External Effects

Top  Previous  Next


The factors evaluated in a conjoint analysis study usually focus on product attributes, ignoring other factors that affect market share.  Consequently, the market simulator (choice simulator) calculates a preference share rather than a market share, emphasizing that important factors for calculating market share are missing from the model.  Such missing factors include: the level and effectiveness of advertising, the size and effectiveness of the sales force, the number of outlets where the product is sold, product awareness, and the length of time the product has been on the market.


The principal value of choice simulators is to indicate the kinds of changes that would make the most favorable differences in customer preferences.  There is no real need for them to match market shares.  All the same, it is sometimes disconcerting for those viewing results to see shares of choice that are very different from known market shares.


External Effect Options


The External Effect options help to account for factors outside the model.  When used properly, External Effects can lead to more realistic predictions.  The External Effect options are:


oApply Product Availability (Multi-store)

oApply Product Awareness

oApply Share Adjustment


Recommended Procedure


If you do plan to implement External Effect options, we recommend the following order of application:


1.First adjust for Product Availability

2.Next, apply any Product Awareness corrections

3.Third, adjust the Exponent

4.Lastly (and this step should only be done if the business situation requires it and all other avenues for improving the validity of the model fail), apply Share Adjustments


Step 1 (Product Availability adjustments) is a quite defensible and correct procedure, assuming you have accurate data about which products are available in which stores (regions).


Step 2 (Product Awareness corrections) makes the strong assumption that people who are not aware of a product will not buy it.  This assumption may or may not be right for your situation and for the product category you are studying, and since it also requires asking additional awareness questions you may decide to omit this step.


Step 3 (Exponent adjustment) is a quite defensible procedure, since the magnitude of respondent errors in the conjoint questionnaire may not necessarily reflect the magnitude of buyer errors in the real world.  Adjusting the Exponent tunes the degree of "flatness" or "steepness" in the share of preference probabilities, but usually does not change the rank order of product preference.  However, it can be challenging to find the right Exponent tuning factor because doing so requires accurate target share data and also assumes that scale factor accounts for the only remaining difference between simulated shares of preference and target shares (after applying Steps 1 and 2).


Step 4 (Share Adjustments) should only be done when all other steps to creating an effective conjoint survey instrument, effectively sampling potential buyers, fielding the study, part-worth utility analysis, product availability adjustments, product awareness adjustments, and Exponent adjustments fall short of producing accurate predictions of actual market shares.  It should only be done if it is absolutely necessary that the market simulator's base case shares of preference match the real world market shares.



Page link: http://www.sawtoothsoftware.com/help/lighthouse-studio/manual/index.html?hid_externaleffects.html