Conjoint Research

Conventional quantitative research (of which online research makes up the biggest part) is mostly used when asking factual (e.g. employment, holiday spend, shopping habits) or attitudinal questions (e.g. political beliefs, or thoughts on current events). However, we know from both neuropsychology and behavioural economics that not all decisions are simple or easily verbalised and that’s where conjoint research proves the most beneficial for brands and orgnaisations.

To find out about sofa buying habits, we could directly ask respondents questions such as what their preferred brand is, what colours or materials they would like, how much they are willing to spend, etc. While we will get answers this way, we are not mirroring the way human beings make choices in the real world. Consumers don’t see just one attribute of a product at the time and make judgements on each one, they will evaluate products as a whole (gestalt) and weigh up the various elements together.

Conjoint research mimics a real-life buying scenario by presenting respondents with a series of (usually binary) choices based around different attributes to see how people value each one in conjunction with the others. This should always include brand and price as standard attributes, with the other attributes depending on what we are asking about. If we think of our sofa example above, the attributes might be:

  • Material
  • Colour
  • Number of seats
  • If it is a recliner

Then within each attribute, there will usually be 2-4 levels, which are the specific values of each attribute. For example the “number of seats” attribute would likely have options of 2/ 3/ 4/ 5 or more.

Respondents are then shown a series of attributes with the levels mixed up between them, to see where their preferences sit. For example, how does a 3-seater leather sofa costing £499 from Brand X, compare to a 4-seater polyester sofa costing £550 from Brand Y?

Once enough respondents have seen the different combinations and compared, we can analyse the conjoint research data to see how important each attribute is compared to the others (e.g. where does brand sit compared to price, material, colour), and we can also see how much weight an attributes level (e.g. different price points) has with respondents decision making.

See an example of our conjoint research Wine Consumption Habits – Online with Conjoint Analysis – OnePoll