Hi everyone,
I am conducting a CBC study for my master's thesis. I would like to estimate the part-worth utilities of win-back offers in the video-on-demand industry. A typical win-back offer is promoted as follows: " Do not cancel your subscription and receive a 40% discount for the next three months plus an upgrade to UHD video quality ".
My attributes are based on Netflix subscriptions. Here are the attributes and levels:
Discount: 0% / 20% / 40% / 60%
Contract duration: 1 months / 2 months / 4 months
Service upgrade 1 (number of devices one can simultaneously watch on): 1 screen at a time / 2 screens at a time / 4 screens at a time
Service upgrade 2: SD video quality / HD v. quality / UHD v. quality
Interaction effects are important to measure (specifically between discount and contract duration). I have read that Balanced overlap is the suggested method for that. However, I have also seen that complete enumeration is better for asymmetric designs. I would like to measure both main effects and interaction effects.
Could you help me out in finding the better method for my study?
Thank you and best,
Max