Know where a launch's volume comes from before you launch. Chosŏn learns from shoppers' choices which products they treat as substitutes, so a new variant takes share where it really would: from a competitor, from your own line, or from shoppers who would have bought nothing.
A new flavor, a bigger pack, a premium version of your bestseller. The survey says shoppers like it. The question that decides whether it pays is where its sales come from.
Most conjoint software assumes a new product takes share from every product on the shelf in proportion to its size, or by one similarity rule set before the study. A new pack of your bestseller then looks like it takes from everyone.
Someone who buys your 250ml can is the first to switch to your 500ml can of the same drink. Near-copies compete with each other first, and the rest of the shelf second.
Overstated incremental volume, a line total that adds up to more than the market will give, and a launch that mostly moves your own customers from one product to another.
The same choice-based conjoint survey you would run anyway. A model that reads more out of it.
From the choices themselves, Chosŏn estimates how closely shoppers substitute along brand, flavor, pack or any attribute you test. No decision tree to draw up front and no similarity rule to tune.
Line extensions, flavors and pack sizes that shoppers see as close substitutes share demand instead of stacking it, so the line's total stays honest.
Add a product in the simulator and see, product by product, where its buyers come from, with a 95% range on each number.
The level of the volume and profit forecast, not only which option ranks first, because a hurdle or an ROI case is decided on the level.
Every account includes a 20-product energy drink shelf, a 1,000-shopper survey already fitted for you. The survey was simulated, so the true answer is known and shown next to the estimates.
Sign up free, then choose 20-product showcase on the Projects page.
In the Simulator, load Add Volt Berry 500ml: a bigger can of a flavor Volt already sells.
Press What if we drop it? on the new can. The chart shows where its buyers would otherwise go, under Chosŏn's model, the standard model and the truth. The standard model spreads them across the shelf. In the truth, the largest single source is Volt's own 250ml can.
The flavor or size that brings new buyers to the brand, not the one that moves your own shoppers around.
Price curves for the new product, and cross-price sensitivities that show how much each product gains when another gets dearer.
If the new pack replaces the old one for most buyers, the shelf space may be worth more elsewhere.
How much of the launch is new to the category, with a range you can defend.
No. A standard choice-based conjoint where products share levels of brand, flavor, pack or other attributes across tasks. Shelf-style tasks with many products work well. Studies of up to 5 products per task are free; larger runs fit up to 60.
Standard tools handle those. The difference shows with near-copies: similar on some attributes, different on others, which is what most line extensions are.
Every result has a 95% range, every fit is checked on choices it did not see, and the tutorials and showcase show the known truth next to the estimates. We have tested it in controlled studies with known answers; see the ranging case study. If you have a real launch to compare against, become a beta tester.
Small studies are free. Larger studies run on the free trial, academic and beta tester routes. See the four ways to start free.