Research Methods

Observational Research in Marketing: Methods, Examples and Limits

Observational research records what people actually do, not what they say. Learn the main methods, examples, strengths, limits and privacy rules.

On this page
  1. Short answer
  2. What observational research is
  3. Main methods
  4. Examples
  5. Strengths of observational research
  6. Limits of observational research
  7. Ethics and privacy
  8. Combining observation with interviews

Short answer

Observational research records what people do by watching them directly, or by recording or measuring their behavior. Because it does not rely on people reporting their own actions, it avoids some of the memory and social-desirability problems of surveys. Its main limit is that it shows what happened, but rarely why. It also raises privacy questions, because it involves recording people. Observation works best when behavior is the question, and it is often paired with interviews that explain the reasons.

What observational research is

Observational research studies behavior in its natural setting. A researcher, a camera, a sensor or a tracking tool records what people do, without asking them questions. A shopper picking up a product, a visitor clicking through a website, or a user struggling with an app are all behaviors that can be observed.

It belongs to the qualitative and descriptive side of research when a researcher watches and notes behavior, and it can be quantitative when behavior is counted or timed. For how it fits with other designs, see types of market research.

Main methods

  • In-person observation. A researcher watches people in a store, a café, a showroom or at an event, often using a checklist or tally sheet. They might note which aisle a shopper visits first, how long they stop at a display, or whether they read the packaging.
  • Recorded and sensor-based observation. Cameras, footfall counters, eye-tracking equipment and smart devices record behavior that is hard to note by hand. Recordings need to be coded, and the coding rules should be agreed in advance.
  • Digital observation. Web and app analytics show page paths, clicks, time on page and where people leave. Session recordings and heatmaps show how people move around a page. These are useful for finding friction, but they are usually collected through tracking tools, which raises privacy questions.
  • Mystery shopping. A researcher poses as a customer to assess service, product availability or how a process works from the customer’s side.
  • Usability observation. A researcher watches someone use a product, app or website to see where they get stuck. Some sessions ask participants to think aloud, which adds a self-report element.
  • Ethnographic observation. A researcher spends extended time in a setting, observing how people live with a product over time. Netnography applies this idea to online communities.

Examples

The examples below are illustrative, to show how observation can be used. They do not describe specific studies.

  • Shelf placement. A supermarket team observes which shoppers notice a new product on a particular shelf height, and whether they pick it up, to decide where to place it in the store.
  • Self-order kiosks. A café owner watches customers use a new kiosk, noting where they hesitate or ask staff for help, to see whether the screens need redesign.
  • Checkout drop-off. An online retailer uses analytics to see at which step of the checkout most shoppers leave, then observes a few sessions to understand what is happening at that point.
  • Packaging. A food company watches shoppers in a store as they compare two package designs, noting which one they examine first and for how long.

In each case, the observation shows the behavior. The reason behind it usually needs a follow-up question.

Strengths of observational research

  • Actual behavior. It records what people do rather than what they remember or think they do.
  • Less affected by self-report bias. People are less likely to adjust their behavior to what they think the researcher wants to hear.
  • Captures the unexpected. Observers can see behavior the research team did not think to ask about.
  • Useful for usability. It is one of the most direct ways to find where a product or process causes problems.

Limits of observational research

  • It shows what, rarely why. A researcher can see a customer hesitate, but not always understand the reason.
  • The observer effect. Some people change their behavior when they know they are being watched, which can distort what the researcher sees.
  • Sampling is difficult. Observers must decide who to watch, when and where. Results from one store at one time may not represent others.
  • Coding can vary. Two observers may record the same behavior differently unless they agree clear rules in advance.
  • Some behavior cannot be observed. Attitudes, intentions, private decisions and rare events are hard to see directly.
  • Time and cost. In-person observation takes many hours, especially when it needs to cover different times of day.

Ethics and privacy

Observation often involves recording people who have not agreed to take part, so the rules matter.

  • Check local law. Recording people, and processing data about them, is regulated in many countries. Follow the rules where you work.
  • Give notice. Signs, notices or a clear statement in an online journey can tell people that recording or tracking is happening.
  • Limit what you collect. Record only the behavior your research question needs, and avoid identifying details where you can.
  • Protect the data. Store recordings and tracking data securely, restrict access, and delete them when they are no longer needed.
  • Avoid private spaces. Observing people in places where they expect privacy, such as changing rooms, is not acceptable.

Combining observation with interviews

Observation and interviews are often used together. A researcher observes a shopper choose between two products, then asks, “What made you pick that one?” The observation shows the choice, and the interview gives the reason. This pairing is one of the most common ways to add depth to observation. For a one-to-one method, see how to conduct in-depth interviews.

Observation can also track behavior over time, which is useful when the question is about repeat purchase. Panels that record purchases over time follow this principle, as described in consumer panels in market research. For more on online observation methods, see online market research methods.

The complete guide to research methods

Types of Market Research: Methods, Designs and When to Use Each
Read the guide

Keep exploring

Guides from other topics

All topics →