Usability Testing: what, how and why?

We all know that building a robust information architecture (IA) can make or break your product. And getting it right can rely on robust user research. Especially when it comes to creating human-centered, intuitive products that deliver outstanding user experiences.

‍

But what are the best methods to test your information architecture? To make sure that your focus is on building an information architecture that is truly based on what your users want, and need.

‍

What is user research? 🗣️🧑🏻💻

‍

With all the will in the world, your product (or website or mobile app) may work perfectly and be as intuitive as possible. But, if it is only built on information from your internal organizational perspective, it may not measure up in the eyes of your user. Often, organizations make major design decisions without fully considering their users. User research (UX) backs up decisions with data, helping to make sure that design decisions are strategic decisions. 

‍

Testing your information architecture can also help establish the structure for a better product from the ground up. And ultimately, the performance of your product. User experience research focuses your design on understanding your user expectations, behaviors, needs, and motivations. It is an essential part of creating, building, and maintaining great products. 

‍

Taking the time to understand your users through research can be incredibly rewarding with the insights and data-backed information that can alter your product for the better. But what are the key user research methods for your information architecture? Let’s take a look.

‍

Research methods for information architecture ⚒️

‍

There is more than one way to test your IA. And testing with one method is good, but with more than one is even better. And, of course, the more often you test, especially when there are major additions or changes, you can tweak and update your IA to improve and delight your user’s experience.

‍

Card Sorting 🃏

‍

Card sorting is a user research method that allows you to discover how users understand and categorize information. It’s particularly useful when you are starting the planning process of your information architecture or at any stage you notice issues or are making changes. Putting the power into your users’ hands and asking how they would intuitively sort the information. In a card sort, participants sort cards containing different items into labeled groups. You can use the results of a card sort to figure out how to group and label the information in a way that makes the most sense to your audience. 

‍

There are a number of techniques and methods that can be applied to a card sort. Take a look here if you’d like to know more.

‍

Card sorting has many applications. It’s as useful for figuring out how content should be grouped on a website or in an app as it is for figuring out how to arrange the items in a retail store.You can also run a card sort in person, using physical cards, or remotely with online tools such as OptimalSort.

‍

Tree Testing 🌲

‍

Taking a look at your information architecture from the other side can also be valuable. Tree testing is a usability method for evaluating the findability of topics on a product. Testing is done on a simplified text version of your site structure without the influence of navigation aids and visual design.

‍

Tree testing tells you how easily people can find information on your product and exactly where people get lost. Your users rely on your information architecture – how you label and organize your content – to get things done.

‍

Tree testing can answer questions like:

• Do my labels make sense to people?
• Is my content grouped logically to people?
• Can people find the information they want easily and quickly? If not, what’s stopping them?

‍

Treejack is our tree testing tool and is designed to make it easy to test your information architecture. Running a tree test isn’t actually that difficult, especially if you’re using the right tool. You’ll  learn how to set useful objectives, how to build your tree, write your tasks, recruit participants, and measure results.

‍

Combining information architecture research methods 🏗

‍

If you are wanting a fully rounded view of your information architecture, it can be useful to combine your research methods.

‍

Tree testing and card sorting, along with usability testing, can give you insights into your users and audience. How do they think? How do they find their way through your product? And how do they want to see things labeled, organized, and sorted? 

‍

If you want to get fully into the comparison of tree testing and card sorting, take a look at our article here, which compares the options and explains which is best and when. 

In UX research we are told again and again that to ensure truly user-centered design, it’s important to test ideas with real users as early as possible. There are many benefits that come from introducing the voice of the people you are designing for in the early stages of the design process. The more feedback you have to work with, the more you can inform your design to align with real needs and expectations. In turn, this leads to better experiences that are more likely to succeed in the real world.It is not surprising then that paper prototypes have become a popular tool used among researchers. They allow ideas to be tested as they emerge, and can inform initial designs before putting in the hard yards of building the real thing. It would seem that they’re almost a no-brainer for researchers, but just like anything out there, along with all the praise, they have also received a fair share of criticism, so let’s explore paper prototypes a little further.

‍

What’s a paper prototype anyway? 🧐📖

‍

Paper prototyping is a simple usability testing technique designed to test interfaces quickly and cheaply. A paper prototype is nothing more than a visual representation of what an interface could look like on a piece of paper (or even a whiteboard or chalkboard). Unlike high-fidelity prototypes that allow for digital interactions to take place, paper prototypes are considered to be low-fidelity, in that they don’t allow direct user interaction. They can also range in sophistication, from a simple sketch using a pen and paper to simulate an interface, through to using designing or publishing software to create a more polished experience with additional visual elements.

‍

‍

‍

Showing a research participant a paper prototype is far from the real deal, but it can provide useful insights into how users may expect to interact with specific features and what makes sense to them from a basic, user-centered perspective. There are some mixed attitudes towards paper prototypes among the UX community, so before we make any distinct judgements, let's weigh up their pros and cons.

‍

Advantages 🏆

‍

• They’re cheap and fastPen and paper, a basic word document, Photoshop. With a paper prototype, you can take an idea and transform it into a low-fidelity (but workable) testing solution very quickly, without having to write code or use sophisticated tools. This is especially beneficial to researchers who work with tight budgets, and don’t have the time or resources to design an elaborate user testing plan.
• Anyone can do itPaper prototypes allow you to test designs without having to involve multiple roles in building them. Developers can take a back seat as you test initial ideas, before any code work begins.
• They encourage creativityFrom both the product teams participating in their design, but also from the users. They require the user to employ their imagination, and give them the opportunity express their thoughts and ideas on what improvements can be made. Because they look unfinished, they naturally invite constructive criticism and feedback.
• They help minimize your chances of failurePaper prototypes and user-centered design go hand in hand. Introducing real people into your design as early as possible can help verify whether you are on the right track, and generate feedback that may give you a good idea of whether your idea is likely to succeed or not.

‍

Disadvantages 😬

‍

• They’re not as polished as interactive prototypesIf executed poorly, paper prototypes can appear unprofessional and haphazard. They lack the richness of an interactive experience, and if our users are not well informed when coming in for a testing session, they may be surprised to be testing digital experiences on pieces of paper.
• The interaction is limitedDigital experiences can contain animations and interactions that can’t be replicated on paper. It can be difficult for a user to fully understand an interface when these elements are absent, and of course, the closer the interaction mimics the final product, the more reliable our findings will be.
• They require facilitationWith an interactive prototype you can assign your user tasks to complete and observe how they interact with the interface. Paper prototypes, however, require continuous guidance from a moderator in communicating next steps and ensuring participants understand the task at hand.
• Their results have to be interpreted carefullyPaper prototypes can’t emulate the final experience entirely. It is important to interpret their findings while keeping their limitations in mind. Although they can help minimize your chances of failure, they can’t guarantee that your final product will be a success. There are factors that determine success that cannot be captured on a piece of paper, and positive feedback at the prototyping stage does not necessarily equate to a well-received product further down the track.

‍

Improving the interface of card sorting, one prototype at a time 💡

‍

We recently embarked on a research project looking at the user interface of our card-sorting tool, OptimalSort. Our research has two main objectives — first of all to benchmark the current experience on laptops and tablets and identify ways in which we can improve the current interface. The second objective is to look at how we can improve the experience of card sorting on a mobile phone.

‍

Rather than replicating the desktop experience on a smaller screen, we want to create an intuitive experience for mobiles, ensuring we maintain the quality of data collected across devices.Our current mobile experience is a scaled down version of the desktop and still has room for improvement, but despite that, 9 per cent of our users utilize the app. We decided to start from the ground up and test an entirely new design using paper prototypes. In the spirit of testing early and often, we decided to jump right into testing sessions with real users. In our first testing sprint, we asked participants to take part in two tasks. The first was to perform an open or closed card sort on a laptop or tablet. The second task involved using paper prototypes to see how people would respond to the same experience on a mobile phone.

‍

‍

Context is everything 🎯

‍

What did we find? In the context of our research project, paper prototypes worked remarkably well. We were somewhat apprehensive at first, trying to figure out the exact flow of the experience and whether the people coming into our office would get it. As it turns out, people are clever, and even those with limited experience using a smartphone were able to navigate and identify areas for improvement just as easily as anyone else. Some participants even said they prefered the experience of testing paper prototypes over a laptop. In an effort to make our prototype-based tasks easy to understand and easy to explain to our participants, we reduced the full card sort to a few key interactions, minimizing the number of branches in the UI flow.

‍

This could explain a preference for the mobile task, where we only asked participants to sort through a handful of cards, as opposed to a whole set.The main thing we found was that no matter how well you plan your test, paper prototypes require you to be flexible in adapting the flow of your session to however your user responds. We accepted that deviating from our original plan was something we had to embrace, and in the end these additional conversations with our participants helped us generate insights above and beyond the basics we aimed to address. We now have a whole range of feedback that we can utilize in making more sophisticated, interactive prototypes.

‍

Whether our success with using paper prototypes was determined by the specific setup of our testing sessions, or simply by their pure usefulness as a research technique is hard to tell. By first performing a card sorting task on a laptop or tablet, our participants approached the paper prototype with an understanding of what exactly a card sort required. Therefore there is no guarantee that we would have achieved the same level of success in testing paper prototypes on their own. What this does demonstrate, however, is that paper prototyping is heavily dependent on the context of your assessment.

‍

Final thoughts 💬

‍

Paper prototypes are not guaranteed to work for everybody. If you’re designing an entirely new experience and trying to describe something complex in an abstracted form on paper, people may struggle to comprehend your idea. Even a careful explanation doesn’t guarantee that it will be fully understood by the user. Should this stop you from testing out the usefulness of paper prototypes in the context of your project? Absolutely not.

‍

In a perfect world we’d test high fidelity interactive prototypes that resemble the real deal as closely as possible, every step of the way. However, if we look at testing from a practical perspective, before we can fully test sophisticated designs, paper prototypes provide a great solution for generating initial feedback.In his article criticizing the use of paper prototypes, Jake Knapp makes the point that when we show customers a paper prototype we’re inviting feedback, not reactions. What we found in our research however, was quite the opposite.

‍

In our sessions, participants voiced their expectations and understanding of what actions were possible at each stage, without us having to probe specifically for feedback. Sure we also received general comments on icon or colour preferences, but for the most part our users gave us insights into what they felt throughout the experience, in addition to what they thought.

‍

Further reading 🧠

‍

• Why You Only Need To Test With 5 Users - Nielsen Norman Group’s Jakob Nielsen explains reveals his thoughts on only using five users when conducting usability testing.
• Sketching for better mobile experiences - Lennart Hennigs explains how to sketch and why you should do it in an article for Smashing Magazine.
• Paper prototypes work as well as software prototypes -  Bob Bailey explains why he thinks paper prototypes are just as good as software prototypes for usability testing.

With UX research so closely tied to product success, setting up a dedicated research practice is fast becoming important for many organizations. It’s not an easy process, especially for organizations that have had little to do with research, but the end goal is worth the effort.

‍

But where exactly are you supposed to start? This article provides 6 key things to keep in mind when setting up a research practice, and should hopefully ensure you’ve considered all of the relevant factors.

‍

1) Work out what your organization needs

‍

The first and most simple step is to take stock of the current user research situation within the organization. How much research is currently being done? Which teams or individuals are talking to customers on an ongoing basis? Consider if there are any major pain points with the current way research is being carried out or bottlenecks in getting research insights to the people that need them. If research isn't being practiced, identify teams or individuals that don't currently have access to the resources they need, and consider ways to make insights available to the people that need them.

‍

2) Consolidate your insights

‍

UX research should be communicating with nearly every part of an organization, from design teams to customer support, engineering departments and C-level management. The insights that stem from user research are valuable everywhere. Of course, the opposite is also true: insights from support and sales are useful for understanding customers and how the current product is meeting people's needs.

‍

When setting up a research practice, identify which teams you should align with, and then reach out. Sit down with these teams and explore how you can help each other. For your part, you’ll probably need to explain the what and why of user research within the context of your organization, and possibly even explain at a basic level some of the techniques you use and the data you can obtain.

‍

Then, get in touch with other teams with the goal of learning from them. A good research practice needs a strong connection to other parts of the business with the express purpose of learning. For example, by working with your organization’s customer support team, you’ll have a direct line to some of the issues that customers deal with on a regular basis. A good working relationship here means they’ll likely feed these insights back to you, in order to help you frame your research projects.

‍

By working with your sales team, they’ll be able to share issues prospective customers are dealing with. You can follow up on this information with research, the results of which can be fed into the development of your organization’s products.

‍

It can also be fruitful to develop an insights repository, where researchers can store any useful insights and log research activities. This means that sales, customer support and other interested parties can access the results of your research whenever they need to.

‍

When your research practice is tightly integrated other key areas of the business, the organization is likely to see innumerable benefits from the insights>product loop.

‍

3) Figure out which tools you will use

‍

By now you’ve hopefully got an idea of how your research practice will fit into the wider organization – now it’s time to look at the ways in which you’ll do your research. We’re talking, of course, about research methods and testing tools.

‍

We won’t get into every different type of method here (there are plenty of other articles and guides for that), but we will touch on the importance of qualitative and quantitative methods. If you haven’t come across these terms before, here’s a quick breakdown:

‍

• Qualitative research – Focused on exploration. It’s about discovering things we cannot measure with numbers, and often involves speaking with users through observation or user interviews.
• Quantitative research – Focused on measurement. It’s all about gathering data and then turning this data into usable statistics.

‍

All user research methods are designed to deliver either qualitative or quantitative data, and as part of your research practice, you should ensure that you always try to gather both types. By using this approach, you’re able to generate a clearer overall picture of whatever it is you’re researching.

‍

Next comes the software. A solid stack of user research testing tools will help you to put research methods into practice, whether for the purposes of card sorting, carrying out more effective user interviews or running a tree test.

‍

There are myriad tools available now, and it can be difficult to separate the useful software from the chaff. Here’s a list of research and productivity tools that we recommend.

‍

Tools for research

‍

Here’s a collection of research tools that can help you gather qualitative and quantitative data, using a number of methods.

‍

• Treejack – Tree testing can show you where people get lost on your website, and help you take the guesswork out of information architecture decisions. Like OptimalSort, Treejack makes it easy to sort through information and pairs this with in-depth analysis features.
• dScout – Imagine being able to get video snippets of your users as they answer questions about your product. That’s dScout. It’s a video research platform that collects in-context “moments” from a network of global participants, who answer your questions either by video or through photos.
• Ethnio – Like dScout, this is another tool designed to capture information directly from your users. It works by showing an intercept pop-up to people who land on your website. Then, once they agree, it runs through some form of research.
• OptimalSort – Card sorting allows you to get perspective on whatever it is you’re sorting and understand how people organize information. OptimalSort makes it easier and faster to sort through information, and you can access powerful analysis features.
• Reframer – Taking notes during user interviews and usability tests can be quite time-consuming, especially when it comes to analyze the data. Reframer gives individuals and teams a single tool to store all of their notes, along with a set of powerful analysis features to make sense of their data.
• Chalkmark – First-click testing can show you what people click on first in a user interface when they’re asked to complete a task. This is useful, as when people get their first click correct, they’re much more likely to complete their task. Chalkmark makes the process of setting up and running a first-click test easy. What’s more, you’re given comprehensive analysis tools, including a click heatmap.

‍

Tools for productivity

‍

These tools aren’t necessarily designed for user research, but can provide vital links in the process.

‍

• Whimsical – A fantastic tool for user journeys, flow charts and any other sort of diagram. It also solves one of the biggest problems with online whiteboards – finicky object placement.
• Descript – Easily transcribe your interview and usability test audio recordings into text.
• Google Slides – When it inevitably comes time to present your research findings to stakeholders, use Google Slides to create readable, clear presentations.

‍

4) Figure out how you’ll track findings over time

‍

With some idea of the research methods and testing tools you’ll be using to collect data, now it’s time to think about how you’ll manage all of this information. A carefully ordered spreadsheet and folder system can work – but only to an extent. Dedicated software is a much better choice, especially given that you can scale these systems much more easily.

‍

A dedicated home for your research data serves a few distinct purposes. There’s the obvious benefit of being able to access all of your findings whenever you need them, which means it’s much easier to create personas if the need arises. A dedicated home also means your findings will remain accessible and useful well into the future.

‍

When it comes to software, Reframer stands as one of the better options for creating a detailed customer insights repository as you’re able to capture your sessions directly in the tool and then apply tags afterwards. You can then easily review all of your observations and findings using the filtering options. Oh, and there’s obviously the analysis side of the tool as well.

‍

If you’re looking for a way to store high-level findings – perhaps if you’re intending to share this data with other parts of your organization – then a tool like Confluence or Notion is a good option. These tools are basically wikis, and include capable search and navigation options too.

‍

5) Where will you get participants from?

‍

A pool of participants you can draw from for your user research is another important part of setting up a research practice. Whenever you need to run a study, you’ll have real people you can call on to test, ask questions and get feedback from.

‍

This is where you’ll need to partner other teams, likely sales and customer support. They’ll have direct access to your customers, so make sure to build a strong relationship with these teams. If you haven’t made introductions, it can helpful to put together a one-page sheet of information explaining what UX research is and the benefits of working with your team.

‍

You may also want to consider getting in some external help. Participant recruitment services are a great way to offload the heavy lifting of sourcing quality participants – often one of the hardest parts of the research process.

‍

6) Work out how you'll communicate your research

‍

Perhaps one of the most important parts of being a user researcher is taking the findings you uncover and communicating them back to the wider organization. By feeding insights back to product, sales and customer support teams, you’ll form an effective link between your organization’s customers and your organization. The benefits here are obvious. Product teams can build products that actually address customer pain points, and sales and support teams will better understand the needs and expectations of customers.

‍

Of course, it isn’t easy to communicate findings. Here are a few tips:

‍

• Document your research activities: With a clear record of your research, you’ll find it easier to pull out relevant findings and communicate these to the right teams.
• Decide who needs what: You’ll probably find that certain roles (like managers) will be best served by a high-level overview of your research activities (think a one-page summary), while engineers, developers and designers will want more detailed research findings.

‍

Continue reading

‍

• Selling your design recommendations to clients and colleagues – Guest writer Jeff Tang outlines some of his techniques for presenting UX recommendations and answering objections.
• Quantifying the ROI of UX – Ashlea McKay delves into one of the toughest UX questions to answer: “What do we get for our money?”
• How do I explain what UX is? – Ashlea McKay covers how you can explain UX to people who may not necessarily have familiarity with the field.