User Testing #1

Testing after coffee shops is cool. You can test with anyone. Rapid testing requires essentially no preparation. If only that were true...

‍

‍

In two years in House of Řezáč I've done hundreds of tests — and that's why I want to share practical tips on what to avoid and what to look out for. We will talk about qualitative user testing according to Krug. Throughout the series of articles, we will analyze in detail the pitfalls of screening, writing a scenario or creating a report.

However, in this first article in the series, we look at why to test, what testing consists of, and what you should do continuously if you want to incorporate testing into your processes.

‍

‍

Testing reveals

At least x times (reach any number for x according to the frequency with which you read articles about UX, design, research, etc.) you have heard that You will not release an untested website into the worldThat it is not only those who are afraid to test, and that the sooner the better. Yes, it is.

But why? For me, the most important thing is the following: You get an *immediate* reaction from users, which is at the same time Argument for constructive discussion about possible errors in the design of the site and their solution. All this leads to saving resources (money, people, time) and increasing profits. Amen.

*Caution! Don't think of a reaction in the sense that users tell you what they like and what they don't. Not by any chance. About that later.

People often use the site differently than intended, and the following statement applies: “Until you see, you will not believe.” Plus, it's a bummer. Sometimes it is a challenge for me to mask the growing joy of seeing that the fourth respondent has already encountered the same problem and is lost. This results in a pattern that, through testing, we discover and then destroy its source.

This sounds very dramatic, but the list of trouble spots you expose through testing is almost endless as the number of projects increases. As well as A variety of strategies that users will developto find what they need. You're not going to invent that. On the other hand — the more you test and force designers testing reflect, the less they suggest the same error spots.

‍

Testing process

Recruiting, screener, script, recording, rewards,... not a few. So let's break down the parts and briefly describe them to see what you need to think about.

‍

1. Recruitment and selection of respondents

You can test with anyone. When you test a fifth website with your mom that she would never look at, you'll understand that you can't. When you see a student of art studies choosing an ERP system for the whole company or shopping at the e-shop of a respondent sitting at a computer for the third time in their life, you will understand that... every test must be preceded by screening of respondents: that is, the selection of those relevant to the project.

‍

2. screenplay

The scenario is quite crucial, because through its lens you then evaluate how the site is doing. If you come up with irrelevant tasks or poorly worded tasks (keywords, hypothetical questions, etc.), the result will be distorted. Even getting reactions from the right people doesn't save you.

‍

3. Site, process & technology

It is necessary to remember to thoroughly prepare for the test, that is, to choose a suitable place, to try the technique and, most importantly, to create the right atmosphere in which the respondent feels comfortable and not like in the exam. You need to prepare an introduction that dispels all of the respondent's concerns and handle questions like, “Should I click on this now?”

‍

4. Debrief and reporting

After each test, it is advisable to immediately record field notes. Although you have a record taken, immediate notes after testing will help you get the most out of your final report.

Individual areas I will discuss in detail in the following articles, but in order not to dwell only on a narrow list, how everything is important and that you will learn everything later, let's finally take a look at recruiting. Respondent recruitment differs from other activities in that It is worth doing it on an ongoing basis. You won't do a script or screening ahead of time if you don't know what you're going to test, but you'll need respondents all the time.

‍

Recruit respondents continuously

Having a tremendous segmented database of available and willing user testing respondents is every researcher's dream. In a few seconds, filter out the target I need right now, set aside a day and test them all pretty well.

In an ideal world of bouncing unicorns from Asana, this would work. How to approach it at least? Recruiting respondents requires systematic work not only on the social capital of yourself and your colleagues (i.e. convince everyone to give you the emails and phone numbers of everyone they know and must meet at least 10 new people a month:), but most importantly active long-term contact gathering.

If we're doing pop-up research at the office and talking to people, we'll ask them to get in touch. If we distribute questionnaires to applicants at Gaudeam, we ask them to contact them. We then, of course, dutifully write them down in a shared document. It's like with saving, the sooner you start, the less it hurts.

‍

‍

In order to get as rich a database as possible in House of Řezč, we launched of course also through the online form Call for interested parties. The risk, however, is that some of these test respondents are very eager to experience testing, God forbid they want to get a job through it, and therefore often disproportionately try to accommodate the researcher. And the researcher certainly doesn't want that. It seeks to create a situation “i sit in the living room with my laptop on my lap and act natural”, not “I am at a job interview, and therefore I will preemptively praise everything and keep the unpleasant to myself”.

For example, you can search for specific respondents in narrowly focused Facebook groups or actively (if possible) go to the given environment and recruit directly there.

It is also a good idea to play on employees of Mary Kay or Herbalife and the respondent who has just been tested ask for contacts to someone else. He has pleasant feelings from the test and is happy to recommend other acquaintances, often new respondents and prepares for you to contact them.

So get contacts where you can and, if possible, also record the basic characteristics of the respondents. Name, surname, email and/or phone number are obvious. Ideally then add the age, what they do (student/unemployed/employed/entrepreneur), what they do it in (IT, education, gastronomy, HR, culture, etc). Also, always keep a box for additional notes, it will come in handy. The most important thing is contact, but even basic categorization will make screening much easier.

Screening? Yes. Seriously about the choice of respondents next time.