7 Steps to Create Effective UX Research

Development of a new product is always a creative and fascinating process. A team generates so many ideas, one better than the other, trying to develop a unique service. 

There is nothing worse than a sudden understanding that your project thrills no one: users do not click buttons or use a new tool set or simply refuse to apply your product.

To save time and help you survive in a competitive market, it is recommended to test all of your hypotheses, preferably, by means of attracting a large number of users. Quantitative UX research is here to help.

This article shows how to create a UX research online, how to find a target audience and what mistakes to avoid in an attempt to get meaningful results.

Why You Need to Conduct Online Research

First of all, it is worth-mentioning, such research can be quantitative or qualitative. Qualitative research is applied when you need to know about the struggle and needs of a small group of people. To get new ideas and hypotheses, it is necessary to hold an in-depth interview and to conduct a field study and an interface testing.

Apply quantitative UX research in case you need to make an important decision, prove a hypothesis or results of qualitative research with some data, and check the quality of a final product, attracting a large number of people.

Quantitative and qualitative research always complement one another and help developing a convenient and efficient product for users. UX research can be conducted at various stages of product development. Within the process of development, you can test ideas, avoid mistakes, and calculate risks before a product launch. 

Examples: You are developing an accounting app for entrepreneurs. With your partner/friend you have decided that you are not going to ask users about anything because you are entrepreneurs yourselves and you are aware of all possible issues. Or there could be another scenario: you conduct a short survey among friends and acquaintances, asking them what tools and features they think are missing in similar apps.

You come to a conclusion that you, your friends and a limited number of people tend to use several accounts. Then, the app definitely needs a tool set to control over, at least, ten accounts. So, you let it be a special feature of the app and you build the entire product around it.

Based on the needs of only a few people you start the development process. Software engineers work on it for six months, developing the first version of the product. Then, you find out that the majority of users do not use these features at all. But you have already put your money on it and spent budgets on its development and design.

Quantitative research is also effective at the stage when you already have an MVP or a launched product.  At this stage it is possible to detect errors that have not been detected before, check the quality of technical support and the need for a new tool set.

Going this way (and we assure you, it is The Way!), you need to accept that UX research is not a one-time thing. It is necessary to conduct it on a regular basis to make sure you keep an eye on your product, keep on looking for true relevant ideas and check if you are ready to launch the app.

Online Testing: Pros and Cons

Online research, especially amid the pandemic, is the best way to attract a big audience and get quick results.

Pros of online research:

  • Big selection. It is way easier to find a target audience that complies with necessary requirements. There is no need to limit yourself with a geographical location, and you can always find any kind of participants from any country in the world.
  • Speed. Online research requires way less time and effort than a personal interview.
  • Rather simple analytic system. The majority of the tool set does everything for you, requiring the least of effort. It is easier to make conclusions and decisions based on data, discovered online than processing miscellaneous pieces of information.
  • User-friendliness. It is possible to complete a survey at any time and in any surroundings.

However, an online survey can cause some difficulties, for example:

  • Time and number of questions are limited. Users do not spend more than 5 minutes on completing a survey. Only a few users are willing to spend time to provide detailed comments.
  • It is hard to motivate a person to provide thought-through answers when there is no personal contact.
  • You cannot track a user’s reaction or ask follow-up questions. 

To avoid such format related issues and create efficient UX research, simply follow our step-by-step instruction.

UX Research Drafting

We created a simple guide that contains such sections as goal determination, audience search, preparation of questions and options for answers, limitation of survey time, toolset, and final checks.

1. Set a Goal

A goal might be to check a particular hypothesis or an update that you are willing to add to a product. Or you might want to conduct research to better understand if a new tool or feature is appreciated. 

Remember: you should always have only one goal. You need to be sure that all questions address the same topic or issue. Otherwise, you are going to end up having a bunch of random and basically useless data. 

Example: Let’s go back to the accounting app development. Let’s assume that you have made a wise decision to conduct research to find out if users need an account management feature at the home page of the app. To do so, it is possible to initiate a survey to find out how many accounts do users actually use.

2. Finding and Selecting Respondents

It is rather simple if you are conducting research in an app itself. You have options of announcing a survey in its interface or on a product website or sending a newsletter to all subscribers. But in case you need a new, previously not engaged audience or you need to conduct research on a product that has not yet been released, then you need to make an extra effort.

There are several ways to find research respondents. 

If you are planning to conduct a UX survey with zero expenses, you need to engage social media. Post a link to a survey in subject-related communities on Facebook, Reddit, or online forums. Offer something for those who complete a survey, for instance, a free trial of a new product or a discount for a subscription.

In case you have a specially designated budget, you also have extra options:

  1. Promote such a post on Facebook, applying ad targeting tools. Spending even the smallest budget gives you a chance to get a decent reach.
  2. Use a service of a recruitment agency. It is the simplest but the most expensive strategy: you give an assignment to professionals and wait until you get the results.

The next important question is: do I need to segment the audience? It depends on the goal of the research. In case you are planning to introduce a new tool set, designed for a specific group of users, then segmentation will make the research more efficient and reliable.

Example: The target audience of a finance tracking app consists of entrepreneurs who have started their businesses more than two years ago and have more than 15 employees. It is possible to filter the audience by means of adding a question on the number of business’ employees. Then you can ask the selected audience relevant questions.

3. Ask Right Question

The first and the most important rule of quantitative research is to ask yes-no questions. A user should have an option of choosing ‘yes’ or ‘no’ or checking one of the available options. Apply the ‘Do not make a user think!’ rule. 

Moreover, using yes-no questions allows you to get understandable and clear data as a basis for further conclusions.

Example: When we first started the research on the finance tracking app, we asked the following question: How many accounts do you use? It was a mistake to let users answer the way they wanted to answer. Some of them answered ‘around five’, others stated: ‘7-10’, and a few people answered that they do not use them at all. So, what did we get as a result? A huge amount of miscellaneous data requiring long and complicated analysis.

Obviously, we needed to start all over again. This time we added possible answers about the number of accounts. Offering a yes-no question to users, you get an instant result, expressed in numbers. For example, 30% of users use more than 10 accounts. Simple. Quick. Clear.

There are surveys where open questions are the only possible option. For instance, you are conducting research on a new feature and a user chooses ‘dislike’ as an answer to one of the questions. To find out what exactly such a user does not like you need to ask an open-ended question. It will help you make a more applicable tool set and perhaps might provide you with new ideas and show you areas of further growth.

What else to keep in mind?

  • Avoid double negations.
  • Do not try asking two questions in one. Divide them.
  • Line up questions the way a following question is related to the previous one, so research is structured properly. Compile questions into modules to make a survey easier to complete.

4. Provide Appropriate Answers

  • Add the option ‘Other’, so a user can leave a comment. The majority do not like to do so but there are always people who are willing to share their opinion.
  • Limit the number of possible answers. Do not include more than 7-8 answer options. Everything else can be left using the ‘Other’ option.
  • Apply interval scales in questions, related to evaluation and assessment. Add ‘excellent’ and ‘below average’ to standard options of ‘good/poor’.
  • Be real: add ‘I don’t know’ or ‘Not sure’ to answer options.

5. Limit Time

Make sure you know how much time does research require. Good research lasts for five minutes or less. Make sure users are aware of that time before they start a survey. Users see that you value their time when you state that the research will not take more than five minutes to complete. When they see that they have the required time, the chance that they will start a survey gets higher. But if you do not say how long it takes to complete research, it is likely to be ignored by the audience.

6. Choose a Tool

It is fast and easy to build-up a survey, applying fee-based UX survey development services, like SurveyMonkey, Type Form, or Crowd Signal. Some of them even provide their databases of pre-selected respondents for an additional fee. 

If you are paying for a survey development service, make sure you create a design that will enhance overall user experience. Divide a survey into additional segments where, depending on the answer to the previous question, a user is transferred to the next related question. Make the progress of survey completion visible. This way a user can track how many questions are already covered and how many questions are left. Survey development services can help you brand a survey, using corporate colors and logotypes. It increases brand awareness and credibility.

Google Forms is a standard free option for that. There is no unique design or pre-selected audience but it helps to get the required data. In case you don’t need to conduct monthly UX surveys, this tool is all you need.

7. Survey Revision

A survey is ready but it is not yet time to relax. It is time for a final revision: make sure that research you are going to conduct has no obvious errors or mistakes.

Complete a survey yourself, then ask your team to do the same. If needed, engage your friends and relatives. We can guarantee that someone will spot a mistake you cannot yet see.

Do not forget to thank survey participants at the end of the research. Make sure they feel you appreciate the time they have spent. If you ask them to provide their e-mails, do not spam after. It is more efficient to offer them a bonus: some useful information, a discount, or a promo code. Later you can call the same people up to conduct a beta-testing and qualitative research.

Done! If your survey has a clear goal, a logical structure of questions and answers, a display of required and spent time, then, results will start rolling in very soon.

Good luck developing a product designed to be loved by users!

Other Publications

Contact Us