UX testing (usability testing) is a set of measures aimed at identifying any problem areas on your resource. That is, it assesses whether is it clear enough, logical, convenient, and whether all its technical elements work correctly.

The result of competent UX testing is a list of recommendations — what and how to change to increase the number of conversions and turn site visitors into regular and loyal users.

Why Do UX Testing?

Usability testing reveals major and minor interface issues, each of which impacts your potential customers.

A common example: You carry out SEO-optimization of the site, run ads, catch up with users, and get twice as much traffic as a result, but two times less sales. There may be several reasons for this, and the task of UX testing is to identify each of them and give recommendations on how to fix it.

Also, UX testing shows how clear your interface is to buyers—whether they use it the way you intended,or in a completely different way. This means that they show how to change the user flow on the site to make it convenient for users.

Who Needs to Conduct UX Testing and When?

UX testing is necessary if you want to test the existing interface for the usability of user scenarios, mark all the “problem” areas, and improve them. It is very important to conduct it on the following:

Online stores that want to increase profits, ensure sales growth (at any stage of their work);

projects that are under development to make the resource effective at the initial stage;

sites with high traffic but low conversion rates; mobile applications that have a significantly higher download rate than actual use; portals with a lot of useful functions that users do not use for unknown reasons, et cetera.

If your site does not fulfill the function assigned to it, if you understand that many users simply leave without reaching the goal (such a goal can be not only buying, but also finding the goods they need, answering their questions, booking accommodation, and so on), usability testing will answer the question of why this is happening.

But there are cases when you can do without UX testing by replacing it with in-depth interviews, diary research (studying how user behavior and experience has changed overtime), or split testing. However, if you want to find out exactly what it is in your resource that prevents you from earning more and fix it, it is better to do full testing.

How UX Testing Works

UX testing or audit is a pains taking and time-consuming process, and it consists of several stages.

Stage 1. Formulation of goals, setting goals, selection of methods, and respondents

The goal of this phase is to identify the problem to be solved by testing. Problems and tasks are formed based on the goal. For example, find the reason for the drop in conversion rates after an online store redesign.

The choice of respondents also directly depends on the product that you are testing: there are tests that can be carried out on your employees and there are those where it is extremely important to find a suitable target audience. For example, you cannot check the convenience of a car dealership website on an old lady who does not have a car simply because their answers will not be at all what your target audience will give you.

Stage 2. Formulation of hypotheses

A hypothesis is an assumption about the reasons for user behavior and the influence of usability elements on them. It is important to understand that a good hypothesis is one that can be refuted. For example, a hypothesis might be the following statement: The location of the order form is unsuccessful. The user simply does not see it.

It is very important not to be scattered over several hypotheses at once. One test assesses one hypothesis. After testing, three types of results can be obtained: the hypothes is was confirmed, not confirmed, or partially confirmed. Regarding the order form, the hypothesis can change until the real reason why users ignore it is found:

The form is bulky and inconvenient to fill out;

Users do not see it;

Users simply do not understand what it is for.

Stage 2a. Creating user scripts

We have identified this stage separately, since it is carried out in parallel with the rest of the work, not after completing a certain stage.

A scenario is a model of user behavior on the site, based on their predetermined goals and entry point. It helps to calculate all possible real situations in which users may find themselves.

For each page, you need to develop a separate script answering four key questions without fail:

Where did the user come from?

What should they see on this page?

For what purpose did they come to the page?

How should the visit be finished?

There is another type of scenario, which also includes information about the character of the user, their age, level of education, and awareness of the topic.

The shorter the possible user scenario on your site, the better. It should be kept in mind that the scenario always directly depends on the business and what groups of clients it works with. For example, a scenario of a simple startup selling the same type of scratching posts for cats may look like this: visited, saw, bought. However, if you have a large resource, and your users are not only retail buyers but also small and large wholesalers, each of these groups will have their own scenario and each of them needs to be well worked out based on the four key questions above.

Stage 3. Determination of the type and method of testing

Having determined the hypotheses, one should choose the optimal type and method of testing and test it on the target audience.

Types of UX Testing

Testing is divided into two types: qualitative and quantitative. The first helps to understand more about the problems of the resource using open-ended questions, and the second aims at specific numerical indicators.

Qualitative researchis suitable when you need to understand how users think and get detailed feedback from them about all the inconveniences. In this case, interviews with respondents are conducted based on any of a variety of techniques.

For example, Kano's model provides responses in the form of scores on a scale of expectations from zero to 10, where each score ranges from “I like it, I am pleasantly surprised”to “I don’t like it, and it is not acceptable to me.” The result is a graph showing which features users consider essential, which are unimportant, which exceed their expectations, which are better to remove, and which can be called a competitive advantage of the site.

Quantitative research aims to get a numerical answer to any research. Such an answer can be: the time spent by the user to complete the task, the number of clicks on the way to the goal, even the “yes” and “no” answers translated into a binary system and presented as numbers.

For such testing, theJacob Nielsen method is often used, with the help of the percentage of successful completion of the task by the respondent calculated. For example: 100% if they completed it entirely independently, 50% if they used the help of a moderator, 0% if they did not complete it. This approach helps to identify the presence of problems and the number of users who stumble on the same section of the path. For a reliable result, it is necessary to give the test participants the same tasks.

This is just a small part of the research that is being applied to make your website UX/UI design more effective. If this is already enough for you, and you have decided to check the usability of your site/application/portal, write or call us and we will gladly conduct it for you.