Dickson Fong

This is the second in a three-part series about using research to design more effective web interfaces. Today, in Part 2, I’ll teach you the basics of conducting ethnographic research.

The one-on-one interview

The best way to study your users is to conduct an in-person one-on-one interview. Depending on the logistics of your project, this might not be possible or even warranted. For the sake of this discussion, let’s assume that you can set up appointments with current and prospective users of your website. We will refer to these people as ‘informants’. 

How you go about obtaining your informants is up to you, but it usually involves dedicated networking, phone calls, e-mails, and persuasion. Your client might be able to help you reach out to their internal and external user base. 

Internal user base

The internal user base usually comprises of the client’s own employees. These people are relevant to your website only if they will be using it themselves. Examples include employee portals and intranets. These informants can usually be arranged with the help of your client because they have some control of their employees’ time. Explain to management that you need to schedule 30 minutes with several of their employees to ensure that you properly understand their website’s internal user base. 

They are paying you a lot of money, and it’s in their best interest to make sure that the final product properly serves their employees needs. Emphasize this and your clients will usually be happy to oblige with your needs.

External user base

The external user base usually comprises of the client’s customers and/or target audience. Your client might be able to help you reach out to them. If not, you will have to be resourceful. There are numerous methods for recruiting informants, but I will not be covering them here today.

Beginning the interview

It’s important to put your informant at ease. Reassure them that you’re not there to judge their work habits or their skills. You simply want to observe, ask questions, and learn. Also tell them that their help is greatly appreciated and that their efforts will directly go into creating websites that are better suited for people just like them. You should be gracious that they are giving you their valuable time and help.

Be organized

As with any other interview, you should have a list of questions and topics that you need to address. This helps you organize your thoughts. You may choose to have a pre-formatted question and answer sheet, or you may opt to just write down your questions on index cards and fill in your information on a blank notepad. Choose whatever method works best for you. 

Show and tell

Contrary to what you might expect, these aren’t your typical interview questions. Questions like “What do you think makes a website easy to use?” and “Do you prefer navigation on the left or on the top?” have no value because they inquire about solutions. This is called ‘solutioneering’, and it is not the dialogue that we want to establish.

Rather, we want our informants to show us how they work and tell us the reasoning behind their decisions. By doing so, we can understand their workflow, skills, and knowledge in a way that reveals the problems to us. Only by having a contextual understanding of users problems can we effectively come up with solutions. 

Establishing dialogue

We begin by having our informants perform some tasks on the existing client website. Alternatively, we may also have them use a similar website for the purposes of a competitive analysis.

Begin by having them use the website as they would normally. While some experts think you should run your informants through a predefined set of tasks, I believe it’s better to let your informants do whatever interests them. There are opportunities to learn something completely unexpected if you let people share what they want to share. Chances are, some of the predefined tasks will be covered during the process, and any that haven’t been touched upon can easily be revisited towards the end of the interview. 

As they work their way through the website, remind them to walk you through the process. Ask them to narrate it for you. For example, if they’re moving their cursor around the screen, they should be telling you why they’re doing so. Perhaps they’re searching for something, or maybe they’re confused. Similarly, if they click on a link, they should tell you the corresponding reasoning and goals. 

Digging deeper

People who are unfamiliar with this experience will usually give you uninformative responses. It’s your job to coax more out of them. If they say they’re looking for a link on the top navigation bar, ask them what particular word or phrase they had in mind. If they say they’re looking for “Store Locations”, ask them what information they’re hoping to find on that page. I know this seems like a rather trivial question—obviously, the person wants to find the store location closest to their house, right? 

Not necessarily. In addition to finding the nearest store location, they might be interested in finding driving directions to a particular store, the number of stores in the entire franchise, the number of stores locally, store hours, or a store’s telephone number. 

By unpacking these finer details, we gain insight and information that can later help us tailor the semantics, information architecture, layout, and other design elements to the goals and expectation of our users. 

Recording your data

Write down your observations and information during the entire process, but not so much that you lose track of what your informant is doing and saying.

Unfortunately, I really can’t help you with this. Everybody works in different ways. Find a solution that works for you and stick with it.

Why we’re doing this

Our motive isn’t just to know the actions that our users take, but also the goals that are driving these actions. We can’t possibly solve interface problems if we just see the actions because we can’t adequately identify where the process goes wrong. By understanding what users want to achieve, we can make educated proposals that solve the root of the problem. 

This blog entry isn’t meant to be the definitive how-to guide for ethnographic user research. I provide this as a basic introduction for those who are interested in the merits of user-centered design. Hopefully, I’ve given you a thought-provoking glimpse into the benefits of user research and its methodologies. For those who want to learn more about how to conduct these studies, I recommend Rapid Contextual Design by Karen Holtzblatt. I own this book myself and think it’s a decent resource.

That’s just about it for now, and thanks for your interest. Please join me next weekend for Part 3 in the “User Research For The Web” series, where I’ll show you how to consolidate and analyze your data.

End

RSS comments for this post Previous post Next post Permalink