Pricing
Monetised incentive
To ensure term-neutral prediction behaviour, IQUO uses dynamic quote pricing. The price grows toward snapshot time to compensate for implicitly increased precision of prediction.
As we discussed it earlier,
Brownian motion deviation ( t ) = [ constant deviation ] * [ square root ( t ) ].
If we have time t left until snapshot time, uncertainty is proportional to square_root (t), thus gradually reducing with time. To compensate for this, we can introduce a time-dependent multiplier that will be applied to quote price:
[ base ticket price ] * [ time-dependent multiplier ], or, using model horizon:
[ base ticket price ] * [ square root ( [ model horizon ] / ( [ snapshot time ] - [ creation time ] ) ].
The chart below demonstrates multiplier curve. If quote duration is equal to model horizon (28 days), the multiplier is 1.00, then it gradually grows up until last week to something around 2.00, with steep growth during the last few days up to 18.33 at cut-off time (2 hours prior to snapshot time).
Last updated
