Benefits and Limitations of UGC

UGC (user-generated content) supports large number of conversations; conversations that emanate from a non-representative albeit influential group of consumers. Information gleaned from these conversations tells us how these highly engaged consumers, who publish content or comment on the web, perceive the brand. However, because the sample is so heavily skewed by individuals with fairly strong opinions, their perception or views may not accurately reflect how consumers in general think about the brand.

Marketers too influence opinions as they interact with consumers online. It serves their interest to engage with these consumers, feed them with relevant information and in so doing favourably influence their opinions. While this helps to protect and strengthen their brand, it introduces a further bias on UGC.

Owing to the lack of rigour and discipline, UGC conversations are not representative of the universe, and may contain a strong bias considering that:

1. Consumers who are engaged and vocal on the net tend to have strong opinions that do not necessarily reflect the population as a whole;
2. their opinions are influenced by other consumers who are active online, as well as marketers who have vested interest in advocating and protecting their brands;
3. and consumers who seek information on categories like fast moving consumer goods online are, in numerical terms, minute compared to the universe of consumers.

In view of the above, interpretation is crucially important. For instance, a media story on Vegemite stated that “IBM research revealed the black spread made down under is the world’s most loved brand on the internet”. That is hardly credible, because outside of Australia, most people, both online and offline, have not even heard of Vegemite. (The metric employed was the ratio of total mentions of the brand to mentions containing expressions that could constitute “love” to consume the brand. While this is a meaningful ratio that reflects affinity amongst those who mention the brand, it remains a ratio, and should not be construed as a measure of “most loved brand” in an overall sense. Besides the choice of metric, a strong bias also creeps in due to the brand’s extensive crowdsourcing campaigns).

While representativeness is important where the intent is to measure, as in quant, it is not of critical importance where the intent is to explore, as in qual. In qual we often (though not always) target engaged consumers, people who like to talk about the brand. Consumers who are actively engaged with the brand online would be appropriate, even sought after, for exploring and investigating issues relating to your brand.

One must, however, remain wary of some of the claims that we stumble upon on this subject. Though it has been touted as the “world’s largest focus group”, the internet is not a focus group. It is not a purposeful sample. It lacks moderation, the discipline in approach, and is devoid of projective techniques that researchers use to unlock people’s minds in focus groups or depth interviews. It does not therefore substitute qualitative research.

It is, nonetheless, a powerful, revealing medium for observation. It allows you to “listen” to unsolicited feedback about your brand from hundreds or thousands or millions of consumers, and “see” how they relate to it and how they use it. As natural language processing and text analytics technologies evolve, you increasingly can listen more efficiently to the glut of conversations on the net. The insights that you glean through listening and seeing enable you to:

• Spot market gaps and trends;
• Generate ideas and insights for new products (ideation);
• Learn about existing needs and preferences; uncover new “organic” needs;
• Uncover feelings and emotions arising from marketing initiatives and activities;
• Understand what product categories, product attributes, themes, and which competing brands are associated with your brand (brand image);
• Test advertising;
• Gauge the extent to which consumers advocate your brand (consumer advocacy).

UGC can trigger new ideas, thoughts and perspectives. You are more likely to encounter the unforeseen via UGC because it yields much more unsolicited information than conventional research which is primarily guided by brand management’s existing thoughts and notions about the brand. UGC also reveals consumer vocabulary, i.e., the labels and adjectives used by ordinary people to describe brands and brand-related concepts or themes.

Moreover UGC-based research need not be expensive. Considering the necessity of monitoring online conversations to protect or enhance the reputation of your brand, the incremental cost of drawing further insights from these conversations is not high and will continue to decline as related technologies improve.

UGC Complements Conventional Market Research

From the perspective of using UGC for research purposes, the iSnack 2.0 fiasco highlights some of the dos and the don’ts. While it is a very rich source of information for marketers, UGC should not form the basis for measurement or validation. It does, however, have the potential to complement, support and strengthen the processes that constitute conventional market research.

The unsolicited information garnered through UGC feeds conventional research. Consumer vocabulary on the net may help refine the language we use in both qual and quant research. The new ideas and thoughts that emerge from UGC may be explored in greater depth through qual. Quant on the other hand helps to measure and validate, draw conclusions, and assess the extent to which the ideas relate to the masses.

In conclusion, UGC-based research works best in conjunction with conventional market research. While it lacks the rigour and discipline that is necessary to form conclusions about the market as a whole, it is a very valuable medium that allows marketers to listen and engage with a large number of consumers, uncover new product ideas, and spot market gaps and trends. The insights that it generates need to be further explored, measured and validated via conventional research.