If you’ve already read part 1 of this blog series on User Testing, we’ve learnt what the benefits of usability testing are and some of the tools available to help. Let’s get into understanding the key stages of conducting the research effectively. Here are the 7 stages of UX research.
Step 1 – Why are you testing?
The first step in any research is to identify why you are doing the test in the first place. Are you trying to understand why conversion is low, or bounce rate is high? Are you trying to determine what products your users would prefer to see or what sort of language/ tone they prefer?
These are all key questions that need to be considered to scope out the objective of the research. Remember user journeys are complex and every user will have their own personal preferences.
Step 2 – Preparing your script
Having a document to hand to guide through the key questions you want to answer benefits two main areas:
1) To help you, as the moderator, stay on track; and
2) Ensure the questioning across the users is consistent and unbiased.
Your script should start by explaining the objectives of the research to the user and some initial probing questions.
Don’t forget to ensure that the tester speaks about their thought process, no one likes an empty silence and we’re trying to understand what the user is thinking and feeling.
Here’s some example script ideas to get you started:
- Background/Context— “How do you usually get around the city? Do you use any apps on your phone?”
- Expectations — “Before clicking on the ‘search’ button, what do you expect to happen?”
- Usability — “How do you save this location for future reference?”
- Rating Scale — “How difficult (1) or easy (5) was it to search for San Francisco on this map?”
The trick is to ask questions that extract useful answers without leading the user or influencing their behaviour. For example, “How do you scroll to see more results?” suggests that I can scroll.
A better question might be “Are there more results on this page? If so, how do you see these results?”
Step 3 – Create a prototype to test
Start by mapping out the user journey and products you want to feature on your website and create this into a prototype solution or MVP (Minimum Viable Product) so you can test without investing heavily in design or development.
This will help inform the wireframes and the key components required in each of the pages across the user journey. Remember, we’re not talking about design or visuals at this stage. We’re looking to nail the user flow and get the prospect to the point of conversion as efficiently as possible.
You can use tools such as Invision to create an interactive prototype and share with users for the testing session. Ensure that you follow through and create all the areas within the prototype which you want to test – it doesn’t have to look pretty at this stage as you’re testing the user flow.
Step 4— Recruit the right users
Understanding your users is one of the most critical stages of conducting qualitative user research. You could always recruit users through social media like Facebook and LinkedIn. Just consider where your prospects are likely to hang out online and go engage with them.
Tip: Don’t invite your friends to test, they’ll only give you the version of truth they think you want to hear.
Recruiting can also be the hardest part, even just finding 6 willing participants can take days or even weeks to find the right candidates that match your profile of users. You may also need to incentivise users to participate, remember each session will likely take around 1 hour of their time so they’ll probably want something in return.
If you already have a database…
Consider inviting them to a sneak peak and use it to create some buzz around your product or launch. These users will already be engaged with your brand; however, their opinions may also be skewed. We’d advocate finding prospects and those who are ‘unaware’ of your brand. This is more pertinent for startups.
Step 5— Where to conduct the testing
Whilst it’s preferable to conduct in-person user testing in a quiet space there is also value in conducting it in a space where the participant is more comfortable.
This could be a café so long as there isn’t too much background noise or external distractions and they have a good internet connection – there’s nothing worse than the WiFi going down half way through a session.
Remote testing is also easier to recruit for as the user doesn’t have to travel to a specific place of your choosing. Always have a backup plan and always conduct a trial run!
You’ll need to ensure that the participant has a laptop or desktop available with a camera and microphone – often a laptop has these built in and so are the preferred tools for the job.
Step 6— Conduct the session
Tools such as Google Hangouts or paid solutions such as Lookback are useful to share the testing link and prep the user in advance of the allotted time. It’s best to dial in at least 15 minutes early to ensure there are no technical glitches.
How to conduct a user testing session
Before the session
Send an email reminder the day before to the user (if possible) and if you’re calling them do so a few minutes before the testing time to ensure they’re ready – the earlier you can get on the call the more time you have available to test and gather insight.
During the session
There should ideally be at least two people conducting the test. The moderator and the observer (or notetaker). The moderator’s job is to ask the questions and following script whilst probing the user to convey their thoughts and feelings.
The observer’s job is to take notes of key insights that the moderator may have missed. It’s impossible to moderate and take decent notes at the same time, in particular contradicting behaviours such as getting stuck on buttons.
In addition, it’s important to ensure that the screen is recorded and, if possible, a separate camera is setup to record the overall session – this can capture facial expressions and body language which often tells a different story to what is being said.
A camera can be hooked up to a separate room for the client to observe with discretion to the participant.
Step 7— Analyse the results
The final stage, if not the most important is to analyse and digest the results. There are many ways to do this and this does depend on the scope of the initial brief and script which was used to conduct the research.
Typically, user testing can be used to score key functionality such as user journey, credibility factors, branding, ease of use, ability to purchase and trustworthiness. Another method is to create four columns to input the findings, these are:
- User likes
- User Dislikes
For each category there should be common themes, trends and form ideas as to how to overcome those barriers. This is where most of the time should be spent, which is why we recommend recording the sessions.
Test and learn
User testing isn’t a one-time process, continuous testing is required through the product lifecycle and post launch so that there is continuous improvement in conversion. Continue to test your ideas, revise, optimise and learn what works well for your users. The more feedback to gather the more refined your product will be.
If you want to find out more about web user testing or would like help with conducting your own research we’d love to hear from you, get in touch.