This chapter of the UX Research Field Guide introduces field studies as a family of research methods, explaining some of the different types and what they are for. The field study is not one research method but a whole category, including all studies of users in their "natural environment," so to speak. This means homes, workplaces, streets, neighborhoods, parks, shops, and other contexts in which your product might eventually be used. If, in a research study, you travel to your participants rather than inviting them to come to your office or lab, you’re conducting a field study.
Field studies provide the most complete, unbiased picture of what potential users actually do. Unfortunately, field studies are expensive and can be time consuming. For these reasons, most research teams use them only when no other method will do.
Field studies are useful for providing context. If your product is designed to function in a particular context, testing in a lab might not give you accurate results. For example, to really find out how people slice their onions, you might want to observe them in their own kitchens. You can also use field studies to explore a need that you hope your as-yet-uncreated product will fill. For example, you could observe the staff at a hospital to look for places where your company might be able to help lower costs or improve outcomes.
Diary studies or surveys can give you real-world data too, but can leave out details that the participants don't realize are important or assume that you already know. They may even bend the truth to appear more favorably to you! With a field study, you know what happens because you watch it happen.
What do you do when you don't know enough about your potential customers to ask good questions about them? What do they need? What do they want? What factors impact their choices? Where do you focus your research? A poorly-designed study—one organized around an inapplicable question—is worse than useless, being not only uninformative but also misleading. A field study is a good way to get the information you need to begin your research in this case.
Even if you can lab-test your product, doing so might not be practical. For example, if your device is designed to function as a component of the navigational system of an oil tanker, going to an oil tanker for final testing makes sense. You probably have an R&D budget for field studies if you're the oil tanker business, too.
There is no bad time for a field study, but you will likely get the most bang for your buck by going out to the field early in the process. In fact, since field studies are excellent for providing an overview of a given context, you can use the same study to inform the development of many different products (all to be used in similar circumstances) at the same time. That way, the cost is spread over the budgets of multiple products. In some cases, field studies can be used in the final testing stages to fine tune details, course correct, or add additional layers to the project.
Field studies can return almost any kind of information you need, but they are most useful for building the groundwork that will help you design more focused studies later. Some of the general questions field studies can address include:
You probably talk about your product and all the need it fills using technical, somewhat jargony language. Use the same language in surveys, interview questions, and recruitment efforts, and you will likely either confuse people or simply put them off. Whether your potential users belong to a culture or subculture different from yours, or simply work in a field unrelated to UX research—and therefore have their own, very different, professional jargon—if you want to be taken seriously, you have to learn their language. Listening in the field is a great way to learn the language.
In designing a product and bringing it to market, you are trying to solve a problem. The problem might be dire (how to keep smartphone users from texting while driving) or relatively mundane (how to change the date of an appointment in a calendar app in fewer steps), but your product will solve somebody's problem. Unfortunately, if your target customers have a life context very different from yours, you are likely to misunderstand, and be unable to solve, their problems. Field studies are an excellent way to get to know your customers well enough that you will be able to offer them something they find meaningful and valuable.
By visiting multiple study sites, or one study site multiple times, you can gain insight into how the situation you are creating your product for varies, what commonalities your potential users have, how they differ from each other, and how adaptable your product (and your marketing) will have to be.
If you don't ask the right questions, you won't get the right answer. For example, if men and women use your product differently, then you'll have to control for gender in your study designs. If the ambient air temperature changes what users need from your product, then you'll have to test it at multiple temperatures. If the age or educational status of your users matters, then you'd better not recruit all of your study participants from an anthropology professor's classroom.
A field study is where you look for patterns that will inform the questions that you ask later.
Again, interviews or surveys can get you some information on what matters to your users and what doesn't, but won't help where interviewees or respondents are unaware of how much something matters, or if they assume you already know and don't need to be told.
The most obvious potential pitfall of field studies is that they are expensive, due to the need to travel, the number of hours researchers need to commit, and the sometimes complex analysis that open-ended, unstructured research requires. It's not that field studies are necessarily difficult—most don't require specialist anthropological researchers, and "time consuming" generally means several days out of the office, not weeks or months. But field studies themselves can't usually be automated or sped up through the use of technology–though tech is easily employed in other stages like recruitment of participants, analysis management, and transcription. As for the study itself, however, it still relies on old fashioned patience and observation. So while the cost of other forms of research has come down, the cost of field studies remains high.
Field studies also stop being an option if your study design calls for a large number of observers—a crowd of researchers pretty much destroys the naturalness of a natural setting—or for recording equipment that aren't an option in the field. If your product is for use in rare, unpredictable circumstances (first-responder mobilization after an earthquake), or places you can't send a researcher (battlefields), then you can't do a field study. If you want to collect sensitive, confidential information, you might have to retreat to the more controllable circumstances of a lab.
Alaska Airlines has been investing in biometrics technology for the past several years to help make the process of boarding planes faster, more efficient, and safer. With a fleet size of 154 aircraft, one of the company’s goals is to create positive customer experiences. For 9 consecutive years J.D. Power and Associates has ranked Alaska Airlines as having the highest customer satisfaction for airlines.
In 2014, Alaska Airlines began building out early versions of a platform that would replace existing ticketing technology. The company had a range of options for solutions to build, but wanted to focus on what customers would value most. Alaska Airlines hypothesized that simpler identification processes and personalized travel experiences would boost customer satisfaction rates.
The research question that the company explored was how to best deliver a seamless identity verification experience. To answer this question, Alaska Airlines set up shop for an experiment in major metropolitan airports, testing 3 different approaches to identification. The first two were face- and fingerprint-based biometric readers. The third was a smartphone-based scanner.
Alaska Airlines didn’t want to risk building technology that customers wouldn’t use, so they ran low-tech experiments. These early iterations, run by people rather than fully- or even partially-built technology, consisted of small experiments to see what customers felt most comfortable with.
Field research guided Alaska Airlines through the process of building a V1 prototype. Members of the airline’s product teams spent time at the San Jose Airport in Silicon Valley, California, learning from travelers’ experiences. In one study, Alaska Airlines representatives recruited field study participants from airport lounges, using the “interception method,” asking if customers would be willing to sign up to use a fingerprint scanner.
The airline’s consumer insights team was hoping for 50% to 60% participation. Actual results were well outside this range—more than 80% signed up. Following the registration flow, Alaska Airlines researchers wanted to know if their customers enjoyed the experience, hypothesizing that 80% would rate their interactions with Alaska Airlines as positive. The actual result was that 92% of travelers who used the tech for check-in said it was good or very good compared to the existing process.
With this data in-hand, Alaska Airlines was empowered to pursue phase II of this initiative, which involved building a V1 of a biometric fingerprint scanner. A year after launching its first in-airport traveler satisfaction experiments, Alaska Airlines built and launched technology that allows passengers to check-in using their fingerprints instead of a boarding pass.
There are two main phases of conducting a field study: the preliminary planning period and the part that actually involves working with participants.
If your study is relatively simple, or if you’re replicating a design that has worked well for you in the past, preliminary work can be quick and straightforward.
Field studies, by nature, can be more open-ended than other types of research, so you don't necessarily need a specific set of research questions but rather defined topics of study. Field studies exist on a continuum from strictly observational, where the researcher is as unobtrusive as possible, to designs involving interviews and product testing. You need to decide where on that continuum you need to be to explore you topic properly.
A related question is who, other than the participants, is going to be involved? Do you need the help of partner organizations? How many researchers do you need to send into the field and for how long? Are there other people, such as internal stakeholders, who will want to be present, and does it make sense to include them? Are there observers who absolutely must be present, such as parents or teachers in the case of a product for children?
If you need to have fifteen stakeholders observing your research at every step, an unobtrusive, strictly observational study probably is not an option.
There are multiple research methods that all fall under the heading of field research. These fall into three broad categories, though the same study may include methods from multiple categories.
Direct observation means simply watching somebody (or a group of somebodies) to see how they behave and why. Ideally, your subject does not care that you are watching and acts exactly as if you were not there. Under some circumstances, you may be able to hide yourself. Researchers sometimes watch shoppers in a mall without anyone knowing they’re being watched: malls, gift shops, department stores. Researchers can observe and take notes on visitor behavior in order to find ways to improve exhibits, services, or visitor management, and most people never even notice they’re being watched. However, there are both ethical and practical limits to incognito observation. In most cases, you will have to explain your presence to participants, and hope they act naturally.
Data recording may take the form of free-form notes, the use of structured protocols and data sheets, or audiovisual recording (supplemented by the researcher's notes).
Direct observation can stand on its own, but is also a great way to gather the information you need to structure later phases of the study.
Participant observation means that the researcher either joins the group of people being studied. Data recording is usually by field notes or diary entries written after the researcher has ceased observations for the day. The classic example of the method is the anthropologist who goes to live with some remote tribe for years on end, but an equally valid example would be a market researcher who makes a habit of inviting herself to cook-outs in order to identify design flaws in popular grill models. The ethnographic method combines informal qualitative interviews with direct observation, except that the researcher has given up on trying to be the proverbial fly on the wall. Instead, the hope is that the research subjects will act normally because they perceive the researcher as one of their own.
Ethically, participant-researchers must identify themselves as such; the word for incognito participant observation is "espionage." Whether people who know they are research subjects will ever act entirely normally is difficult to say, but ideally their behavior is close enough to normal to yield valid data.
Qualitative interviews vary from informal and spontaneous to formal or even structured. Interviews may be a part of a participant observation study, or they can exist independently. The more formal and structured the interview, the easier its data will be to analyze. The less formal and more open-ended, the more likely the interview will be to provide information the researcher didn't think to ask about. Some qualitative interviews in the field are essentially the same as those in the lab, except that they occur on the interviewee's "home turf" and can be spontaneous. Below is a guide for generative interviews that may serve as a useful template here.
Participants in formal interviews will most often give informed consent, documented by consent form.
Informal interviews are simply conversations, though the researcher may have certain areas of inquiry in mind.
A semi-structured interview is formal, and has a designated area of focus. The researcher seeds the discussion with the same series of questions or prompts for every interviewee.
Structured, open-ended interviews allow the researcher to follow a set script, and the questions are carefully planned so as to not suggest answers. For example: “You want electronics that are easy to use, right?" Leading questions are avoided, as always in good qualitative research.
While potentially open-ended, a field study must still be well-planned from a logistical perspective. For example, how long are you going to stay on the study site and where are you going to sleep and eat? How are you going to gain access to your participants? Do you need people to sign consent forms? Your plan need not be complex or involved, but once you are in the field you'll want everything to go smoothly.
Decide also how you’re going to record your data and make sure everyone involved is familiar with your chosen methods.
Your study location may be obvious. If you’re dealing with the public transportation system of a certain city, you will certainly conduct your research in that city. But if you want to learn about schools, construction sites, or hospitals, you need to pick a specific school, construction site, or hospital. In either case, preparation also includes forming the professional contacts and alliances you will need to function in your chosen setting.
Depending on your study design, you could simply be able to show up and start collecting your data. For example, say you have been hired by a museum to create interactive signage for the exhibits. When you field test the first sign, all you really have to do is show up with a notebook or something to shoot video. But for many studies, you will have to introduce yourself to your participants, explain what you are doing, and get your participants comfortable with you. In some circumstances, the introductory process will take five minutes, but in other cases you'll have to work to find a place in the community you have come to study.
Collecting the data might be the simplest part of the entire exercise. Some field studies can stretch on for months or years, but in most cases you can actually get a lot of great data in just a few hours a day for a few days.
Analysis is where you come to understand what you have observed. So far, you’ve gathered your observations and your anecdotal data, but in the analysis stage you can better organize and target your results into data.
You might also find you want to dig deeper, in which case, you may choose to return to your participants for interviews. In either case, another round of analysis awaits you after you have finished the entire study. Highlight key observations in the moment and look for patterns as you go.
Whenever you finish in the field, pay attention to the process of leaving. Unless you are simply taking notes on crowd behavior in public, don't just vanish. Let your participants say goodbye to you. As with introductions, farewells may take five minutes or several weeks, depending on your situation. Take the relationships you have made during your research seriously.
Finally, after all your data are gathered and analyzed, it's time to verify your results with your participants (if applicable), get your notes and other materials into a form where they can be archived (if applicable), and write your report (always applicable).
The logistical side of field research revolves mostly around travel on the one hand and documentation on the other. Consider the below to plan your tools and logistics for a field study:
An important tool of the trade is always courtesy and professionalism. Conducting research on people can be a socially awkward thing, and you must behave in a respectful and trustworthy manner at all times or no one will cooperate with you.
Part of respect and trustworthiness is reciprocity; you are getting something from your study participants, so you should give something of equal value back. The reward can be physical or financial, though it need not be in all cases. Sometimes you just need to say "thank you." If your field study is a product test, allowing your participants to keep your product may be appropriate. In some cases, you may find yourself literally throwing a party for your participants. Free food goes over well. Whatever reward you offer, include it in your budgeting and other planning.
Field studies are generally only the initial stage of a larger research effort. The fieldwork explores the context and helps set the research parameters for the rest of the project. Some aspects of a field study can grade into laboratory-type research as you are essentially creating a temporary lab on site. Explore these discovery research methods to continue your research design. And stay tuned for more validation and post-launch methods coming soon.
Keep it simple
Remember that field research need not be complicated, does not have to take an excessive amount of time, and does not have to involve specialist researchers. In fact, while a good field study always yields a report that can be read and understood by people uninvolved in the study, part of the point of field research is to get out there and talk with the people who will use their product. Get to know and understand them. If you can develop an intuitive (and accurate!) feel for your customers, you will be much better at creating and selling products that feel intuitive to them.