Fixed Competitors vs. Sampling of Scenarios
The Fixed Competitors Simulation Approach
When the competitive set is known, the researcher can specify fixed characteristics of competitors within a simulation scenario. The full set of competitors doesn’t need to be specified; but the important players in the market should be represented. For example:
If the WTP product has either fixed (in the case of "JVC" as the brand) or default levels (as the fourth parameter of the Range function for the other attributes) and if the competitive products are also given fixed specifications, this makes moot the issue of how many market scenario sets need to be sampled. Only one possible base case WTP product and set of competitors is involved, so it isn’t necessary to repeatedly sample across multiple WTP products or competitors.
The Sampling of Scenarios Approach
The Sampling of Scenarios (SOS) approach is useful when you do not have a good sense for your base case product specification and/or to incorporate uncertainty in competitor composition and reactions (e.g., what if competitors also come out with similar enhancements?). What makes a sampling of competitors approach different from the fixed competitors approach is that we repeatedly sample among randomly distributed competitive positioning.
We can use Range() functions to specify which attributes should be sampled for the WTP product and also for competitors:
In the example above, we want to estimate WTP for features of our JVC product on row 1, where we hold brand constant at JVC (level 1). For the two competitors in this scenario, we are sampling randomly across the other two levels of brand (levels 2 and 3).
For the other attributes, we are sampling across all possible levels included in the experiment.
This WTP estimation run will be generalized with respect to JVC's product, where we take the median of the WTP results across 1000 draws involving all possible competitor reactions and prices. (Except that competitors cannot take on the JVC name, of course.)
