How can we get optimal bundle, relative preference of different bundle like we have in CBC, attribute importance and price curve for each attribute.

What all other outputs we can get from MBC other than counts report?

Visit Sawtooth Software Feedback to share your ideas on how we can improve our products.

How can we get optimal bundle, relative preference of different bundle like we have in CBC, attribute importance and price curve for each attribute.

What all other outputs we can get from MBC other than counts report?

0 votes

Optimal bundle depends on what you are trying to optimize. You are probably thinking about optimizing revenue or profit. To optimize for profit, you need to know costs for each item in the potential bundle. So, in the Excel spreadsheet you could add some additional formulas to multiple choice likelihood times price minus cost (for profit). Excel has a Solver function for changing values of cells (prices for items) to optimize the target value of another cell (total profit or total revenue). With some extra work you could use Solver to find optimal bundling pricing from your Excel MBC simulator.

You can also use your MBC Excel simulator to derive price curves for each item in your menu. Set all prices at the base case price. Then, for each item on the menu (holding price of all other items constant), change its prices from low to high prices and record the choice likelihood at each price move. Plot those to obtain price curves per item.

The simulator within the MBC software (not the export simulator to Excel) can do "combinatorial predictions). That allows you to set prices across the menu items and then predict what are the top (say) 3, 5 or 10 bundles that respondents would be most likely to configure. Must be using HB models to make the combinatorial simulations happen.

You can also use your MBC Excel simulator to derive price curves for each item in your menu. Set all prices at the base case price. Then, for each item on the menu (holding price of all other items constant), change its prices from low to high prices and record the choice likelihood at each price move. Plot those to obtain price curves per item.

The simulator within the MBC software (not the export simulator to Excel) can do "combinatorial predictions). That allows you to set prices across the menu items and then predict what are the top (say) 3, 5 or 10 bundles that respondents would be most likely to configure. Must be using HB models to make the combinatorial simulations happen.

...