you've got the concept of voting right then you're probably going to want to ask them about how they might go about 'expressing an opinion' without referring to the page in your question.

You're right to think about not leading the user too much but you can also go too far the other way to the point where the user may not understand what you're asking them to do. It's good to guerilla test your questions on someone else who is not aware of what you're testing for and find out what they understand from your question before you go and do your proper field work.

  • 1
    If you want to test both the discovery and usability of the poll, you can have a second follow-up task that says something like "if you didn't notice the poll on the page yet, please find it and...". So in the case the user doesn't naturally discover it, you mark that task as failed, and still get the benefit of them actually using the poll for the remaining tasks.
    – Kip
    Dec 1, 2015 at 1:59

First I write down all the tasks I'd like the participant to do.

Then I see if the things I want to test can be turned into a higher level task that touches upon the issues that we want to look at in a way that reflects 'normal' usage patterns. For example we might be interested in looking at whether people:

  • Can register
  • Can log in
  • Can log out
  • Can find the latest news page
  • Can post a comment on a page
  • Can edit their profile.

Might turn into:

  • Can you post a comment on the latest news page? (requires them to register, find the latest news page, and post a comment.

  • Can you log out

  • Can you edit your profile (requires them to login and edit profile).


If you are using the concept of User Stories, and you have built your functionality to fulfil user stories, your usability testing could follow the same pattern. (if you are using Cucumber, then your Cucumber tests can also be derived from the user story really easily)

e.g. Example user stories for Amazon.com

User story: As a customer, I want to be able to browse books for sale, in order to decide what to buy

Task for study participant: find a book you want to buy, and add it to your order

User story: As a customer, I want to be able to manage my basket, in order to cancel my purchase

Task for study participant: cancel the purchase of the book that you added to basket in the first task

Your user stories identify the key functionality for your application, and your usability testing should test the key functionality to make sure it works.

The poll on the homepage

Your question seems a bit too obfuscated. What is the purpose / business value of the poll?

In my Facebook group, I have a poll where members can vote on guidelines for social interaction in the group. This means users can see what guidelines are most important to other members, which reinforces the importance of the guideline (i.e. it was not just arbitrarily dreamed up by a moderator, everyone else agrees too).

So, whatever your users are voting on, you could just say something like "vote on the questions about x, y, z" - that should give them enough of a clue without mentioning the word "poll".


The question I would ask is: why do you need people to answer a poll?

Here are few things I recommend doing before writing your tasks:

1. Define what your goals and objectives for your test. One of the best places to start drawing up what questions to ask is starting with the people who work with your website — your developer, content marketer, manager etc. How much do they know the product ?— features, users, competitors etc. Do they know of any constraints of the product? — development problems. Once you gain that knowledge, you start asking a set of questions to help focus the team on research questions — “Why do people sign-up but not do x,y,z?” ←this is better than →”We need to do focus groups now!”

2. Ask yourself - What questions need to be answered to achieve my goals/objectives for this test. For example, your goal may be: to identify navigation issues. So the question you will need answering is, do people have problems using our navigation? Do people have problems using search? etc.

There's some really great information about defining your goals here.

When you've created your tasks/questions, you're ready to start testing. I recommend checking out UserTest.io - they're my favourite alternative to UserTesting.

Hope this was of use to you.


You have designed your web site in order that a visitor may be able to do a things that will give them information (eg the price of an item) or make some actions (eg book a ticket, express an opinion). The key to describing 'user tasks' is to write the task in the users' language, eg: "Book the earliest flight from Washington to Miami on [day] in coach class" or "When is high tide at the Shannon Bridge on [date]?"

If you haven't, then I'm afraid it's time to go back to square one and ask yourself, what do I intend this web site to do or convey?

But assuming you have, you should be able to set half a dozen or so 'user tasks' in this kind a phraseology: that is, starting from the home screen, and leading up to a transaction or getting some information. You might check with airlines involved before you get people making reservations and then not paying for them :)# and in general when faced with situations like that, the task ended with us when the visitor had the 'pay' dialogue pop up. Be warned though, my friend Jared Spool once showed that visitors will be much more thorough and demanding if it's their own money they are actually spending rather than 'play money' or 'stop before you give your card details' scenario. Funny, that.

There are ways of scoring the results: can they do it (well, almost, badly, not), how long does it take them (mind the internet time-lags!), did they thinki it was easy? You might even ask them to evaluate their experience of the website as a whole using some kind of online questionnaire... hmmm... did I mention my little pet duck, wammi.uxp.ie ?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.