Conjoint Analysis – Too Many Attributes!

As part of our research and consulting practice, I get asked to review a lot of projects on Conjoint Analysis. One the most common themes is the concept of measuring every single possible attribute. In many client meetings, I’ll sit through the entire talk about how the product manager would like to determine utility and importance on over 15 attributes and be asked if we can support a conjoint model to do that:

Well the technical short answer is – Yes – of course the software can do that. But the practical answer is – No  – Do you know how impossible it is for a respondent to choose between two concepts with more than 5 or 6 attributes? – Think about it – Even when we make big purchases (like cars) we only look at a few key attributes:

a) Utility
b) Color
c) Brand
d) Price

It’s not that other attributes do not matter – its impossible to put a user-interface (like conjoint) and expect users to actually care about the more than 6 attributes – combined with the levels.

BUT – the business problem still remains. The Product Manager still needs to know how to prioritize what to ask in a Conjoint model.

Like mentioned above, the system does not limit the number of attributes. However from a practical presentation standpoint, it really does not make sense to have choices with more than 5-6 attributes because of the cognitive stress involved. However, if you do have a case where you’d like to test out 10-20 attributes we would suggest you do this as a two part project:

  1. Create a screening/profiling survey and use simple “Multiple Choice (Multiple Select)” to determine viability of attributes – “Pick 6 of 20” etc.
  2. Organize the options into 2 columns (for ease of reading – so users don;t have to scroll)
  3. Use TURF Analysis to pick the Top 5 or 6 attributes with the highest reach.
  4. Then as a secondary wave, run the conjoint study on highest reach attributes.
  5. Profit..