How to specify attributes and levels in conjoint analysis


Attributes are ‘dimensions’ of your product (such as price, colour, shape, size, brand, location). Include the attributes that you believe are most important to your customers when they make buying decisions, as well as any attribute whose importance you would like to check. For example, if you know that customers are driven by price and size, and want to investigate whether colour is important, include all three attributes (price, size, and colour). Try not to include more than six attributes because it might confuse your respondents.

Levels are the ‘values’ that each attribute can take. For example, the attribute ‘colour’ can have levels ‘blue’, ‘red’, ‘transparent’. The attribute ‘features’ can be ‘no updates included’, ‘automatic update every month for one year’, ‘lifetime automatic updates’. Keep in mind that you need to have at least two levels per attribute. If you are looking only at one type of product, make the product description more specific.

Key rules for specifying attributes and levels

  1. When describing both attributes and levels, use the language that customers understand (or even pictures) because in their screen they will see attributes and levels exactly as you describe them. Think of the kind of language you would use in a customer-oriented brochure. If you have salespeople in your organisation, you can ask them to test your conjoint study for understandability before sending it out to customers or panel respondents.

  2. Related to the previous tip, levels are like degrees of a characteristic and should be precise: e.g. the levels of the engine power of a car are 1.5L, 1.8L, 2.0L, not “less than 1.5L” or “more than 2L”.

  3. Levels need to be realistic and feasible to achieve real-life scenarios and should include competitive offerings.

  4. Make sure the levels are mutually exclusive within each attribute. For example, consider textures of tissue paper. For this attribute, one may specify the levels “recycled” and “weave-like”. This is not mutually exclusive because recycled can be “weave-like”. This can be avoided by specifying two attributes:
    • Attribute “Texture”: “weave-like texture”, “simple texture”;
    • Attribute “Recycled”: “recycled”, “not recycled”.
  5. If you are working in an already established market (e.g., FMCG or financial services), for best results in market share simulation, we recommend:
    • You can increase the number of attributes and levels (but be mindful that the required sample size will also increase).
    • Make sure to include top brands by sales.
    • Make sure to include any other salient brands (e.g., prominent newcomers to the market).
    • Make sure to include any variants of your product you are keen to understand market’s reaction to.
    • Ensure there are competitor products for all important variants of your products (for example, if you are testing a larger pack size, make sure that you include any competitive brands with large pack sizes - if they exist on the market).
    • Ideally, think of the items (and their attributes and levels) that will be sold two or three months from now, not what is on the market today.

Specifying prohibitions

Conjoint.ly allows you to prevent specific combinations of levels from showing on the same alternative (under “Advanced settings”). For example, a particular price (say, $5) might be incompatible with a particular size (“recycled”). For best results, we recommend using this option sparingly: not more than four restrictions.

Keep in mind that you only need to prohibit combinations that are unrealistic in the eyes of the consumer, rather than those that are not feasible for the company to make. For example, if consumers would not be surprised to see $5 recycled tissue paper, but it is not feasible for the company, you should not prohibit this pair in the design of your experiment. Instead, you should simply ignore product concepts with this combination when you look at the results of the analysis.

If you find that you need to have more than seven prohibitions, we strongly encourage you to think through these options:

  1. If you are using a generic experiment, you might need a brand-specific conjoint study where you can set which levels apply to which brands.
  2. Remove prohibitions of any pairs that are at least marginally plausible in the eyes of the consumer.
  3. Remove prohibitions of pairs that are plausible to the consumer but not feasible for the company to supply (as discussed above, they should not be prohibited in the design).
  4. Make sure to set other advanced settings to “Automatic”.