A birds-eye view to observe baristas making drinks in a coffee shop.

Contextual inquiry is should not be confused with user interviewing, another very valuable user research method. The key distinction is right in the name — your inquiry must be on context.

So, what is contextual inquiry? Think of it as a combination of light ethnographic work and user interviewing.

Contextual inquiry involves observing people in their natural context and asking them questions to fill in the gaps of your observation. If I’m looking to create a more efficient point of sale system for grocery stores, I would observe someone using an existing point of sale system, taking notes on what works and what doesn’t work.

Contextual inquiry would be an excellent way for us to discover ways to improve the checkout process at a grocery store from both the cashiers perspective and the people buying goods. A user interview where you just ask people pain points is not going to be specific enough or in context enough to yield enough actionable data about that specific system or store.

On the other hand, if I am looking to build an application to help new parents track when their babies last went to the bathroom and wanted to make sure it was easy to use in real-world settings, I would observe new parents in their homes and wherever else they may take their babies.

This is a key point of contextual inquiry: The context of your inquiry depends on what and who you are trying to study. While you can conduct a user interview over the phone, you cannot do contextual inquiry unless you are actually in the real context of what you are trying to learn about

The core of contextual inquiry is

  • Observing
  • Inquiring
  • Documenting

We observe people doing different activities. Why not just ask them after the fact? People don’t remember every detail of what they do, and they will usually give you the high level beats when you interview them.

By observing people, you’ll notice every little workaround they do to get an activity or task done. You’ll observe things about a person that they themselves might not even know.

Contextual inquiry tips and check list

Record if possible

Always ask permission to see if you can record your contextual inquiry sessions. Video is preferable, as you can later on review actions, but audio will also help as well, particularly during portions where you are asking questions. Recording can be immensely helpful for you when you want to review everything, but it can also help those who weren’t there for the session better understand what is happening. Remember to ask permission upfront and go over ground rules. Are there things you can record but others you can’t? Establish this first.

Here is a video of someone conducting contextual inquiry of a cab driver. This observation in context will yield better data than a user interview out of context would. Notice, however, that contextual inquiry also involves asking questions.


Take detailed notes

Recording or not, take detailed notes. This is a combination of step-by-step notes about what a person is doing, direct quotes that are interesting, and observations about what is and isn’t working.

I always carry around a little Field Notes notebook on my back pocket along with one of my favorite pens. This allows me to jump into observations at any time, even if I wasn’t intending to do observations. I also use these little notebooks to help me sketch up different thoughts and ideas during contextual inquiry.

Take photos

Taking notes is helpful. Recording any direct observation or interview responses is also very helpful, but grabbing photos and creating sketches can help give you a sense of key issues and can help give you a better sense of place. Take photos of everything you can and make sure to label them when you are done so that you understand what you are looking at.

In addition, you can combine your photos with some of your observations and quotes to give you a better sense of what you have observed and learned. This will also be very helpful when you are sharing your user research with others. Even a simple Google Doc file that combines different artifacts from contextual inquiry can help put all of your data together.

In the grocery store point of sale system example above, having photos of what the system looks like, along with different screens and buttons will help everyone better understand the problem. If your research uncovers a major usability issue with part of the point of sale system, having photos of the issue and of a real person working with that system will help you during the next phases of your work.

Below is a photo of a modern touchscreen point of sale system using an iPad. In this example, you are going to want to do a combo of activities with a situation like this. Takes photos of key interactions during your contextual inquiry, observe what is happening and take notes, ask clarifying and expansive questions, and record video if you can.

Photos of the point of sale system in action can help give your research a sense of place, and will be very powerful when combined with observations and quotes.

Create sketches

What’s the difference between taking photos and creating sketches for contextual inquiry purposes? Photos are of things that exist today. Sketches can be of the way things should work.

Sketches can also be a way to annotate what is or isn’t working. Sketches can also be a way to create a conceptual model of what you are observing. Having a good conceptual model of how a current system works will help you create better recommendations and solutions later on.

I also often find that contextual inquiry and user interviewing jog my creativity. I begin to see the world more clearly, and I do some of my most creative work when exposed to these kinds of user research. I often start sketching up ideas while I am conducting the user research.

Your sketches should remain high level, however, as you don’t want to collapse upon solutions too soon, but sometimes my best ideas come when I am witnessing a person use a product or system. If you wait weeks or months to start doing any sketching, some of what you learn may get lost in translation.

Prepare questions ahead of time

Just like with good user interviewing, you should put a lot of thought and care into any questions you might ask. And just like with good user interviewing, you should ask the same questions of everyone you are working with. If you don’t prepare questions ahead of time, the questions you ask will be ad hoc, and you won’t ask everyone the same questions.

It’s important when gathering good quantitative data that you are controlling for variables. One variable is that you ask everyone the same set of rigorous and thoughtful questions.

This doesn’t mean you won’t ask additional questions. This means that you will ask the same core questions.

In any user interview or contextual inquiry session, questions will come up that you’ll want to ask. Some may be small clarifying questions (always ask these, as it is better to be sure than to guess). A round of contextual inquiry will involve observing several different people, even if they are doing the same thing. Your follow-up questions may differ by person, because you may be seeking to understand something that is slightly different about this session than a previous one or a person might be doing different things.

Just remember to work on your core set of questions ahead of time. Run them past other people to make sure they make sense. It’s even good to pilot test them on one or two users to make sure your questions work well and are helping you find out the information you need.

Test the tests.

It’s usually better to ask for forgiveness

Usually being the operative word here.

This advice does not apply to certain areas where asking for permission first is a must. Do not do any contextual inquiry in a hospital, for instance, without getting express permission and having a chaperone the entire time. But a lot of contextual inquiry is on much less privacy focused areas or on more seemingly mundane topics.

I had one team last year work on trying to make improvements to a large dining hall at University of Maryland. First, they just needed to observe how people used the dining hall, how everything was laid out, and just get a general sense of how everything worked. They took notes and photos.

This initial observational work helped inform the questions they were going to ask of people eating at the dining hall and of various stakeholders.

The university, for whatever reason, is skittish of people taking photos in the dining hall of the food. I don’t know why, and I don’t care. But they kicked the team out.

Had they decided to ask for permission ahead of time, they probably would have been immediately shut down. Instead, they got the observation work they needed done, and then had to ask for forgiveness when they were kicked out. Forgiveness was granted.

This observational info was key, however, because they needed to understand the lay of the land for dining halls and nutrition at University of Maryland. Without this observational work, it is doubtful they would have been able to provide a meaningful solution to some of the problems that their interviewing uncovered.

This team then identified students to interview who ate at the dining hall. They also interviewed people working at the dinning hall. They were able to get all of the information they needed, but it required a dose of forgiveness to get it done.

For certain areas, such as anything healthcare related, always ask for permission first. Asking for forgiveness is only for areas where there isn’t a rational reason for someone telling you no, but you might suspect they will tell you no anyway.

Doing this in stages may work best

Contextual inquiry doesn’t all have to be done at once. Sometimes it is best to split up the work into multiple stages.

Often observing first without doing any inquiry and interviewing helps set up the rest of the work better. Your initial observations will help you understand the problem space and will help you formulate what questions you might want to ask.

In the grocery store point of sale system example, you probably want to just observe people checking out at a grocery store and what sales clerks have to do for each order. You will also want to observe as much as possible in the store. You don’t want your user research to be too narrow early on, and observing people doing different activities around a grocery store help you understand the entire system better.

Once you have done your initial observations, it is time to write questions and figure out what you need to take a deep dive into. You may realize that you need to do separate contextual inquiry on point of sales systems used by cashiers in standard lines, by the people who work in customer service, and of the automated, self checkout systems. You may also realize that you need to better understand how items and prices are put into the point of sale system.

By first observing before jumping into the rest of this, you can make sure that you cover everything that needs to be researched.

Below is a student performing contextual inquiry on someone grocery shopping. Notice how this goes from before the shopping trip, to observing during the trip, to a post-trip interview. It’s a good, quick demonstration of contextual inquiry. The only thing I don’t recommend are the parts where the video is sped up, as it’s good to watch some of that data at real time speeds.