Netnography: how to study online communities when surveys fail to capture customers’ language

A questionnaire returns answers to questions that someone previously came up with in a conference room. Meanwhile, on an allergy forum, in a Facebook group for parents of premature babies, or in a Wykop thread about mortgages, customers describe the same problem in completely different words – and point to barriers nobody asked them about. Netnography is a method that makes it possible to collect this material systematically rather than anecdotally.

What is netnography, and when does it offer more than a survey?

Netnography is an adaptation of ethnographic methodology to digital environments: for a specified period, the researcher observes an online community, collects its naturally occurring statements, contextualizes them, and interprets them qualitatively. The term was introduced in the 1990s by Robert Kozinets, who used it to describe research into consumer cultures emerging around discussion groups. The key word is culture – netnography does not count mentions but reconstructs the group’s ways of thinking, values, hierarchies, and vocabulary.

The difference from a survey is fundamental and concerns the source of the stimulus. In questionnaire research, the stimulus comes from the researcher: the respondent reacts to a question, scale, and list of responses. In netnography, the stimulus may come from the participant’s own life situation – someone writes a post because they are genuinely looking for a solution, want to complain, warn others, or praise a purchase. The statement is often produced without the researcher’s direct involvement, which can reduce interviewer effects and some distortions resulting from the desire to present oneself favorably. This does not mean, however, that audiences, community norms, or the specifics of the platform have no influence on the content of the statement.

Netnography is particularly useful in situations where traditional interviews or surveys fail for structural reasons. Below are typical contexts in which online community research provides information unavailable through other methods:

• Embarrassing or intimate topics – health, personal finances, addictions, fertility problems, hygiene. In anonymous communities, people may write about matters they would not disclose to an FGI moderator sitting across from them.
• Expert and hobbyist categories – photography equipment, aquariums, automotive, IT, active skincare. Communities develop their own jargon and product evaluation criteria, often differing from marketing language.
• Identifying the problem before formulating hypotheses – when the team does not yet know what it should ask about in quantitative research.
• Reconstructing the decision-making journey – threads such as “what should I buy, budget X” show the actual sequence of considerations, the list of brands under consideration, and the point at which an option is ruled out.
• Analyzing reactions to change – price increases, changes in formulation, packaging redesigns, service outages. Communities may react faster than periodic research measurements.
• Groups that are difficult to reach through a research panel – narrow professional segments, people with rare conditions, users of niche B2B categories.

An important limitation of scope: netnography is a qualitative method. It does not answer the question “what percentage,” but rather questions such as “why,” “in what language,” “in what sequence,” and “in what social context.” Netnography findings are, however, highly suitable as input for questionnaire design – they provide wording, response options, and real-life selection criteria that can then be quantified.

How do you conduct a netnographic study step by step?

Rigorous netnography follows a defined procedure. Browsing forums “by feel” and pasting interesting quotes into a presentation is not research – it is illustrating pre-existing beliefs. The research process usually includes the following stages:

1. Defining the research question and unit of analysis. It is necessary to determine whether the subject of the study is a product category, a specific consumer problem, a brand, or a cultural phenomenon. This determines the selection of communities.
2. Mapping the field. Identifying the places where a given group actually talks: topic-specific forums, Facebook groups, subreddits, Discord servers, comment sections under video content, e-commerce reviews, and threads on industry portals. Mapping ends with a list of sources describing their characteristics – size, activity level, degree of moderation, and dominant user type.
3. Selecting communities based on qualitative criteria. The selected groups are active, topically focused, rich in content, and diverse in terms of participants, rather than simply being the largest.
4. Immersion and observation. Over an agreed period, the researcher follows the community and learns about its norms, roles (moderators, veterans, newcomers), rituals, and taboo topics. This stage distinguishes online ethnography from a one-off data export – without understanding the context, an individual statement can be misinterpreted.
5. Collecting material. Archiving threads, posts, and comments together with metadata: date, source type, the author’s role in the community, number of reactions, and the course of the discussion. Analysis of online forums requires preserving entire threads rather than isolated sentences – the meaning of a statement often only becomes clear in other participants’ replies.
6. Coding and analysis. Open coding, followed by the development of thematic categories, language analysis (vocabulary, metaphors, descriptions of problems and solutions), and reconstruction of typical narratives and points of contention within the group.
7. Triangulation and validation. Comparing findings with another source – IDIs, sales data, quantitative research results, or offline observation. This reduces the risk of overestimating the views of the most vocal participants.
8. Reporting with anonymization. Quotes are paraphrased or stripped of data that could identify the author, and the context is described rather than a specific profile being identified.

Netnography can be observational in nature (the researcher only observes) or participatory (the researcher discloses their role, asks questions, and conducts conversations within the group with the consent of administrators and – depending on the nature of the community and the project – its participants). The participatory variant makes it possible to explore topics in greater depth, but it introduces the researcher’s influence on the situation and requires transparency toward the community.

However, the method’s greatest value lies not in the procedure but in the material it produces. As Hume’s Institute experts point out, “the most valuable aspect of netnography is what no questionnaire can generate – customers’ natural language. The words they use among themselves to describe a problem, rather than the ones suggested by a list of responses. The difference between ‘a long-lasting moisturizing product’ and ‘something that does not get sticky under a face mask’ later determines whether a message is noticed at all.”

This lexicon has highly practical research applications: it provides wording for response options in quantitative research, makes it possible to build realistic interview scenarios, identifies phrases used in search engines, and reveals discrepancies between manufacturer terminology and user terminology.

How does netnography differ from social listening, and what are its limitations?

This is the most common misunderstanding in conversations about the method. Social listening is primarily the monitoring and analysis of mentions: a tool searches the content available to it for defined keywords and may return volumes, changes over time, source share, reach, and automated sentiment classification. It primarily answers the questions “how much,” “where,” “when,” and “how often,” although its findings require qualitative interpretation. Netnography is qualitative immersion in a selected community, conducted by a researcher with the goal of understanding cultural context and meaning. It answers the questions “why” and “what does it mean to them.”

The practical implications of this difference are as follows:

• Unit of analysis. Social listening often operates on individual mentions and aggregates, although it may account for the context of discussion; netnography focuses on the thread, relationships between participants, and group norms.
• Reach vs. depth. Monitoring covers a broad range of sources, while netnography deliberately limits itself to a few communities and analyzes them in depth.
• The researcher’s role. In social listening, the analyst interprets data and aggregates generated by the tool; in netnography, the researcher is the key research instrument.
• Content availability. Many important discussions take place in closed groups and forums that monitoring tools may not be able to access. A researcher conducting netnography may gain access to some such spaces, but this does not exempt them from the obligation to follow their rules and make appropriate ethical decisions.
• Sentiment. Automated sentiment classification may fail when dealing with irony, jargon, and conditional statements, which is why its results require qualitative verification.

The two methods complement each other well: social listening makes it possible to identify where and when the intensity of discussion is increasing, while netnography explains what is happening in that discussion. Consumer ethnography conducted offline remains a separate category – observing behavior at home, in a store, or in the workplace. Offline ethnography helps observe what people do; netnography shows what and how they say about it to one another. Declarations and behavior can differ, which is why projects covering the entire consumer journey use both perspectives.

However, netnography has limitations that should be understood before commissioning a study. The most important are:

• Lack of representativeness. Active participants in online communities are an exceptionally engaged group. Proportions observed on a forum must not be extrapolated to the population.
• The vocal minority effect. A small group of users may generate a large share of the content, while the silent majority remains invisible.
• Lack of demographic data. It is rarely known who the author of a statement really is, and profile declarations may not reflect reality.
• Sponsored content and bots. Some recommendations may be commercially driven or generated inauthentically. Netnography should include an assessment of the credibility of the material, although it does not always make it possible to conclusively confirm an author’s authenticity.
• Risk of confirmation. A researcher looking for confirmation of a hypothesis will always find it on a forum. Systematic coding, team-based analysis, and reporting statements that contradict the main narrative help protect against this.
• Legal and ethical issues. Public availability of content is not the same as the author’s consent to its analysis and quotation. The processing of personal data is subject to GDPR, and the terms of use of some platforms restrict the collection of content.

When should you commission netnography, and when should you use another method?

The following list organizes typical decision-making situations related to method selection in online qualitative research:

• Netnography – when the goal is to understand language, cultural context, non-obvious barriers, and the way customers themselves define the problem; when the topic is embarrassing; when an initial understanding of a new category is needed.
• Social listening – when volumes, time trends, share of discussion, and ongoing monitoring are needed.
• Online IDIs or FGIs – when further probing is necessary, control over sampling is required, and specific hypotheses need to be verified with a known respondent structure.
• Offline ethnography – when the subject of interest is actual behavior rather than statements about behavior.
• Quantitative research – when the scale of a phenomenon previously identified qualitatively needs to be estimated.

A common and effective approach is a sequence of netnography at the exploratory stage, followed by interviews that explore selected topics in greater depth, and finally quantitative research to quantify the findings. The reverse order – netnography after quantitative research – is sometimes used when numerical results show an unexpected distribution of responses and the reasons behind it need to be explained.

Frequently asked questions

How does netnography differ from social listening?

Social listening is primarily the tool-based monitoring and analysis of mentions: based on keywords, it can return volumes, trends, reach, and automated sentiment. Netnography is qualitative research conducted by a researcher who immerses themselves in a selected community, analyzes entire threads in context, and reconstructs the group’s meanings and norms. Social listening primarily shows how much is being said and when; netnography explains why and in what language.

How do you select communities for a netnographic study?

Sampling is purposive, not random. It assesses activity (the regularity of new threads and responses), topical focus, richness of statement content, participant diversity, the level of moderation, and the presence of sponsored content. Usually, several different types of sources are selected – for example, an expert forum, a support group, and product reviews – in order to capture different registers of conversation rather than the perspective of a single subculture.

What ethical principles apply to netnography?

The foundation is distinguishing between public and quasi-private spaces: the more closed the group and the more sensitive the topic, the greater the obligation to consider obtaining administrators’ consent, disclosing the researcher’s role, and – depending on the project – informing participants or obtaining their consent. Anonymization is required – removing usernames, paraphrasing quotes so they cannot be found through search, and avoiding screenshots containing identifying data. The processing of personal data is subject to GDPR, and the use of material should comply with the platform’s terms of use and industry standards, including the ICC/ESOMAR Code.

Ask about netnographic research in your category

If you need to understand the real language and decision-making context of your customers, Hume’s Institute will design a netnographic study tailored to the specifics of your category and the communities available. Contact the Institute to discuss the project scope and possible timeline.