Hypotheses, Prioritising, Methods and Tools – Why They Are So Important In User Testing22 October 2015, Igor Farafonow
Every usability test aims at confirming or denying some hypotheses. The test should not be conducted only for the sake of conducting a test.
Defining proper research hypotheses is the key to the result of any study – so that the money spent on testing is not wasted. The importance of this stage is as high as e.g. defining project objectives within the project development process. If it’s wrong at the beginning we’ll just waste money and time on it 😉
Research hypotheses will help us choose suitable research techniques, scope and quality of the sample. They will also be helpful at a test design stage – they will help us choose questions and tasks that will be provided to test participants.
So let’s start 🙂
Step 1- Hypotheses
Research hypotheses are assumptions concerning certain regularities – sources of issues, bottlenecks etc. How can we identify them? By searching in various places.
Most of all you should take into account:
- Google Analytics data
- client feedback (e.g. from a helpline)
- conversations with customers
- conversations with employees
- your own assumptions (at the very end because you’re the least objective 🙂 )
Your initial hypotheses can be formulated in a following way:
“Complicated and incomprehensible terminology used in filters has impact on using these options.”
“Registration button is not clearly visible which causes skipping this button.”
“Users don’t use additional functionalities at the site – probably because they don’t see them / they seem redundant to them.”
Hypotheses don’t need to include the knowledge on the cause of the issue, they can be formulated in an open form like in the examples presented below:
“Users are lost at the registration screen.”
“Users experience issues at a shopping cart stage.”
Step 2- Prioritising
After writing hypotheses the next stage is prioritising them. At this level we take the following factors into account – severity of an issue (how important it is for the conversion rate) and frequency of occurrences (what percentage of users will encounter this issue). Of course the most important issues are the ones that rank high on both these levels. But if we come across “high severity – low frequency” issue and the other way round “low severity – high frequency” issues there are two ways of approaching them: based on satisfaction (firstly, we focus on issues that occur more often, even if they are minor UX errors) or based on business efficiency (firstly, we take care of issues that influence business, even if they are not frequent).
Step 3 – Choice of methods and tools
In this part we’re moving towards creation of the test itself. At this stage the most important aspect is the correct choice of a method, techniques and tools. Below you’ll find possible issues with suggested solutions:
- if you noticed a high cart abandonment rate and you don’t know what is the reason behind this issue use e.g. remote or lab usability testing – let users complete tasks with you around. Talk to them about aspects that influence their shopping decisions, maybe something scares them off, or something is unclear, or maybe they just have doubts that competitors have a better offer / lower prices. You can also use the so called Customer Satisfaction Survey (pop up window with a questionnaire – targeted at customers who have just abandoned their carts) where you’ll ask two ad hoc questions in a context (don’t you dare asking more questions!) and you’ll get a valuable feedback about the current customer experience,
- if you feel that users have a problem with finding a given product while using navigation at your site (using one path they enter different locations in a navigation pane for a few times and then return) test the information architecture of the site using card sorting. Ask users where they would organise given products or in which categories they would expect finding these elements,
- if you think that users get lost at the site (they come back to the main page many times) and your problem lies in the naming you can use remote testing methods e.g. click and comment. Show users your site and ask them to enumerate the names that they don’t understand. You can also ask them to describe what they would expect by clicking on a given navigation element.
- if you suppose that the problem lies in search filters (e.g. users enter a site for a given product and get back to another product or they don’t use filters at all which is of great importance for e-commerce) it is worth to employ a usability test that will enable your testers to go through the whole path up to the choice of a product. Create a detailed scenario and observe how users follow it. You can use a name of a chosen product, price range, additional functionalities of a product,
- if your customers call a helpline and ask about basic information at the site conduct a usability test (remote or in a lab) that will involve a scenario task that aims at finding these pieces of information. In this case it is worth to recruit your customers to the test since they might have already experienced similar issues.
You can also read our text about Remote User Testing – 10 tips to improve your user research.
We wish you luck!