For example, to study the process of booking rental cars for leisure travel, we recruited subjects that specifically said they had an upcoming leisure trip they had not yet booked a car for to come in for an interview.

When the subjects arrived we executed our interview protocol. A protocol is a cross between an agenda and a script you use to run interview sessions. 

After introductions and going over the logistics of the session, we asked them to tell us about their upcoming trip, and to engage in the process of shopping for and booking their rental car in front of us using a laptop we’d provided or their phone.

The subjects were usually happy to get a “to-do” item out of the way during a paid market research session and, of course, they knew they could always cancel the reservation if they changed their mind later.

We got to observe a variety of subjects who used differing shopping tactics. 

Some went to sites like Travelocity or Expedia that compare many rental car providers. 

Some logged into their employer’s business travel portal even though it was a leisure trip because they wanted the corporate discount. 

Some Googled for rental car companies serving the city they were traveling to. 

And some went to the websites of specific car rental brands that they were interested in, such as Hertz, Avis, or Enterprise. 

One subject just texted her assistant with the travel dates and told her to book whatever was a good deal.

Seeing this kind of detail and “real world” activity is far more accurate than simply interviewing subjects about their travel habits.

HAVE THE SUBJECT VOCALIZE THEIR THOUGHTS 

During these types of observational activities, we ask subjects to “talk out loud” so we can know what they’re thinking. 

It can feel strange to verbally share every thought one has, so we demonstrate. 

We dramatize playfully such as, “Hmm, I need to find where to enter the date. Oh, there it is. Ok, so I’m putting in my departure date. Oh, I see I need to pick it from a calendar,” and so on. 

Subjects generally understand pretty quickly and are usually willing to play along.

This method allows us not only to observe what they do, but based on their words, understand their thoughts, and based on their tone of voice and body language, infer their emotions.

In fact, some subjects did get emotional in our rental car study as they encountered obstacles and frustrations on various sites.

We also occasionally ask subjects questions during their task if we need more information. For example, there may be a key moment where the subject completes an action, an expression appears on their face that seems important to understand, and they completely forget to keep verbalizing their thoughts. 

If you encounter this situation, and it seems like a good moment, it’s ok to ask, “I’d love to know what you are thinking right now,” or “Can I ask why you clicked on that? What were you expecting?” 

The most accurate information about what someone is thinking or feeling is always extracted at the moment rather than later when they need to reconstruct that moment from memory.

The downside is that asking questions does pose the risk of pulling the subject out of their task. 

So, researchers need to use their judgment during these types of observational situations and only interject when necessary.

ENCOURAGE THE SUBJECT TO STAY INSIDE THE TASK

You always want to avoid making the subject feel as though the observational session is a conversation with the researcher. 

So, if you’re asking questions during the activity, it’s important to direct the subject back to the activity as quickly as possible.

Whether or not questions are asked, if the subject’s “talk aloud” turns into an explanation to the researcher of why or how they are doing something rather than their natural “stream of consciousness,” it is time to redirect them.

Remind the subject, “Just focus on the task and verbalize whatever you are thinking and feeling as you’d hear it in your own head. Afterward, I’d love to hear anything else you’d like me to know.”

BE WILLING TO LEAVE THE SUBJECT IN PAIN

Also, it may be hard, but when a user is engaged in a task, don’t help them. 

It’s a natural human reaction. If we see someone struggling, trying to figure out how to navigate a web site or an app for example,  we want to provide assistance. 

After all, the person is experiencing some discomfort, and if we can take away that discomfort by giving them the answer, shouldn’t we? 

Isn’t it kind of cruel to just let them suffer, confused and frustrated? 

If you feel yourself burdened by this moral dilemma, remember the words of Mr. Spock from Star Trek:

“THE NEEDS OF THE MANY OUTWEIGH THE NEEDS OF THE FEW.”

As a researcher, you must suppress the instinct to “save” the subject and instead let them have their own experience while you remain an observer.

Remember, there’s going to be some pain and that is what we are looking for.

You are trying to understand the experiences of customers and, out in the world, they don’t have a researcher hovering over them and able to assist them when they get confused. 

If we use a few research subjects’ pain to discover and fix problems that affect thousands of real customers, then the research subjects have sacrificed their comfort for the delight of a far larger group.

Plus, they are getting paid.

But what if the subject asks you a question? 

They might ask, “I can’t seem to find the search feature, does this app have one?” 

A good response is, “We’re really interested to see how you complete this task without assistance. I’d be happy to answer that question later, but please continue as if I wasn’t here.” 

Observational research is one of the best ways to gather deeper customer insight, especially when done correctly.