Course · 2. Choice-based conjoint (CBC)

The simulator

The simulator is where decisions get made. Build a shelf (your products and competitors', with prices), and Chosŏn predicts each product's share of choices, with a 95% range.

What you can ask

  • Shares: change a level or move a price slider; shares update.
  • What if we drop it?: where a product's buyers go if it disappears. Read the other way round, it is a product's source of volume: where its sales come from. This is the launch and range question.
  • Price curve: one product's share across the tested price range.
  • Price sensitivity: how each product's share responds to a 1% price change of any product. Off-diagonal entries show who gains when a rival gets dearer.
  • Chance A outsells B: how sure the data is about a ranking.
  • Compare with standard HB logit or the aggregate logit, to see where the answers differ.

Example: the tutorial launch

Launch Volt Berry next to Volt Original, Surge Berry, Apex Citrus and Nova Mango. With the default settings our model says about 38% of Volt Berry's volume comes from Surge Berry (same flavor), 28% from Apex Citrus, about 21% from Volt Original, and about 10% from people who would otherwise buy nothing. The data was simulated, so the truth is known: 42%, 25%, 19% and 10%. Our model's ranges contain all four; the standard HB logit's miss two. That is one simulated study, an illustration rather than a guarantee.

Limits

Prices must stay within the range tested in the survey; the simulator will not extrapolate. Shares are shares of choices among the products you put on the shelf, not market shares: distribution, awareness and repeat purchase are not in a conjoint survey.

Try it in the tutorial: the launch question

Next: The optimizer

Tutorial links open the example project in the app; it is free and needs a free account.

← Reading the results The optimizer →